FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

CASES & RESULTS

FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

CONTENT LIBRARY

Link to Original content

Conversational AI

June 2023 Create connection. Value conversation.

01 Foundever overview for digital CX

02 Trends

03 Definitions

04 Pain points & benefits

05 Key selling points & Foundever differentials

06 Chatbot details

07 Voicebot details

08 Demos TBD

09 Generative AI TBD

10 Global partner: Cognigy

11 Success stories

2

01

Foundever / CX / digital overview

3

Our history

A foundation of expertise. Ever evolving.

Started in 1994 Founded with an entrepreneurial spirit…

From our origins as Groupe Acticall built by co-founders Laurent Uberti and Olivier Camino

With a heritage of supporting customer experiences for

+40 years  and grounded in proven CX expertise…

Backed by long-term shareholders Creadev, we’ve grown as a global brand, both organically and through strategic acquisition of global CX providers in 2015 and 2021

2023 ,today, we’re an innovative CX leader

Combining global scale, award-winning culture and digital know-how, we’re creating the future of CX for the world’s leading brands

Key global facts

Foundever at a glance

We’re one of the largest customer experience companies in the world

Foundever combines global strength and scale with the agile, entrepreneurial approach of our founder-led culture.

170K associates across the globe

9M customer experiences delivered daily

+750 brands partner with Foundever

+60 languages supported

45 delivery countries

57% of our team working from home

CX trends 2023

A changing customer landscape

We share the experience of +9 million customer conversations each day to turn insights into the inspiration that shapes your CX development strategy

CX needs to go back to basics

Only 49% of U.S. consumers think they’ve seen an improvement
in CX levels

Seize every
CX opportunity

62% of consumers admit social media influences their choice of brands and organizations

Develop your employee experience

29% of consumers would quit a brand that didn’t take staff welfare seriously

Build trust to build data

38% of consumers would share personal data in exchange for a personalized CX

Avoid the technology trap

Just 27% of Gen Z consumers are confident that digital forms of CX can meet their needs

Build today
for a metaverse tomorrow

15% of consumers are interested in virtual reality’s potential to improve CX

See more at Foundever.com/insights

Our solutions

Enterprise CX know-how

Designed to support your customer needs

Digital CX built with a human touch

Fully customizable Designed to integrate Secure & compliant

How we work | Performance What are businesses top customer service goals for 2023?		Grow the business Improve operational excellence Optimize costs
9 Gartner 2023 Top Priorities for Customer Service & Support Leaders Digital market And how are they planning to achieve these goals?

Digital market	Why do businesses adopt digital CX? Improve customer experience Ensure a consistent and engaging CX that drives positive business outcomes, focusing on personalization and omnichannel delivery Process optimization To identify bottlenecks in business processes, remove redundant steps, and develop efficient workflows. Strategic decision- making Data-led approaches weave together insights from multiple touchpoints to provide a 360-degree view of customers Enhance agent experience and productivity Digital CXM solutions can mitigate key agent pain points and provide agents with tools to improveproductivity Cost savings By reducing the cost of operations, deflecting voice volume to non-voice channels, removing redundant steps, encouraging self-service and more Revenue growth By enabling organizations to have greater visibility in operations, increase sales, and make more informed as well as customer-centric decisions. Market differentiation Unparalleled experience to differentiate organizations from the competition, ensuring customer stickiness and brand loyalty
Digital We mix creativity with innovation & data to assure the best results to our clients in digital channels while helping them wow their customers through great experiences

Let’s walk through…

A day in the life of a customer

That’s what digital CX is about:  Great experiences great results for our clients.  

4

02 Trends

15

Conversational AI is the hot topic of 2023 ate connection. Value conversation.

State of the market

Popularity of channels among customers before and after the pandemic

Social messaging

Over the last few years, we’ve been changing the way we communicate

Zendesk 2021

Sources: Capgemini, Facebook Insights, Gartner

Did you know that…

80% of consumers will interact with a machine if it saves them time

+70% of customers report positive experiences with chatbots

79% of consumers prefer chatbots when needing an instant response

83% of consumers would like to speak freely in self-service options

State of the market

Customers
want to have digital conversational experiences that are easy & faster

69% of customers say they’re willing to interact with a bot on simple issues (a 23% increase from 2022)

74% of customers are more likely to spend more with brands that allow them to avoid contacting a live agent

87% of customers expect an immediate response when contacting a company

By 2027 chatbots will become the primary customer service channel for 25% of organizations

Sources: Zendesk Trends 2023, AWS/ESG survey 2022

And unhappy clients don’t come back

$4.7 trillion could be lost to businesses in consumer spending, as customers refuse to spend money with businesses that don’t meet their expectations

Qualtrics 2022

74% of customers are more likely to spend more with brands that allow them to avoid contacting a live agent

AWS 2022

Vertical trends

21

E-commerce & retail E-commerce & conversational AI trends & statistics 590% rise of conversational commerce market by 2025 47% of chatbot users purchase items 58% consider “convenience” a main decisive factor, alongside “price,” when deciding where to shop $145B estimation of chatbot e-commerce transactions by 2025 40% of shoppers don’t care if they’re helped by an AI or human, as long as their question gets answered -$80B reduction of agent labor costs by 2026 generated by Conversational AI and Generative AI   Source: Juniper Research, Facebook Insights, HubSpot, Gartner

E-commerce & retail

Why conversational commerce?

Digital salesperson

A strong conversational commerce strategy will allow agents not only to be the voice of the brand but also a salesperson for the brand, in order to convert prospects into customers or customers into returning customers.

Online shopping

Since the pandemic, much of our lives have moved online, and new consumer patterns have been established. According to McKinsey, more than 75% of consumers who turned to online shopping did not return to pre-pandemic shopping patterns.

Omnichannel experience

Services and support that customers used to get in stores migrated online and now exist as chatbots and virtual on-demand assistance — so they’re becoming more and more important.

AI- powered analytics

Recording customer interactions and analyzing them with AI technologies allow businesses to find the best communication and sales scenarios, as well as monitor the general level of satisfaction.

Proactive personalization

AI and machine learning technologies allow retailers to create comprehensive customer profiles and offer products and services based on their preferences and anticipate their needs. Thus, customer experience changes from reactive to proactive.

Source: VTec

E-commerce & retail

Virtual influencers Big numbers

Why do people follow virtual influencers?

27% for their content

19% for the storytelling

15% because they inspire them

58% of users follow at least one virtual influencer

35% of users said they have already purchased a product or service promoted by a virtual influencer

35-44 year-old users are the age cluster who are the most trusting and find virtual influencers relatable

42% think virtual influencers will become mainstream in the next 12 months

30% of celebrity budgets will be dedicated to virtual influencers by 2026

24

Gartner, nfluencer Marketing Factory 2022

Healthcare

Medical chatbots

are one of the leading trends in the healthcare industry

The industry’s continuous innovation and technological advancements, along with increasing patient wait times and a lack of patient engagement, are leading to a growing adoption of healthcare chatbots.

By 2023, cost savings attributed to healthcare chatbots is estimated to be $11B

Between 2019 and 2029, the medical chatbot market is expected to grow at an annual growth rate of 26.29%

By 2027, the healthcare chatbot market is expected to reach approximately $543M

Source: Market Research Future, Botsify, Get Dear Doc

Healthcare

The growing importance of automation for healthcare and the aging market

Elderly people are a major target for healthcare

And it’s pretty hard to revolutionize the industry without impacting their lives.

The old cliché of elderly people who have no clue how to handle a computer is less and less true. They are now very familiar with smartphones.

In the past two years alone, smartphone ownership among people over 65 years old increased from 53% to 61%.

Longevity economy

The so-called longevity economy, which serves the needs of Americans over 50, currently has a market size of $7.1 trillion.

As baby boomers continue to age, by 2032 the market is forecasted to grow to more than $13.5 trillion.

Medtech

The future of medtech is about connectivity and data

Data is the lifeblood of healthcare

Realizing the vision of a connected health system — which provides better care, value and population health — requires a system driven by data.

Data is the lifeblood of that system; without data, you can’t make decisions.

Medtech must overcome data connectivity barriers

The vision of a connected health system is exciting, but the reality is not yet there. Data and analytics will create value, but the problem is that it’s still in so many silos, in so many disparate places.

Experts advocate for industry-wide standards to drive data sharing — standards for interoperability that allow data to be taken, normalized and pulled together.

Medical devices are becoming connected

Connectivity provides the ability to extract data from a device. That data can be reviewed and interpreted to derive insights, influence therapy and eventually provide new types of services.

Experts envision a future of “truly connected health” and “hospitals without walls,” which eventually leads to ubiquitous personal care, with care delivered anywhere, anytime, on-demand.

Using data to
create a holistic view and drive action

If these barriers can be addressed, silos broken and data shared, we can have a holistic view of the patient. This holistic data-driven view provides the potential to fundamentally disrupt how healthcare is delivered.

Instead of providing reactive and episodic care, care can be proactive, preventative, and personalized. This shift to prevention can involve predictive algorithms and decision support tools to assist clinicians.

Will medtech become more commoditized?

Experts see a future where medical devices become more commoditized, with similar core functionalities, a future with smart homes, smart hospitals

and smart devices — all driven by data.

Source: Medical Design and Outsourcing

Medtech

4 key challenges for the medtech sector in 2023

Ensuring innovation throughout the entire organization

How can you effectively embed innovation throughout the whole organization? And how can you create a culture of innovation? To be at the forefront of change and lead the market toward the future, you need a clear innovation strategy.

Managing data efficiently and ethically

Management of patients’ data should be used to grow products or services, optimize user experiences and keep both patients and specialists content and loyal. But privacy concerns and local regulations need to be adhered to, and collecting and using data requires specific expertise.

Expanding offerings from product to platform

A challenge of the medtech sector is the shift from selling one-off products to providing integrated service solutions that generate recurring revenue opportunities.

Putting a focus on consumers

The way companies communicate is important but so is how, where and how often they interact with their target consumers.

Source: Nextrday

03 Definition

One-pager Redefine customer experience through AI-powered conversations A comprehensive solution combining experts, strategy and tools to empower memorable customer interactions across all your social channels as well as review websites Your challenges • High volume of live support through chat and voice channels • High volume of low-value requests, with no automated channels • Unable to manage peaks in call volume during online campaigns • Low customer satisfaction due to long waiting times and low resolution Solution: How we can help • Immediate response 24/7 and full automation: Chat and voice sessions contained from end-to-end • Scalability: One solution in different channels and different languages • Sales boost: From post-sales support to proactive lead generation, converting users to customers • Customer experience & brand loyalty: Quick, seamless, customized and human-like experiences aligned with your brand goals • Specific needs & use cases that go beyond support: Bots for employee experience, campaign engagement and/or peak seasons • Powerful analytics & insights: Complete analytics and reports with customer behavior insights to help better decision making		Key benefits For your customers • Empowerment to solve problems on their own • Enhanced experience on all channels • Availability 24/7 — no more waiting in queues • Reduced frustration and resolution time For your business • Chat and voice sessions contained end to end • Lower cost to serve • Increased customer stickiness and long-term loyalty • Improved customer retention and lower customer churn
31	FoundeverTM Convternstational AI Leverage human-like conversations to redefine customer experience, improve results and save costs

Conversational AI

What is FoundeverTM Conversational AI?

For users, it means empowering them to solve their problems and enhance their experience through conversations on all channels.

For companies, it means solving their main contact requests, deflecting human contact, increasing sales and saving costs.

Conversational AI

Quick, human-like and on all channels

Different needs, audiences and moments can demand different channels — we’re ready to provide conversational experiences in all of them

Chatbot

Easy, efficient chat conversations directly from customer-preferred channels

Voicebot

Provides a natural calling experience and immediate vocal responses to customers’ common queries

Conversational AI

Truly natural and human-like experiences

Voicebots

Foundever voicebots bring together a compelling combination of conversational AI, native voice connectivity and prebuilt contact center integrations to help you deliver engaging, sophisticated voice experiences at scale, driving

Chatbots

Built on deep learning technology and augmented with Large Language Models (LLM), Foundever chatbots deliver superior intent mapping, entity recognition, and contextual mapping, gracefully handling every customer request and navigating non-linear dialogues with dynamic context switching.

AI-powered NLU engine

Cutting-edge natural language understanding for human-like, memorable service experiences. Powered by advanced machine learning and neural

machine translation technology to offer outstanding recognition accuracy in any language.

04 Pain points & benefits

Customers want fast, easy channels but it’s not just that.

Conversational AI brings incomparable solutions to the main pain points of customer service.

.

.

There’s a reason why. Conversational AI can be the answer to many customer service pain points High volume of contact Limited hours of service High cost per contact High average handling time Low rate of first contact resolution (FCR) Peak seasons

Problems that can turn into: Poor customer experiences Unhappy clients

Pain points

Operational problems can turn into expensive, slow and poor customer experiences.

High volume of contact

Limited hours of service

Peak seasons/
high operational costs to cover seasonality

Low rate of
First Contact Resolution (FCR)

High cost

Long waiting times and/or average handling
time (AHT)

High volume of repetitive and low- value tasks

Low customer satisfaction

3

Benefits & outcomes Leverage conversations to improve results
Availability and agility • No more queues — 24/7 response on all channels • Solutions customized to where your customers are and how they like to interact, by voicebot or chatbot • Agility support from simple tasks to complex situations • Reduced agent effort and reduced handling time Sales boost • Chatbots and voicebots that go beyond post-sales support and convert users to customers • Proactive bots for lead generation in inbound and outbound conversations • Strategies and personalization of conversations to leverage sales or cross-sales opportunities • Agentassisttosupportsalesbyanagentwhenneeded Customer experience Cost & brand loyalty Happy customers become loyal customers: • Quick,seamlessandeasyexperience savings • Sessions contained by the bot from end to end: fully automated Deflection of live chat- and voice-based interactions • Agent usage reduction — fewer FTEs to cover the same sessions • Continuous improvement for better customer engagement and results • Easy scalability to different channels Powerful analytics & insights • Completeanalyticsandreportswithpersonalizedcustomer behavior insights • Sentimentanalysisandvoiceanalysis • Powerfulinsightsthathelpwithbetterdecisionmaking • • Human-like and personalized conversations Brand personas that translate brand personality and generate engagement Specific needs & use cases Different chatbot and voicebot uses for different needs: • Peakandpromotionalseasons • Improvingcampaignengagement • Employeeengagement • Specificsupportareas
	Benefits & outcomes We leverage conversations to improve results Availability & cost savings Immediate response 24/7 and fully automated: Chat and voice sessions contained from end to end Customer experience & brand loyalty Quick, seamless, customized and human-like experiences, because happy customers become loyal customers Scalability One solution in different channels and languages Specific needs & use cases Going beyond support: Bots for peak and promotional seasons, employee experience and campaign engagement Sales boost From post-sales support to proactive lead generation: Chatbots and voicebots that convert users to customers Powerful analytics & insights Complete analytics and reports with customer behavior insights to help better decision making

Benefits & outcomes Expected outcomes When leveraging conversational AI Reduced demand for agents More efficiency for agents Containment of spikes in volumes Compliance & consistency of content Superior customer experience Opportunities to sell, cross-sell and upsell
Benefits & outcomes Unlock your CX with conversational AI Lower cost to serve through best-in-class AI capabilities, agent productivity and increased customer experiences Accessibility Cost Superior Agent Sales on all channels savings CX productivity support

Benefits & outcomes

Creating value for all

Because happy customers make happy businesses

For customers

Empowered to solve problems on their own Enhanced experience on all channels Availability 24/7, no more waiting in queues Reduced customer frustration and resolution time

For businesses

Chat and voice sessions contained from end to end Lower cost to serve
Increased customer stickiness and long-term loyalty Improved customer retention and lower customer churn Increased FCR

For agents

Equipped with useful information and actionable insights

Improved agent satisfaction by eliminating monotonous, repetitive tasks

Real-time feedback

Benefits & outcomes Going beyond support Chatbots and voicebots can do much more than customer support. Together, we can define use cases, align with your business goals and achieve the results your company needs.

Promotional seasons A surge in customer inquiries doesn’t need to mean increased costs Marketing and campaigns Engage customers and employees in important campaigns Conversational AI Customer support Provide immediate support through engaging conversations Sales boost Increase conversion through personal shopper and personalized recommendations Specific use for verticals Each industry has their own specific needs — conversational AI can help them all

Benefits & outcomes

Conversational AI through the customer life cycle

Consideration

• Information about products
• Personal shopper
• Proactive messages
• Capturing customers’ positive/negative emotions

Purchase

• Moving customers toward purchase

• Step-by-step assistance • Personalized offers
• Identifying opportunities to cross-sell and upsell

• Identifying leads • Outbound calls

Awareness

• Target marketing campaigns • Segmenting leads

Post-call support

• Capturing feedback
• Post-purchase information
• Automated image analysis for exchange requests
  • Personalized offers for new purchases

05 Key selling points & Foundever differentials

But if everybody is all about conversational AI, why choose Foundever?

The Foundever way

Leading-edge technologies

Best in-class AI foundation in partnership with leading global platforms that can be integrated with CRM, generative AI, natural language processing, machine learning and more

Team of CX experts

+30 years of experience championed by the passion and knowledge of our digital experts

Security & privacy compliance

In fast-changing times, trust and safety are cornerstones for scalable growth

Customer-centric and result-driven

With a proven track record and results that can be better than market average

Because we know that technology doesn’t work alone

Leadership & expertise

Leaders in CX for some of the world’s most amazing brands

End-to-end solution

Including assessment, strategy, deployment, management and continuous improvement that delivers results

4 Customer experience & generative AI

At Foundever, we’re solution integrators

Global platform of

intelligent CX services

Foundever is well positioned to support brands in the next phase of

digital transformation, capable of building a real hybrid model between humans and machines to deliver a CX strategy based on artificial intelligence.

Access to data

Data from +9M conversations daily to feed and optimize the AI model

Global CX talent

Engaged and culturally diverse human talent capable of training and supervising the machine

Technology ecosystem

Connecting proprietary digital capabilities and leading technology partners

Multi-CRM experience

Proficient across multiple CRM platforms and tools used within a CX workflow

The Foundever way

Our offer is a complete end-to-end solution

Because a great car is only as great as the driver behind it

Leading-edge Technology

Results-oriented strategy & expertise  

The Foundever way

Foundever is a solutions integrator.

40 years of expertise

And proven leadership in customer experience management

More than 9 million conversations daily

Access to a level of data and intelligence that can shape innovation

Hybrid model, diversified solutions

Integration between humans and machines to deliver holistic CX strategies

We’re among a small number of companies capable of building a real hybrid model between humans and machines to deliver an integrated CX strategy

Proven record & guaranteed performance

Through a results-oriented approach and performance-based economic models

Our differentiators

In a time of easily generated chatbots, why Foundever?

Personalization & customization

Tailoring chatbot interactions to individual users can enhance the user experience

Hybrid models with human agents

Ensure higher accuracy, handle sensitive situations, and provide a more empathetic and nuanced user experience

Privacy & compliance first Essential with the increasing concerns around privacy and security

Domain expertise

Some models may excel at general conversations, but they might lack in-depth knowledge in specific domains

Multimodal, multichannel & integrated
To guarantee good results, bots need to go beyond the simple conversations, integrating different media, channels and possibilities

Strategy, insights & results
A team of experts overseeing the project from end to end

The Foundever way Premium delivery Our offer is a complete end-to-end solution including consulting and management that delivers results Leading-edge technology With the world’s best partners Continuous improvement KPI monitoring, ongoing reports and insights for enhancement Full consultancy & management CX expertise, training, strategy and deployment

Project Dashboard for improved stakeholder management and transparency Data analytics monitoring NLU weekly training Monthly reports with updates, chatbot performance results and personalized customer behavior insights Extension of the bot’s comprehension capacity by increasing the knowledge base Benchmarking and industry trends

How we work

Is your bot performing at its peak?

Optimizing and maintaining your chatbot or voicebot’s effectiveness is crucial to remaining competitive and serving your customers’ ever-changing demands.

But this is a complex challenge that requires strong expertise and customized, one-of-a-kind strategies built to your needs.

5

4

06 Chatbot details

Conversational AI by chat

We know that chatbots don’t have the best reputation, because of things like these.

Foundever Conversational AI mixes expertise and technology to ensure the best resolution and experience for the customer

Leading-edge AI
Conversational design
Full automation of simple and complex tasks Generative AI integration
Strong brand identity
Compliant with privacy laws
Multilingual

Process used as a benchmark by the client  

KPIs & results

Thi is your title –

Because KPIs are just 3 letters

If it doesn’ t bringsreal change. Bol 40pt

U.S. beaty etail

Cost Savings / More efficiency

11 months (June 2022 – April 2023)

Fashion

Lower customer effort

(Project average: 34 months)

U.S. eyewear retail

Sales generation

2021 – Q1 2023

2.5K/sec Simultaneous accesses during promotional day +55K access/day

$3,2M Savings (7 months of project)

-9%Human chat volume

90% Conversations contained by bot

95% Bot understanding rate

$1,9M Savings (2020 – Q1 2023)

+12% Average order valueof whom used the bot compared to those who did not

$1,8M Savings (2021 – Q1 2023)

+232%Conversion rate increased in one year after bot improvements

Up to 94% Customer Satisfaction (CSAT)

The best experience for customers means the best results for clients.

“Our chatbot is our most productive agent at 2.3M interactions and growing!

Since the launch, we’ve seen an average of 71% of these interactions retained within the bot, eliminating the need for customers to speak directly to an agent and offsetting contact volume.”

Customer Care Manager

U.S. retail

A digital channel designed
to understand customers’ needs

Keeping your customers happy.

Our CX expertise significantly enhances your chatbot performance — increasing customer retention
and satisfaction.

Source: 1. https://www.Businessinsider.Com/chatbot-market-stats-trends?IR=T,
improve-digital, 3. https://c1.Sfdcstatic.Com/content/dam/web/en_us/www/assets/pdf/salesforce-state-of-the-connected-customer-report-2019.Pdf,

30% FAQ chatbot handling rate1

53% AI chatbot handling rate2

85% Foundever chatbot average handling3

2. https://www.Gartner.Com/en/documents/3983091/deliver-customer-service-during-covid-19-  

Addressing your CX needs Our end-to-end, fully managed chatbot solution is created based on your customers’ needs and expectations. Meet Sarah One of our chatbots

Engage Check out our new winter campaign! Business Help My voucher credit has expired. Customer Trade I need to make a payment. Customer Grow You just received the points from your last purchase. Business   Inform What is my order status? Customer

Custom-built personas to speak your brand language

Our conversational design team immerses themselves in your brand to characterize a very detailed persona describing how your chatbot should look, think and sound.

The result is a digital brand ambassador that’s an extension of your brand experience in an always-on, automated channel of support.

Steps to build a chatbot

From a deep understanding of your organization’s culture and brand voice to the

implementation of the chatbot, our digital experts will guide you every step of the way.

4-9 weeks deployment

Analysis Deep analysis immersion

Strategy Content and UX strategy

Content Content creation

Implementation Content implementation

Integration System integration

Metrics KPI monitoring and reporting

Improvement Training and improvement

07 Voicebot Details

Conversational AI by voice

.

Voice automation can mean  two very different kinds of experience Traditional IVR Guides callers through a series of sub-menus before connecting them with a live agent Foundever Voicebot Uses natural language understanding to process customer requests and provide automated, voice-based customer self-service, sometimes without involving a live agent

It’s all about seamless technology —  as if you were talking to a live agent

Voicebot

Traditional IVR vs. Voicebot

Understanding the differences

Traditional IVR

“This is what I can do for you”

• Robot-like responses and rigid menu options
• Endless hold times
• Need to repeat the query every time they call
• Cannot get to an agent when it is urgent
• Multiple contacts and even that does not resolve the request
• Single input and single action
• Business sets the structure and flow, not the customer

Voicebot

“What can I do for you?”

• Human-like experience with Natural Language Understanding

• No wait time

• Short and long-term memory

• Seamless handoff to agents

• Integrations and automation to enable true self-service

• Capable of processing multiple inputs and actions

• Customer-driven journey

Natural conversational experience and immediate vocal responses

Our solution Integration with telephony and business systems Triggering omnichannel engagement like outbound calls, SMS and WhatsApp Huge impact for peak seasons, nonworking hours and avoiding queues Agent productivity and lower AHT thanks to prequalification and context

Voicebot

Benefits of automated, voice-based 24/7 support

Ensure loyalty, growth and profitability

Be reachable 24/7

Agent productivity and lower costs

Delight customers with high-quality CX

Support digital transformation

Sentiment analysis & customer insights

Fast and convenient

Screenless, inclusive and hands-free

Human interaction always possible

Aligned with customer habits

Personalized
and consistent CX

Feedback acknowledgement

Voicebot

Benefits of automated, voice-based 24/7 support

Your customers can now resolve multiple queries and transactions in a single automated call with ease.

Superior customer experience (CX)

Customers enjoy the convenience
of immediate, intuitive services
and smoother resolution without going through a maze of options and selections or waiting for a long time to get answers.

Support agents work

The automation of Tier 1 questions can ease staff pressure, support agents’ work, reduce AHT and lower costs. Our voicebot allows agents to remain highly engaged in their day-to-day work by allowing them to avoid repetitive and low-value phone calls.

Conversational by design

Our voicebot is configured to respond to your customers’ needs and queries to provide the best response and fastest customer support.

24/7 instant channel + cost savings

Uninterrupted customer support
that helps resolve customer inquiries autonomously without having to wait, saving on costs while freeing-up live agents to focus on more complex queries.

Aligned with new customer habits

We talk faster than we can type. Trends show that each day people are sending and receiving more voice texts and consuming more audio media like podcasts.

Sentiment analysis

Being able to capture not only what is said but how it is said is extremely valuable data. We extract deep insights from the tone, pitch and frequency of a customer’s voice to assess sentiment.

Screenless, inclusive and hands-free

Voice allows for a communication that
is connected to a screenless and IoT-based future, hands-free for all interactions, and inclusive for users with motor, linguistic
or other difficulties.

Crisis management

During a crisis, the voicebot can support your workforce by absorbing huge volumes of calls. It can bring immediate customer reassurance and resolution

to maintain trust in the company.

Voicebot

Voicebot + visual interface

Our innovation to boost sales and deflection

We customize and seamlessly blend rich, interactive micro web applications into your conversational experiences.

We let companies leverage the best of text, graphical, touch and voice interfaces to delight customers with outstanding service on their preferred channels, all within a single experience.

Best-in-class experiences use your existing interfaces and content to:

• Show and select products
• Enable authentication
• Allow payment
• Share location

Voicebot Building your voicebot Dedicated digital experts at all stages of the voicebot development cycle, from strategic assessment to continuous improvement and management. 8-12 weeks deployment End-to-end project management Analysis Deep analysis immersion & use cases identification Strategy UX strategy, interoperability & architecture Design Conversational design Implementation Activation & testing Integration Systems integration & testing Launch Go live Metrics KPI monitoring & reports Improvement Continuous training & improvements

Voicebot

How does our voicebot work?

Understanding customer intent to drive conversations

Powered by voice technologies, our voicebot can resolve customer queries in a natural conversation. By understanding customer intent, reasons of contact and pain points, it has the ability to drive customer conversations and provide real-time support for simple queries 24/7.

Our voicebot provides immediate answers for frequent and low- value queries, reducing customer waiting time and deflecting contacts for human agents.

In case of complexity or voicebot misunderstanding, calls can be transferred to live agents intelligently with smart routing during opening hours or pre-processing of the query for asynchronous support during non-opening hours and/or in case of agent unavailability.

Customer call

Transfer Availability opening hours Live agent hands over with awareness of context to finalize customer call with accuracy

Answer vocalization Immediate reply to the user query

Query vocalization Unavailability & non-opening hours  Transcription, qualification and transmission to agent for asynchronous support (call-back).

Transcription ASR/NLU Automated speech recognition (ASR) Natural language understanding (NLU)

Response match verification

Use case database match verification, Distant service interrogation (CRM), Natural language processing (NLP)

Voicebot

Outbound calls

How does it work?

Automation of high-volume outgoing calls and automatable process.

A dialer connected to the voicebot makes it possible to manage large volumes of calls in an automated way. Outgoing calls are initiated by the dialer according to a list of numbers provided, the voicebot then takes over conversations with customers. It uses natural language recognition and processing to tailor the conversation to the customer’s responses.

At any time, if the customer requests to speak to a human agent, the voicebot has the ability to transfer the call to a qualified agent. This combination of advanced technologies and the ability for an agent to take control of the conversation ensures a smooth and efficient customer experience. Routine tasks are automated, allowing agents to focus on higher-value interactions while providing proactive customer service.

Dialer

Initiates calls to a list of qualified numbers

Receiving the call by the customer

Continuation of the conversation

The voicebot engages the tour of conversation

The customer responds

Transcription

ASR/NLU

• Automatic speech recognition (ASR)
• Natural language understanding (NLU)

Transfer

Business hours / immediate availability of an agent Pre-qualified transfer to the agent to finalize the treatment accurately

Immediate unavailability of an agent
Transcription, qualification and VR taken for asynchronous processing (call-back).

Announcement of the transfer by the voicebot

Vocalization of the answer

Processing according to use cases and defined business rules

• Checking in the knowledge base
• Querying remote services (CRM, DMP…) • Natural language processing (NLP)

Voicebot Voicebot capabilities Mix and match capabilities with your objectives for the best CX outcomes Inbound Routing Efficient routing to avoid waiting time and/or customer redirection by agents Pre-qualification Prepare for and support agent handover, reducing human agent AHT Outbound Full processing & pre-processing Appointment setting, reminders, proactive information, pre-qualification and lead generation, payment reminders, debt collection…. High peaks & non- working hours Easily manage peaks of calls and remain accessible 24/7 Full processing Automate simple and common queries for human agent call deflection

Different applications for different customer needs Customer support Guidance and FAQs Order tracking and status checks Bookings and cancellations Returns and refunds Reminders, payments, collections, lead generation

Voicebot

Full consultancy

throughout your conversational AI journey

Before launch Assessing customer journey to identify and prioritize use cases,  Assessing integration with telephony, business rules and systems, Creating conversational design persona and journey(s) / flows

Get ready to go live,  Implementing your bot(s), Connecting your bot(s) into back-office systems, Building KPI dashboards

Post-launch, Centralizing and consolidating conversation and KPI data for easy visualization, Continuous improvement, Executive-ready performance reports with meaningful insights ,

“Once built, the Conversational AI capabilities must be continuously supported, updated and maintained.”

Daniel O’Connell, Research VP, Gartner

KPIs & results

Natural & inclusive conversational solution for peaks & crises

Number of conversations + Accessibility 24/7

Health Ministry

2.000 calls

Handled Simultaneously (During peak of crisis – Nov 2020 to Feb 2021)

Average Handling Time + agent productivity

Fashion Industry

1 min 30 sec

AHT reduction in 2022 for calls pre-processed by the voicebot

Number of conversations + accessibility 24/7

EU Food Industry

10.000 calls

Per hour (During peak of crisis)

08 Demos – TBD

e connection. Value conversation.

Demos | chatbot Personal Shopper Improve shopping experience, build loyalty and drive sales • Tailored and personalized shopping experience • Users can provide information about their style preferences, body type and desired clothing items • Customized recommendations based on users’ inputs • Knowledge of current fashion trends and fit guides • Helps users make informed decisions about purchases • Convenience of having someone curate a selection of items • Saves users time and effort • Allows users to explore and discover new styles • Seamless end-to-end shopping experiences with the ability to make transactions • Cross-sells by suggesting complementary items • Remembers user information to offer tailored recommendations • Helps to build trust and loyalty • Fosters long-term relationships • Drives sales growth

Demos | voicebot

Automatize a vast amount of outbound calls Engage with customers easily

Provide efficient service at lower costs Settle customer payments faster

Conversational AI: Outbound Voicebot Demo

With our outbound voicebots, built on our partner leading edge technology, created and trained by Foundever, we can easily manage a huge number of outbound calls, partially or fully automating the processes (payment collection, authentification, reminders….), improving productivity and performance while saving costs.

09 Generative AI – TBD

Research A recent study conducted by Capgemini Research showed that the majority of consumers trust Generative AI

73% of surveyed users  trust AI-generated content 53% of global consumers trust generative artificial intelligence (AI) advice when it comes to investing and financial planning 67% Ask chatbots for medical advice 66% Consult them for relationship problems 55% Gen Z users are utilizing AI for financial planning 49% Are unconcerned about risks of fake news stories

Customer experience & generative AI

Integrating Conversational AI in CX services

As generative AI and large language models (LLMs) evolve, Foundever is committed to supporting leading brands in the next stage of digital transformation by building a hybrid approach integrating both human and machine across two primary models.

• AI is directly completing the tasks of an agent — either partially or fully automating the conversation
  • Human agents become AI supervisors and trainers

AI automation 10-30% Volume automation in 3-5 years

Co-pilot model 5-10x efficiency

• Augmenting and assisting agents with insights and suggestions
  • Automating repetitive background tasks
  • Focusing human conversation where it matters most  

10 Global partner: Cognigy

Global partner

And all of this is brought in partnership with a global leader

Cognigy

Fast and up-to-date: Generative AI integrations

Our official partner Cognigy is already integrating with
Azure OpenAI / GPT-3 to enhance the experience and make the process of building chatbots even faster and more efficient.

• 125 languages managed by proprietary AI
• Voicebot
• Sentiment analysisComplementary bot / live chat without additional costs
• Multilingual, multichannel chatbot
• Easy integrations
• Unlimited scalability
• Reporting 360°

Worldwide presence, USA, Germany, Australia, Corea, UK Japon

Recognized by analysts, Gartner, IDC, ISG Named a Leader in Gartner’s® 2022- 2023 Magic Quadrant for Enterprise Conversational AI Platform

IVR + VOICEBOT VIDEO TO BE INSERTED

89 Global partner Cognigy xApps Mobile-first, micro web applications that empower enterprises to elevate and unify channel experiences with multimodal service uniting text, multimedia & voice

Messaging • Growing ubiquity • Text and visual elements • Rich media support Phone • Personal experiences • Fast and simple • Most pervasive channel Web apps • Customizable interface • Transaction-friendly • Access to device sensors Channel interconnection Enable customers to effortlessly engage and transition across channels — all without losing the conversation context. Fully customizable Tailor xApps to your business services and style to deliver a unique and consistent brand experience. We support complete white labeling. OS-agnostic xApps run in any modern web browser on any OS. That means nothing to install or configure for customers. One-stop solution xApps are tightly integrated into Cognigy.AI’s low-code ecosystem, meaning you can build and launch intuitive web apps to level up your CX — all without IT!

Global partner Agent+Assist Real-time intelligence for excellent live support AI-powered coaching and direct access to essential resources at every step of customer-agent interactions, so agents can deliver high-quality support at speed Warm handover with context Emotional intelligence with a boost Speak any language Possibility to act in seconds Drive customer-centric upsells Real-time knowledge lookup

50% decrease in agent turnover 30% CSAT improvement 15% shorter AHT

Global partner

Cognigy differentials

1  Built for scale
Run dozens of bots across multiple languages. Work with multi-disciplinary teams.

2 Low-code experience
Industry-leading flow editor with rich media previews, SSML and custom flow node extensions.

3 Powerful NLU
Fully composable NLU pipeline. Runs where you need it to, not tied to a single cloud. Extend as you like.

4Insightful analytics
Industry-leading conversational analytics with state-of-the-art tools. Truly understand your users.

5 Made for contact centers
Contact center voice is a first-class citizen with end-to-end contact center integration and extensibility. Agent assist capabilities.

6 Unlimited extensibility
Extend on your own terms with prebuilt modules, leading developer tooling and open API. Built-in integration with open AI LLM.

11 Success stories

01 Beauty U.S. retail

.

	CASE STUDY Chatbot | U.S. retail E-commerce chatbot to provide immediate support and handle peaks in contact volume Our client, a U.S, beauty, cosmetics and fragrance retailer, needed to provide instant and personalized support to customers shopping online and find a way to handle peaks in contact during campaigns. In June 2022, we launched the client’s chatbot that could assist customers by answering frequently asked questions, helping them with their orders and offering exclusive information like product discounts. By December, the chatbot was handling 800K sessions.	Client challenges • High volume of calls and messages, with no automated digital solution to handle chat  • High volume of low-value requests, including questions about the products and order statuses • Unable to manage peaks in call volume during online campaigns • Lack of a personalized online shopping experience • Bot persona to provide a fun and human-like experience — that’s going beyond the digital and becoming the ambassador of the brand •Conversational content: +600 contents to answer the customer’s main doubts • Personal Shopper with product recommendations to help increase sales • Customized integration with Zendesk and CRM, having high-volume contact reasons completely automatized within the bot: Order status API: + 37% in retention | +40% in accesses Order cancel API: +50% in retention | +135% in accesses
CASE STUDY Chatbot | U.S. retail 2022 milestones Order status API with an increase of 37% in retention and 40% in accesses SEPT Order detail API launched integrating full order info — 5 top main contact reasons JAN 23 JUN 22 Chatbot deployment OCT Order cancel API with an increase of 50% in retention and 135% in accesses With this integration, customers can cancel an order in a 100% touchless manner The first integration had such a positive impact that it was shared by the client as a benchmark for other providers		2023 roadmap Integrations • Account verification • Cancel order improvements • Product availability (warehouse & store) • Order on behalf of damaged item • Loyalty (add a purchase) New channels • Facebook New features •Avatar creation — to be used as brand ambassador • AI for product recommendation • Personal shopper • New conversational flows Based on improvements assessment matrix

CASE STUDY Chatbot | U.S. retail

2022 – 2023 results

11 months (June 2022 – April 2023)

2.9M Sessions

10.5M Customer Interactions

2.5K Simultaneous accesses during promotional day, early access during 1 sec +55K access/day

800K 2022 December sessions Foundever chatbot record

96% Chatbot understanding

70% Chatbot containment rate

+88% Chat sessions increase from 141,220 to 265,000 per month — 24/7 channel

-9% Human chat volume from 141.220 to 130.248 per month

$3.2M Estimated savings (USD) with bot containment rate

CASE STUDY Chatbot | U.S. retail

The best experience for customers means the best results for clients.

“Our chatbot is our most productive agent at 2.3M interactions and growing!

Since the launch, we’ve seen an average of 71% of these interactions retained within the bot, eliminating the need for customers to speak directly to an agent and offsetting contact volume.”

Customer Care Manager

U.S. retail

02 Havaianas

CASE STUDY Chatbot | Brazilian retail Iana, Havaianas’ award-winning chatbot Havaianas partnered with Foundever to launch a digital agent named Iana in March 2020. Besides personalized and fun 24/7 support, Iana also offers personalized shopping advice, order tracking and product exchange automation through automated image analysis. World presence 9 bots, 6 languages, 32 countries

Chatbot evolution Effective solution To help with their high volumes of contact and provide support on the brand’s website, Facebook & WhatsApp — Cost savings of $1.8M Automated image analysis Completely automatized their main contact reason (73% of calls related to exchange product request) bringing real-time resolution from 10 days to 15 seconds. Integration with Salesforce Chatbot integrated with Salesforce to create tickets automatically and automatize main contact reasons as track my order & order returns — ROI of $59.000 (24 months) for track my order Personal shopper Acts as a sales consultant, suggesting looks, gifts and presenting product portfolio — 62% sales conversion increase Brand ambassador 3D bot persona to provide a fun and human-like experience — that went beyond the digital and became the ambassador of the brand on their social media. The persona for the brand on Twitter is the Iana chatbot. 2022 results

CASE STUDY Chatbot | Brazilian retail

Project results

Global bots 2020 – 2023 (Q1)

+$1.9M Project savings (USD)

+244% Average ROI growth per year (2020-2022)

+7.4M Customer interactions

95% Bot understanding

+2M Sessions

+90% Conversations contained by Iana

+100K Peak of more than 100K sessions during Black Friday month

CASE STUDY Chatbot | Brazilian retail Advanced image analysis to automate the exchange process After a successful partnership with Foundever to launch a customer support chatbot in 2020, Havaianas defined a new mission to further enrich the customer experience by extending the bot’s capabilities. Results 86.3% accuracy Peaks of 6.7K cases analyzed per month Resolution process: from 10 days to 15 seconds 80% automation average (July 2022) 2022 results

Challenges • Makethechabotsolution,Iana,more efficient and faster in processing return and exchange requests for pairs of flip- flops • Delivercompellingefficiencywhile reducing operational costs Solutions • Leveragedautomatedimageanalysis to create a bot that can quickly and accurately respond to customer product exchange requests based on images submitted through digital channels • Fullyautomatedend-to-endexchange process from request to approval, powered by deep learning combined with AI for faster and easier support • Facilitatedassistancethroughcontacts escalation to agents whenever needed or requested by the customer

CASE STUDY Chatbot | Brazilian retail

Iana chatbot + automated image analysis

Human call deflection

With the implementation of automated image analysis for Iana, all other channels (voice, email and chat) began to direct customers with this request to the bot, optimizing human operation time and maintaining the quality of service at peak hours, as the bot has no limitation regarding the volume of contacts it can receive.

Human contact represented 96.2% of contacts in 2020

Chatbot reached a peak of 99.6% of contacts in 2021

FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

Conversational AI QUESTIONS

1. What is Conversational AI?
2.  How does Conversational AI work?
3. Can you explain the difference between a chatbot and a voicebot?
4. How has Conversational AI evolved over the years?
5. What are the key components of a Conversational AI system?
6. How can Conversational AI benefit businesses and customer service?
7. How can Conversational AI enhance customer support?
8. Can you explain how Conversational AI can improve accessibility for users?
9. What are the challenges in developing Conversational AI solutions?
10. Can you provide examples of industries that have successfully implemented Conversational AI?
11. How do you ensure that Conversational AI systems understand and respond accurately to user queries?
12. What is the role of natural language processing (NLP) in Conversational AI?
13. Are there ethical considerations when designing Conversational AI systems?
14. How do you handle issues related to privacy and data security in Conversational AI?
15. Can Conversational AI systems learn and adapt to user preferences over time?
16. What is the impact of Conversational AI on human jobs and customer support roles?
17. How do you measure the effectiveness of a Conversational AI system?
18. How can businesses measure the ROI of implementing Conversational AI?
19. Are there any best practices for designing user-friendly Conversational AI interfaces?
20. What are some common misconceptions about Conversational AI?
21. Can you explain the concept of “conversational design” in Conversational AI?
22. What is the future of Conversational AI, and what advancements can we expect to see?

FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

FAQ CHATBOT

Question 1: What is the definition of a chatbot?

End-to-end conversational solution with advanced artificial intelligence (AI) that solves the main contact reasons through digital channels, increasing customer satisfaction and achieving important call deflection rates.

Question 2: How does the Chatbot function?

Three main elements enable the creation and operation of AI-powered solution:

•           Leading-Edge Technologies: Leading conversational AI technology enables us to build and continually improve the chatbot with natural, non-linear and transactional conversations.

•           Conversational Design: Using CX data sources and our user-experience design expertise we build efficient and engaging human-like interactions.

•           Leaders in CX and Team of Experts: We are leaders in CX for some of the world’s most amazing brands, and the work created by our team of experts is recognized by top world awards.

Question 3: What is the difference between Decision-Tree based Chatbots (FAQ) and Natural Language Processing (NLP) based Chatbots?

Decision-Tree Based chatbots, also known as “Rule-Based” chatbots, are a series of pre-defined rules to drive visitor conversation, offering them a conditional [if/then] statement at each step. Decision trees can also replace general FAQs.

Natural Language Processing (NLP) based chatbots, or AI chatbots, use machine learning to understand the context of unstructured inputs from the visitor. The bot in this case provides them with a response through actual understanding of the visitor’s input.

Question 4: What are the key benefits of having a Chatbot?

Having a Chatbot has many benefits; here are the essential ones:

•           Customer Experience Enhancement

•           Human calls deflection

•           Cost reducing

•           Unlimited hours and capacity of service (availability 24/7)

•           Automation of repetitive tasks – from simple to complex ones

•           Reduction of AHT

•           Increase first contact resolution (FCR)

•           No extra costs to cover seasonality

•           No more queues

•           Multilingual and multichannel capacities

•           Solve the main contact reasons

•           Increase in CSat /NPS levels

Question 5: What are the most relevant KPIs?

Metrics will depend on your Chatbot strategy. These are some of the recommended ones:

·           Handling Rate/Transfer Rate

·           Client Satisfaction (Feedback/CSAT/NPS)

·           Bot Understanding

·           Session Volume

·           AHT

Question 6: How can a Chatbot help me to deflect human support?

Reproducing the agent requests into the Chatbot flow can identify resolution patterns and automate responses, deflecting unnecessary calls. Call deflection improves productivity, allowing agents to focus on added value conversations while saving time and money.

In addition to call-deflection, Chatbot consolidates users from multiple channels such as websites, social networks and other conversational apps such as Whatsapp

Question 7: Why should my company have a Chatbot?

To be where the customers are, to enhance the experience by offering 24/7 support. The Chatbot solution will provide an instant response to the user with fluid and targeted experience, including instant transactions, personalized journeys, through scenarios where they originally needed a human representative.

Question 8: What Chatbot’s unique selling points?

• We are leaders in CX for some of the world’s most amazing brands
• 10 years of experience in chatbots and recognized by top world awards
• Our offer is a complete end-to-end solution including consultancy and management, potentializing the results
• Amazing results than can be better than market average

Question 9: Why should I choose Foundever to build my chatbot?

Foundever Group offers a complete consultancy solution that complements the mere technological implementation. The Chatbot solution is backed by a skilled group of experts, UX strategy, best practice inclusion, and content creation capabilities to build the strategy and architecture to reach and exceed every goal. Foundever Group provides an end-to-end solution, from the setup to maintenance offering continuous improvement.

Our Chatbot solution can have multiple approaches. Besides answering the main reasons for contact,  HR chatbots will engage with employees and candidates, while E-commerce chatbots focus on increasing sales. 

Question 10: What is a content element?

A content element is each answer given by the chatbot. A flow is a set of content elements or intents about a particular subject or category that are part of the conversation. 

Question 11: Why is it important to incorporate Persona and Tone of Voice when designing a Chatbot?

Persona & Tone of Voice can give the Chatbot a personality that follows the brand’s identity and resembles a person more than a robot, transforming the conversation into a truly human-like experience. These increase engagement, build rapport and emulate empathy.

Therefore the user creates more empathy with the Chatbot, increasing the engagement.

Question 12: What is tone and voice?

Tone and voice are how the Chatbot persona comes through vocally in a conversation. It’s not about what is said but rather how it is portrayed, as it generates an impression on your customer. 

Question 13: How much can I personalize the Chatbot persona’s visual identity?

It is possible to create an illustration (avatar 2D, 3D, static or motion) that shows the appearance, style, gestures and dress code. When it’s used in advertising, it can also change the components according to a specific situation or campaign.

Question 14: What’s the value of including a persona and tone of voice in a more traditional corporation?

It is important to have a persona and tones of voice to transform the conversation into a human-like interaction and increase engagement. If the company is more formal, this persona’s tone and manner will project a more adequate response. The objective is to speak the brand’s language.

Question 15: What are the costs of having a Chatbot? 

The setup includes the configuration of the platform and homologation until its launch, including in-depth analysis, content strategy and production.

After that, a monthly fee applies for the operational maintenance, analytics, AI training and content production for continuous chatbot improvement. 

In addition, there is also a platform cost depending on the number of sessions. Third-party Integrations are tailor-made based on your brand’s specific needs to ensure data consistency and performance.

Question 16: What types of analytics are generated by the Chatbot?

Based on the performance of the Chatbot, there are three types of reports to choose from:

Standard Report

• Main KPI´s Monitoring 
• Onepage KPI´s Report (Monthly)
• Main questions without answer (new content opportunity)

The Complete Report

• KPI´s Monitoring
• Quantitative and Qualitative Analysis Report Presentation (Monthly)
• Insights for improvement

Premium Report
All the data from the Complete Report, plus:

• Analysis Report including Bit.ly tracker and reasons of human transfer
• Insights for improvement
• Exclusive reports for specific campaignsWeekly Report Resume.

Question 17: How do you measure the  Chatbot’s success?

We take into account the main KPIs that should represent the average of Foundever projects:

• Handling Rate/ Transfer rate: The Chatbot from Foundever have an average of 65% to 96% sessions without transferring to a human. Each Chatbot has variables that will define the handling score. It is important not to promise these rates as an absolute certainty for other scopes.  
• CSAT/NPS: There are different metrics to measure the customer satisfaction, as: CSAT, NPS, Feedback of answers and Sentiment Analysis. It’s important to define a strategy and always measure it for continuous improvement. The best case from Foundever is Zenni, with 94% CSat average rate.
• Bot Understanding: Considering the continuous improvement and NLP training of the chatbot, it is possible to reach 98% of answers provided to the user’s requests.
• Volume of sessions: The volume means that the Chatbot has engaged. It will need to consider the volume of sessions in the digital channel.

Question 18: How do we calculate the ROI?

The ROI calculator was designed to give a more accurate estimation of the savings when implementing a chatbot solution. Considering the cost of humans and their human counterparts, the ROI Calculator gives a more conservative, but more accurate cost of the expected return of hiring a chatbot.

We gather the estimated volume of monthly sessions which will be answered by humans multiplied by the cost of FTE. The total amount is spread in monthly costs and according to the crescent handling rate expected for the chatbot along the period, we spread the cost in the month to have the annually ROI.

Question 19: Is Chatbot GDPR compliant?

Yes, working with partners that support full GDPR compliance and any other data protection laws is compulsory. We prioritize your customers’ privacy rights regarding handling personal data in every interaction with your brand.

Question 20: How do you process personal data?

 Foundever Group only processes personal data if the user voluntarily consents, as stipulated by the Global Data Protection Regulation GDPR. These include, for example, hardware, web services and software suppliers, hosting providers or data centers.

All service providers are contractually obliged to comply with data protection by means of a so-called order processing agreement. The user’s personal data will never be shared, exchanged, sold or otherwise transferred to third parties. Anything to the contrary shall only apply if this is necessary for the execution of the notified or agreed purposes and is permitted under applicable data protection law without consent, if the users have expressly given their consent or if we are legally obliged to do so.

Implementation

Question 21: How long does it usually take to deploy a Chatbot?

It will mostly depend on specifics regarding projects’ deliverables. The first step is the Assessment, a deep dive analysis into the client scenarios. It can be a Standard Assessment (1 week) or Premium Assessment (2 weeks). Later, the persona and tone of voice are created or adapted. This process can take between 2 days to 2 weeks. The chatbot strategy comes next when the architecture design and flows are structured, usually taking 1 week. Then, the content creation and deployment that includes the creation, implementation and testing of the content pack contracted will take from 1 week to 1 month. Finally, the content creation and deployment, including the creation, implementation and testing of the content pack, that can take from 1 week to 1 month.

A Chatbot project can take between 1 and 3 months. In the case of third-party integrations, the technical team validates and estimates the development duration.

Question 22: Does the Chatbot need a specialized team to be implemented?

Yes, the main factor of a successful Chatbot is its experienced team of UX strategy specialists. The core team includes the following profiles:

• Project Manager: Responsible for ensuring the project deliverables are within the agreed deadline and in accordance with expectations.
• Creative Leader: Creatively assists the UX writer and ensures the best UX strategy and conversational design.
• UX Writer: Creates customized conversational AI, personas, tone of voice, and manages the platform (from implementation to maintenance.
• Tech Support: Analyzes systems and assists in their integration process. 
• BI Analyst: Analytic profile that monitors the KPIs, analyzes data and develops reports.

Question 23: Can I have a Chatbot on WhatsApp?

To have a Chatbot on Whatsapp is necessary to have a business account on the channel. Foundever Group will manage the process with the broker. The steps for enabling it are:

• Filling out the approval form;
• Business verification – Facebook Business Manager (only for unverified cases);

ID and validation – Activation code submission.

Technology

Question 24: What are some of the bot-specific terms I should know?

Intent: It is the reason that leads the visitor to the site; This is embedded into the questions asked by the visitors.

Entities: Entities are the couriers of intent, they give context to intent by adding dates, places, people names, brands or any other information that might be necessary to clarify further why the visitor is looking for a specific item.

Natural Language Processing (NLP): Computers do not think like us. To make them understand intent or even entities one needs to teach how humans talk, thus the field of Natural Language Processing (NLP). It gives the computers tools to understand how humans speak and their intentions behind a given sentence or phrase.

Machine Learning: It is a field of computer science that aims to make programs that learn and evolve over time. Think of it like a child learning how to talk. By exposing the child to an environment with many stimuli the child learns how to speak by themselves. The same can be said for Machine Learning: by feeding the system with a considerable amount of data, it can learn to better interact with humans.

Decision Tree: Another part of machine learning, a decision tree, can be described as ‘’If A happens, I will do B, Otherwise I will do C.  The sentence “If it is sunny, I will go to the beach, otherwise I will stay at home” can be modeled by a computer as a decision tree allowing the system to “think” in a more human-like way.

Question 25: How do we select the technology partner?

 The best partner would be the one that better accommodates your needs.

• Aivo: Main partner of chatbot products and the most cost-effective platform. It is a solution with advanced features for a complex journey integrated with other systems (CRM). It is relevant, especially in America. It has native integration with most digital channels and is a great platform when considering Whatsapp (they are Whatsapp brokers).
• Cognigy: It is a complete platform regarding voice bots, but it also provides excellent functionalities for a chatbot solution. Offering both channels together is highly cost’effective, but it is also a new partnership with no current project in the Foundever knowledge base. The system has proprietary AI in more than 125 languages and it is very applicable in Europe.

Question 26: What main features can I leverage with Chatbot?

Chatbot will provide interactions via messages enriched by images, buttons, emojis, carousel, video, digital documents, gifs, hyperlinks, avatar, window customization, satisfaction survey and others. Depending on the communication strategy chosen, it can be integrated into CRM to help customers exchange products, find stores and find other information.

Question 27: Is the environment PCI Compliant?

The platforms have different ways to assure security about sensitive information. Even thou they do not have PCI certification, it is possible to configure when certain sensitive data needs to be masked or deleted from the system.

Question 28: What are NLP and NLU?

Computers do not think like us. For them  to understand intent or entities, they need to learn how humans speak. Natural Language Processing (NLP) gives the computers tools to understand how humans talk and what  areare their intentions are behind a given sentence or phrase.

NLU, short for Natural Language Understanding, is a subfield of NLP concerned with teaching machines to identify, extract, interpret, and understand textual and non-textual forms of natural human language. The role of NLU is to create a valid interpretation of the data given, much like a human would write an essay given a piece of text.

Question 29: What is an AI chatbot (Artificial Intelligence chatbot)?

Artificial Intelligence is the part of computer science that tries to make the computer think like a human. Traditional computers can only follow instructions; machine learning and AI are the fields that make the program evolve over time and go beyond their set of initial parameters. A chatbot enriched with AI allows the user to use free typing at any time during the conversation without the necessity of following the flow. Once the user makes a question, the bot will answer straight away about this specific intent, positively impacting  customer satisfaction.

Question 30: Why do we use proprietary AI instead of AI like Watson,  DialogFlow,  etc?

The bot understanding from standard AI has not the same performance as the proprietary AI due to a better understanding of context, typing errors and local idioms. To assure the best AI accuracy of our partners, we tested and got proof of their effectiveness.

Question 31: What languages and alphabets are supported? 

Aivo supports proprietary AI for English, Portuguese (Portugal and Brazil), Spanish, German, French and Italian.

Cognigy supports over 125 languages: Afrikaans, Albanian, Arabic, Azerbaijani, Bahasa Indonesia, Bahasa Melayu, Belarusian, Bengali, Bosnian, Bulgarian, Catalan, Chinese (Traditional / Cantonese), Croatian, Czech, Danish, Dutch, English (Australia, Canada, India, UK, US), Estonian, Finnish, French, Georgian, German, Greek, Gujarati, Haitian, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Irish, Italian, Japanese, Kazakh, Korean, Latin, Latvian, Lithuanian, Macedonian, Malay, Norwegian, Persian (Farsi), Polish, Portuguese (Brazil, Portugal), Punjabi, Romanian, Russian, Serbian Sicilian, Slovak, Slovenian, Spanish, Sundanese, Swedish, Tamil, Thai, Turkish, Ukrainian, Urdu, Uzbek, Vietnamese, Laothian, Sinhalese.

Question 32: What systems can be integrated into the Chatbot?

It is possible to integrate with Back Office Systems and CRMs like Zendesk, Salesforce and RightNow. Also, with Genesys, Avaya, and Payment Gateways. Integration with other systems is supported by a team of digital experts.

Question 33: Is it possible for the data not to travel outside a specific country?

Yes, it is possible. The client can choose a server provider to store the data inside its country.

Training

Question 34: How is the maintenance and training done?

Maintenance includes KPIs monitoring and the training to improve the performance of the Chatbot. We keep a closer look and gather insights to create, update and optimize content. Your chatbot’s training consists of testing and analyzing the conversations to add new ways of asking to increase NLP knowledge. Data analysis can indicate topics that users recurrently ask for, prompting the need for new content and continuous improvement.

Question 35: Can agents or the client be in charge of the ongoing training of the chatbot?

To have the best training and the AI accuracy of the Chatbot, we recommend experts make it. And for this reason, we do not sell the setup from the monthly maintenance separately. It is essential to achieve the continuous improvement promise.

Question 36: If the bot has machine learning, why do I need someone to train it?

The chatbot’s platform never works without human contact. Even the most advanced machine learning-driven systems need experts adding nuances. Human training offers control over what the customer wants the bot to respond to a particular customer request.

Question 37: How do we use Analytics to improve bot quality?

Like with any aspect of a successful business, having chatbot analytics in place allows you to keep track of your goals and make continuous improvements along the way. With chatbot analytics, they are developed to employ a self-learning process that takes systemic records of data, metrics, preferences, and trends which eventually help chatbots in monitoring user interactions, and adapt relevant responses accordingly. This allows a significant role in data analytics, and for these reasons are development companies always on the lookout to leverage best strategies, tools, and technologies in business.

Chatbots play a significant role in data analytics, having a direct impact on the following KPIs:

• Handling Rate/ Transfer rate: This rate shows the number of contacts through the chat that did not need to be transferred to a human.
• Feeback/CSAT/NPS: The customer satisfaction with the chatbot service and answers provided.
• Bot Understanding: It says if the bot understands the user and it is well trained.

The volume of sessions: It shows if the bot has engaged and has been well tapped.

FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

FAQ – VOICEBOT

DESCRIPTION AND CAPABILITIES

Question 1: What is the definition of a Voicebot?

A Voicebot offers a natural calling experience and instant resolution with 24/7 accessibility for customers.

Modern human-machine interface powered by AI, a Voicebot “understands” customer questions and provides immediate vocal responses or transfers complex queries to a human agent. It’s an ideal and cost-efficient solution to manage a high volume of calls for common and simple queries.

Question 2: How does the Voicebot function?

Three main elements enable the creation and operation of our AI-powered Voicebot:

• A well designed UX strategy, design, and architecture
• A conversational AI and automation platform
• A contact center infrastructure

Based on conversational AI, the Voicebot flow starts with the user making a call to a predefined phone call number that is retrieved by the bot. When the user expresses its requests vocally, the AI engines detect its voice (Automated Speech Recognition), transcribe it to text (Speech To Text), analyse and detect (Natural Language Understanding) the intention of the caller to drive the follow up of the conversation and choose the most relevant answer to give back (Text To Speech), transforming it into a voice message read by the bot (Text To Speech).

Question 3: What are the main uses of a Voicebot?

Voicebot can have different capabilities:

• Routing (Efficient routing avoiding waiting time and/or customer redirection by agents)
• High Peaks and Non-working Hours (Deal easily with peaks of calls and remain accessible 24/7)
• Pre-Qualification (Pre-qualification to prepare for agent handover, reducing human agent AHT)
• Full Processing (Automatize simple and common queries for human agent call deflection)

And there are a lot of use cases for the Voicebot, being the main ones the following:

• Customer Support
• Guidance and FAQs
• Order Tracking & Status Checks
• Bookings and Cancellations
• Returns and Refunds
• Account, Billing or Payments

Question 4: What are the key benefits of having a Voicebot?

Having a Voicebot has many benefits for customers, operations and enterprises; here are the essential ones:

Benefits for Customers:

• Accessibility and inclusivity
• Effort reduction / Waiting time avoidance
• Faster resolution
• Enhanced CX and CSat

Benefits for Operations:

• Smarter routing
• Call containment
• Agent productivity
• Agent engagement

Benefits for Enterprise:

• Accessibility 24/7
• Lead generation and Contact optimization (HNO, lack of agent availability, seasonal peaks)
• Brand image and trust
• AI scalability perspectives
• Cost optimization

Question 5: What are the most relevant KPIs?

Metrics will depend on the client’s Voicebot strategy and use cases. These are some of the recommended ones:

• Number of conversations
• Deflection rate
• Resolution or treatment rate
• Number of transfers or Live Agent Escalations
• Handover rate
• Rate of automation realized
• Voicebot understanding rate
• Average Handling Time
• CSAT/NPS/Level of customer effort

Question 6: How can a Voicebot help to deflect human support? 

By automatizing the answers to the common and simple questions that generate a high volume of calls, the Voicebot induces a reduction of inbound calls to live agents, saving up time and money related to the human workforce. It allows human agents to focus on high-value phone calls and avoid low-value repetitive call answers.

If Voicebot can answer successfully to a high volume of calls, we always recommend keeping the ability to handle customer request like “I want to speak to a human” within the conversation with the Voicebot and in case it fails to answer accurately. This is a best practice, key to ensuring customer satisfaction however their experience with the Voicebot went.

Question 7: Why should my company have a Voicebot?

To be reacheable by phone for 24/7 support. In addition, a Voicebot is an inclusive solution for providing support to the customer on what remains the first contact channel: the phone. Even non-digital customers can benefit of this 24/7 support. The Voicebot solution will provide an instant response to the user with fluid experience, and if needed prepare the callback by a human agent when the call center re-opens.

Question 8: What are Foundever Voicebot unique selling points?

Strong partnership with leading conversational AI and automation platfoms covering all conversational channels (voice but also chat and messaging) for a maximum of scalability and advantageous negociated prices. Strong partnership with conversatonal AI technology suppliers (text to speech, speech to text, NLU…) and advantageous negociated prices.

CX Digital experts teams skilled and certified in Conversational AI strategy, design and UX, bot training and data analytics.

In-house experts in knowledge management for customer care to feed the AI.

In-house experts in telephony and web architecture.

In-house experts in integration (CRM, ERP, DMP…).

Ability to consolidate conversational data analytics with other contact channels managed within Foundever to provide clients with a consolidated view of KPIs.

Question 9: Why should I choose Foundever to build my Voicebot?

Foundever offers a complete CX consultancy solution that complements the simple technological implementation, adding long term improvement and management. The Voicebot solution is backed by a skilled group of experts, highly experienced in UX strategy, conversational AI, design and content creation to build the strategy and architecture to reach and exceed goals. Foundever provides an end-to-end solution, from the setup to training and maintenance, offering continuous improvement to maximize performance.

Other strong differentials:

• Persona and tone of voice design for personalisation of the experience consistent with our client brand identity,
• Team skilled in human-centered design and conversational design, key to deliver the best customer experience,
• Solid partnerships with bot solutions providers,
• CX services to manage the project end to end (from creation to long term management and improvement) and scale with other conversational channels (eg: chatbot, voice applications for the assistants like Alexa for example)

Question 10: What is a content element?

A content element is each answer given by the Voicebot for each question that the caller may have.

In addition, to train the machine/the Voicebot to understand callers’ queries, content elements also include what we call “utterances”, which is the large variety of ways and questions that the callers may use to express their needs or “intentions”.

Question 11: Why is it important to incorporate persona, voice and tones when designing a Voicebot?

Persona, Voices and Tones can give the Voicebot a voice personality consistent with the brand’s identity and convey a more human-like experience, transforming the conversation into a more natural and personalized experience. A well built persona, voice and tones convey more empathy and reinforce customer experience and satisfaction.

Question 12: What is voice and tones?

Tones and voice are how the Voicebot comes through vocally in a conversation. It’s not only about what is said but also how things are said and expressed, and the voice chosen is part of the experience proposed to the customer by the brand. Similarly to how human agents represent the brand, the Voicebot also embodies the brand, its values and is part of the global experience of the customer with the company.

Question 13: How much can I personalize the Voicebot speech voice?

It is possible to define gender, speed, prosody, accent, language… Integrations towards further customization with a human or synthetic voice created specifically for the brand can also be developed depending on client expectations and budget.

Question 14: What’s the value of including persona, voice and tones in a more traditional corporation?

It is important to have a persona and tones of voice to transform the conversation into a human-like interaction and increase engagement. If the company is more formal, this persona’s tones and manners will project a more adequate response. The objective is to speak the brand’s language and convey the brand image.

Question 15: In how many languages the Voicebot is available? And if more than one, it is available at the same time – regardless of the language the customer is calling at the moment?

Everything is possible, it’s a question of configuration. In addition, besides the proprietary languages natively available on the partner platform we can leverage 3rd party engines such as Google Dialog Flow, Amazon, and Microsoft in order to expand language capabilities.

Question 16: What are the costs of having a Voicebot? 

The client pays for the setup that includes the configuration of the platform and homologation until its launch, including in-depth analysis, conversational AI strategy, design of the experience, contents and production.

Once set-up done, a monthly fee applies for the operational maintenance, analytics, AI training and content production for continuous Voicebot improvement. 

In addition, there is also platform cost depending on the number of sessions (calls) and ASR (minute of conversation performed). Third-party Integrations are tailor-made based on your brand’s specific needs to ensure data consistency and performance.

Question 17: What types of KPI reporting are proposed by Foundever?

There are three types of reports to choose from:

Light report:

– Main KPIs Monitoring

– Onepage KPIs Report (Monthly)

Standart Report:

– KPIs details Monitoring and evolution / Previous month

– Quantitative and Qualitative Analysis Report (Monthly)

– CUX Improvement recommendation (content)

– Real-Time Dashboard Access (1 accesss)

Premium Report:

– Launch report – each week the first month included

– Monthly reporting after

– KPI´s Monitoring

– Quantitative and Qualitative Analysis Report

– Consolidation with Front data analysis: Insights and CUX improvement recommendation (content, use cases, process) 

– Conversational path analysis

– Real-Time Dashboard Access (1 access)

– Weekly KPI Digest

Question 18: How do you measure the  Voicebot success?

We take into count the main KPIs that should represent the average from Foundever projects:

• Number of conversations managed by the Voicebot
• Deflection rate: = (Number of conversations – Escalations) / Number of Conversations. This may be assessed by use cases managed by the Voicebot for granularity in the analysis
• Resolution or treatment rate: Conversation fully managed by the Voicebot with goal success / conversations
• Number of transfers or Live Agent Escalations
• Handover rate: = Live agent escalations/Number of conversations
• Rate of automation realized: Voicebot integrally managed conversation / Total of calls received by the call center
• Voicebot understanding rate: Natural Language Understanding performance
• Average Handling Time: Voicebot average handling time in seconds
• CSAT/NPS/Level of customer effort: Customer surveys

Question 19: How do we calculate the ROI?

Foundever Voicebot ROI Calculator was designed to give an estimation of the savings when deploying a Voicebot solution. Considering the average cost of human agents for a client, the ROI Calculator gives a conservative evaluation of the return on investment while “hiring” a Voicebot.

We evaluate the estimated calls handled by the Voicebot per month to evaluate the human agent Calls Deflection and/or Average Handling Time improvement for human agents (for pre-processing Voicebots) and the financial impacts related. The total amount is split in monthly amounts adjusted to the resolution rate evolution expected over the period to evaluate the annual ROI.

Question 20: Is Voicebot GDPR compliant?

Yes, working with partners that support full GDPR compliance and any other data protection laws is compulsory. We prioritize customers’ privacy rights and a reasonable approach of data collection.

Question 21: How do you process personal data?

Foundever only processes personal data if the user voluntarily consents, as stipulated by the Global Data Protection Regulation GDPR. These include, for example, hardware, web services and software suppliers, hosting providers or data centers.

All service providers are contractually obliged to comply with data protection by means of a so-called order processing agreement. The user’s personal data will never be shared, exchanged, sold or otherwise transferred to third parties. Anything to the contrary shall only apply if this is necessary for the execution of the notified or agreed purposes and is permitted under applicable data protection law without consent, if the users have expressly given their consent or if we are legally obliged to do so.

IMPLEMENTATION

Question 22: How long does it usually take to deploy a Voicebot?

It will mostly depend on specifics regarding projects’ deliverables. The first step is the Assessment, a deep dive analysis into the client scenarios. It can be a Standard Assessment (it lasts 2 weeks) or a Premium one (2-3 weeks).

Later, the persona and tone of voice are created or adapted. This process can take between 1 to 2 weeks (or several weeks for full customized synthetic voice creation).

The Voicebot strategy comes next when the architecture design and scripted flows are structured, usually taking 2 weeks.

Then, the content creation and deployment that includes the creation, implementation and testing of the content pack contracted will take from 1 week to 1 month depending on the use cases.

Finally, the calls activation to go live lasts 1-2 week(s) and the first pilot with daily analysis takes around 1 month.

The complete project starts from 1,5 months to be launched and takes no more than 3 months. In the case of complex integrations (with CRM, DMP, third party services…), the technical team needs to analyse the documentation in order to give an estimation of potential additional time needed, depending on the complexity. 

Question 23: Does the Voicebot need a specialized team to be implemented?

Yes, the main factor of a successful Voicebot is its experienced team of UX strategy specialists. The core team includes the following profiles:

• Project Manager: responsible for ensuring that the project is delivered within the agreed deadline and in accordance with expectations.
• Digital Strategist: responsible for the immersion and high potential use cases identification and KPI recommendations + potential evaluation of scalability with other conversational channels.
• BI Analyst: Analytic profile that monitors the KPIs, analyzes data and develops reports.
• Conversational Designer: responsible for building conversational scenarios and flows.
• Conversational UX Writer: Creates the UX strategy and writes the contents according to the project needs.
• Creative Leader: Creatively assists the Conversational UX Writer and ensures that the UX strategy is delivered.
• Tech Support: Analyzes systems and assists in their integration process.

Question 24: What is “Conversational Design”?

Conversational design consists of analyzing, modeling and streamlining the exchanges between humans and machines. Conversational design encompasses several activities: user research, data analysis to identify relevant use cases, ideation, conversation design, prototyping, testing, iterating and refining the experience and of course continuous improvement.

Question 25: Is it possible to feed AI with any script? Like Tech Support, Sales Support, and things like that?

Not directly that way. If scripts are already available, our design team will adapt the script with the persona and tone of voice selected, design the conversation flow based on those scripts adapting them for the interaction with the voicebot. The design process is important to guide the conversation taking into consideration the automation but also adjusting the amount of information shared with the customer for better cognitive load and capacity to remember or keep memory or trace of the information provided to them.

Question 26: Can we transfer to the right agents skilled as well (depending on the type of question from the customer at the beginning we were not able to match answers in the database)? Use the voicebot to transfer to the right agent. 

Yes, this is a matter of configuration, but we always recommend allowing the user to be transferred at any time (once it is within the opening hours an agent will take over the conversation / during non-opening hours the context will be kept, summarized and send to an agent for a call back). For specific cases, the voicebot can also identify emergency situations and automatically transfer the call to a human.

Question 27: What are the technical requirements for operating a Voicebot?

A Voicebot is an artificial intelligence (AI) conversational solution for automated support and interactions through the phone. It requires an automation platform that enables to create and operate the solution, as well as a contact center infrastructure to connect the bot.

Question 28: Can we use the client’s existing platform?

If the client already uses a platform, our team can do a deep dive into the technology and be trained to check if it meets the minimum technical requirements necessary to run the Voicebot.

Question 29: Can it be used for security authentication? Is this solution able to do voice biometric verification?

It depends on the voice authentication provider and the selected partner. Several discussions have been started with Voice Authentication providers in order to assess their solutions. At the moment, we are still investigating their offers and any partnership has been signed. Nevertheless, we are convinced that it is something that we will need to operate for some clients. Please contact the Voice SME or your solution team if you have any project needing Voice authentication and we will assess the possibilities. 

Question 30: Could we use this in PCI operations?

Our selected partners are not directly managing PCI, it should be handled by the selected payment partner if we need to connect the voicebot with a payment solution. If your client project needs it, please provide your solution team with this information and the name of the payment solution of your client and we will investigate.

Question 31: Can we link with the back-end system to share status and updates (for example, logistical, payment, and reimbursement status)? 

Yes, we can connect to client systems and also third-party API. Please indicate the name of the system we will need to interoperate to your Solution team in order to assess the complexity of the integration involved. Our selected partners offer native integration with several major distant services allowing us to build interoperability quite easily, so this information can orientate the choice of the partner.

If the targeted distant service integration is not supported natively by our partner, it can induce complexity, additional delays, and impact the cost estimate. That is why our solution team needs to have that information to assess such an impact.

Question 32: How complex is the client’s involvement in IT integration? Is the solution IT intrusive?

Our delivery team is committed to reducing the maximum solicitation for our client time and managing the complexity as far as possible. Nevertheless, we have at least mandatory meetings needed with the client in order to understand their telephony architecture and data interoperability. We try to handle the complexity on our side but depending on the use cases and the complexity of the integration targeted, the client can be solicited to be part of the implementation.

Question 33: Is there an option to build the bot for a client, but have it shelved until it is required? To allow for a faster implementation if needed (let’s say from a BCP perspective)?

Yes, we can propose this kind of option. It will induce pricing adjustment. In such a case, please mention the information to your Solution internal partner in order to allow him to take this into consideration in the pricing and timeline proposed.

TECHNOLOGY

Question 34: What are some of the bot-specific terms I should know?

Prompt: A prompt corresponds to any interaction of the Voicebot with a user. A prompt can be informative, interrogative, imperative, disambiguative… To illustrate, a prompt can be, for example, a response from the Voicebot to a user’s request or even a question to elicit a response.

Reprompt: A special kind of prompt vocalized by the Voicebot when a response is not heard or clearly understandable, usually in the form of a question.

Intent: Nn intent represents the task or action that the caller wants to perform. To express his need, query or ask… the user vocalizes it in his own way. That’s what we call an “Utterance”.

Utterances: The main ways (phrase) a customer can use to express his query or problem.

Entities: Entities are the couriers of intent, they give context to intent by adding dates, places, people names, brands or any other information that might be necessary to clarify further why the visitor is looking for a specific item.

Natural Language Processing (NLP):  To allow the machine (Voicebot) to understand intent or even entities one needs to teach how humans talk, thus the field of Natural Language Processing (NLP). It gives the computers abilities to “understand” how humans speak and their intentions behind a given sentence or phrase (utterance).

Natural-language understanding (NLU): A subtopic of NLP that deals with the machine / robot comprhension of natural language.

Machine Learning: It is a field of computer science that aims to make programs that learn and evolve over time. Think of it like a child learning how to talk. By exposing the child to an environment with a many stimulus the child learns how to speak by themselves. The same can be said for Machine Learning: by feeding the system with a considerable amount of data, it can learn to better interact with humans.

Logigram, decision tree or flows: Another part of machine learning, a decision tree can be described as “If A happens, I will do B, Otherwise I will do C”. The sentence “If it is sunny, I will go to the beach, otherwise I will stay at home” can be modeled by a computer as a decision tree allowing the system to “think” in a more human-like way.

Question 35: What is ASR, STT, TTS acronyms?

ASR: Automated Speech Recognition takes human voice as input and converts it into readable text thanks to STT technology. 

STT: Speech to Text technology is for transcribing spoken voice to text.

TTS: Text to Speech technology is for synthesizing written text to human-like voice.

Question 36: What are NLP and NLU?

NLP: Natural Language Processing combines artificial intelligence and linguistic processing. The latest generation of NLP technologies is based on artificial neural networks or simple statistical machine learning models. Learning models will have been trained on large volumes of text.

The objective can target several types of automatic processing: speech-to-text and text-to-speech, recognition of named entities (names of people, places, etc.), sentiment analysis (positive, negative, neutral), text synthesis, topic extraction (targeting text intent) or topic modeling.

In its latest iterations, the NLP relies on machine learning models aimed at developing solutions capable of capturing the meaning of a written text or an oral discourse. Today, they are used in particular by the different types of bot: chatbot, voicebot and voice application. But also by smart assistants like Alexa, Google Assistant or Siri.

NLU: Natural Language Understanding is a subfield of natural language processing (NLP) and artificial intelligence that has been researched since the 1960s. NLU focuses on understanding natural language in context.

Question 37: What is an AI Voicebot (Artificial Intelligence Voicebot)?

Artificial Intelligence is the part of computer science that tries to make the computer mimic human behavior. Traditional computers can only follow instructions; machine learning and AI are the fields that make the program evolve over time and go beyond their set of initial parameters in an unsupervised or supervised way (bot training supervised by a humain).

Our Voicebot enriched with AI allows the user to use free speeching at any time during the conversation without the necessity of following a rigid flow or script. Once the user makes a question documented in the bot knowledge base, the Voicebot will answer straight away about this specific intent, positively impacting customer satisfaction.

Question 38: Is it possible to use another NLU than the one of our Voicebot partners?

If our client already uses a NLU for other conversational project, for example, it can be relevant to also consider it for his Voicebot project.

Question 39: What languages and alphabets are supported natively by our technological partners?

Zaion supports French, Spanish and Germany and the company is expanding fast his capabilities to deal with other languages.

Cognigy supports over 100 languages: Afrikaans, Albanian, Arabic, Azerbaijani, Bahasa Indonesia, Bahasa Melayu, Belarusian, Bengali, Bosnian, Bulgarian, Catalan, Chinese (Traditional / Cantonese), Croatian, Czech, Danish, Dutch, English (Australia, Canada, India, UK, US), Estonian, Finnish, French, Georgian, German, Greek, Gujarati, Haitian, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Irish, Italian, Japanese, Kazakh, Korean, Latin, Latvian, Lithuanian, Macedonian, Malay, Norwegian, Persian (Farsi), Polish, Portuguese (Brazil, Portugal), Punjabi, Romanian, Russian, Serbian Sicilian, Slovak, Slovenian, Spanish, Sundanese, Swedish, Tamil, Thai, Turkish, Ukrainian, Urdu, Uzbek, Vietnamese, Laothian, Sinhalese.

Question 40: What systems can be integrated into the Voicebot?

It is possible to integrate with Back Office Systems and CRMs like Zendesk, Salesforce,  RightNow… Also, with Genesys, Avaya, and Payment Gateways. Integration with other systems is supported by a team of digital experts.

Question 41: How do we select the technology partner?

We define the best partner according to the client’s needs:

• Zaion: Main partner of Voicebot in Europe, supporting French language for being the most cost-effective platform. 20 years experience in Customer Relations coupled with expertise in Conversational Artificial Intelligence, plus partner on the Covid-19 business case for the French Government.
• Cognigy: It is a complet platform regarding voicebot, but it also provides great functionalities considering chat and voice solutions and high scallability in term of language supported. When offering both channels together, has great cost-effectiveness.

Question 42: Is it possible for the data not to travel outside a specific country?

Yes, it is possible. The client can choose a server provider to store the data inside its country.

Question 43: Is the environment PCI Compliant?

Today we do not process any payment inside the Voicebot so it is no need for the PCI certification. Please consult Voice SME for specific project evaluation.

TRAINING

Question 44: How is the maintenance and training done?

Maintenance includes KPIs monitoring and the training to improve the performance of the Voicebot. We keep a closer look and gather insights to create, update and optimize content, user experience and understanding performance.

The Voicebot training consists of testing and analyzing the NLU performances, the main conversations paths and train the Voicebot with undetected utterances (training phrases) to reinforce its level of accuracy and performances. Data analysis can indicate topics that users recurrently ask for, prompting the need for new content fitting with our client business evolution and continuous improvement. Training a Voicebot is a marathon, not a sprint!

Question 45: Can agents or the client be in charge of the ongoing training of the Voicebot?

To have the best training and continuously reinforce the AI accuracy of the Voicebot, we recommend experts handle this task. For this reason, we do not sell the setup from the monthly maintenance separately. It is essential to achieve the continuous improvement promise.

Question 46: If the bot has machine learning, why do I need someone to train it?

Even the most advanced machine learning driven systems need experts to understand and add nuances. Human training offers control over what the customer wants the bot to respond to a particular customer request and avoid bad training due to the limitation of machines to understand nuances in the language.

Question 47: How do we use Analytics to improve bot quality?

Like with any aspect of a successful business, having Voicebot analytics in place allows you to keep track of your goals and make continuous improvements along the way. With Voicebot analytics, they are developed to employ a self-learning process that takes systemic records of data, metrics, preferences, and trends which eventually help Voicebot in monitoring user interactions, and adapt relevant responses accordingly. This allows a significant role in data analytics, and for these reasons are development companies always on the lookout to leverage best strategies, tools, and technologies in business.

Regular monitoring of analytics seen above makes it possible to assess the progress of the Voicebot.

TELEPHONY INTEGRATION

Question 48: What is the difference between SIP Trunk and PSTN for telephony integration?

SIP Trunk and PSTN are both ways to connect to a telephone network, but they work in different ways. PSTN, or Public Switched Telephone Network, is the traditional phone system that has been in use for many years. It uses copper telephone lines to connect calls and has a limited capacity for call handling.

SIP Trunk, on the other hand, is a more modern way of connecting to a telephone network. It uses the internet to make calls, instead of traditional telephone lines. This means that it can handle many more calls at the same time and is often more cost-effective. SIP Trunking is a method of providing telephone services over the internet. It allows a company to use their existing internet connection to make and receive phone calls, instead of having a separate telephone line for each phone.

In summary, PSTN is a traditional telephone network that uses copper lines to connect calls, while SIP Trunking is a newer technology that uses the internet to make and receive calls, which is more flexible and cost-effective.

Question 49: Which one is the most secure in term of data privacy?

Both PSTN and SIP Trunking can provide secure connections for telephone calls, but there may be some differences in terms of data privacy.

PSTN, being a traditional telephone network, typically uses circuit-switched technology, which means that the connection between two people is established and remains open for the duration of the call. This makes it difficult for third parties to intercept the call and listen in. However, PSTN is less secure than SIP Trunking when it comes to data privacy.

On the other hand, SIP Trunking uses packet-switched technology, which breaks the call into small packets of data that are sent over the internet. This allows for more flexibility and scalability, but it can also make it easier for third parties to intercept and listen in on the call if the connection is not properly secured. To mitigate this issue, many SIP Trunking providers use encryption to protect the call data and ensure data privacy.

In summary, both PSTN and SIP Trunking can provide secure connections for telephone calls but SIP Trunking has the capability to encrypt data which makes it more secure from data privacy perspective. However, it’s important to note that security is never a one-time effort and it’s crucial to have regular security audits and updates to make sure your communication is secure.

Question 50: Which one is the most cost effective?

SIP Trunking is typically considered more cost-effective than PSTN for several reasons:

Lower costs for long-distance and international calls: SIP Trunking uses the internet to make calls, so long-distance and international calls can be made at a lower cost than with PSTN.

Reduced need for dedicated telephone lines: With SIP Trunking, a company can use their existing internet connection to make and receive phone calls, eliminating the need for a separate telephone line for each phone.

Increased call capacity: SIP Trunking can handle more calls at the same time than PSTN, which can be more cost-effective for businesses that have a high volume of calls.

Flexibility to scale up or down as needed: With SIP Trunking, a company can easily add or remove lines as needed, which can save costs compared to PSTN where adding lines can be more complex and expensive.

Reduced maintenance costs: SIP Trunking does not require the same level of maintenance as PSTN, which can save costs in the long run.

However, it’s worth noting that the cost of SIP Trunking can vary depending on the service provider and the specific plan that is chosen, so it’s important to do research and compare costs before deciding.

In summary, SIP Trunking is generally considered more cost-effective than PSTN due to its lower costs for long-distance and international calls, the reduced need for dedicated telephone lines, increased call capacity, flexibility to scale up or down as needed, and reduced maintenance costs.

FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

GENERAL QUESTIONS

1. How Does Voicebot/Callbot Work?: Questions about how exactly the system operates, how it recognizes users’ speech, processes requests, and provides answers.
2. What Are the Use Cases?: Questions about the situations or scenarios in which the voicebot/callbot is used, such as customer service, scheduling appointments, technical support, etc.
3. What is the Technology Behind It?: Questions about the technologies and algorithms used for natural language processing, speech recognition, and response generation.
4. Supported Integrations and Platforms?: Questions about which platforms and systems the voicebot/callbot can integrate, such as CRM systems, ERP, etc.
5. What Are the Benefits?: Questions about the advantages of using a voicebot/callbot, such as cost reduction, 24/7 availability, better customer experience, etc.
6. What does the User Experience look like?: Questions about how users interact with the system, whether it’s easy to use, whether the answers are accurate and useful, etc.
7. Personalization and Brand Adaptation?: Questions about how the voicebot/callbot can be customized to fit the brand and company values.
8. What Are the Limitations?: Questions about the limitations of the system, such as understanding accents, complex answers, unexpected situations, etc.
9. How Is Training Done?: Questions about how the system is trained to understand and answer questions appropriately.
10. How Does Technical Support Work?: Questions about how technical issues are handled, how customer support is provided in case of difficulties.
11. Is Human Interaction Possible?: Questions about whether the system is fully automated or whether there is the possibility of human interaction at critical moments.
12. How Are Updates Maintained?: Questions about how the system is updated and improved over time.
13. What Are the Costs Involved?: Questions about the costs of implementation, maintenance, and possible licenses associated with using voicebot/callbot.
14. What Are the Challenges?: Questions about the challenges faced in the implementation and operation of the system, how to deal with language variations, ensure accurate answers, etc.
15. What Are the Measured Results?: Questions about how voicebot/callbot results and success are measured, such as customer satisfaction, operational efficiency, etc.
16. difference of IVR and voicebot

FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

GENERATIVE AI QUESTIONS

CHAT GPT QUESTIONS/ANSWERS

1. How does Generative AI work?

Generative AI is a category of artificial intelligence that focuses on creating data, often in the form of images, music, text, or other types of content. It involves using models to generate new content that is similar to a given dataset. One of the most common approaches to generative AI is using Generative Adversarial Networks (GANs) or Variational Autoencoders (VAEs). GANs consist of two neural networks: a generator and a discriminator. The generator creates content, while the discriminator tries to distinguish between real and generated data. Through a process of competition and improvement, the generator learns to create increasingly realistic data.

2. What are Generative AI key principles?

Key principles of Generative AI include:

Learning from Data: Generative AI models learn from large datasets to capture patterns and features that are present in the data.

Randomness: Generative models often incorporate randomness to introduce variety in the generated outputs.

Feedback Loop: Models like GANs operate in a feedback loop where the generator and discriminator improve each other iteratively.

Representation Learning: Generative models create an internal representation of the data, allowing them to generate new content that aligns with the learned features.

3. Examples of what Generative AI can generate:

Generative AI can generate a wide range of content, such as:

• Images: GANs can generate photorealistic images of people, animals, landscapes, and more.
• Text: Language models can generate coherent and contextually relevant pieces of text, including articles, stories, and poetry.
• Music: AI models can compose new music tracks in various genres.
• Art: AI-powered tools can produce artistic styles, paintings, and visual designs.
• Video: AI can synthesize realistic video sequences based on input data.

4. How Generative AI differs from other AI approaches:

Generative AI is distinct from other AI approaches like supervised learning, reinforcement learning, and unsupervised learning. While other AI methods focus on classification, prediction, or optimization tasks, generative AI is primarily concerned with creating new data that is similar to a given dataset. It’s more about creativity and content generation rather than making predictions or decisions.

5. How Generative AI is used in everyday life:

Generative AI has a growing presence in everyday life, often without people realizing it. Some examples include:

• Art and Design: AI-generated art and design elements are being used in various creative industries.
• Content Generation: Chatbots and language models use generative AI to compose emails, articles, and other written content.
• Entertainment: AI-generated music and visual effects enhance gaming and multimedia experiences.
• Fashion and Product Design: AI can assist in creating new fashion designs or product prototypes.
• Healthcare: Generative models can assist in medical image generation and analysis.
• Essentially, Generative AI contributes to various industries by automating creative processes and providing innovative solutions.

6. Benefits of using Generative AI:

Generative AI offers several benefits, including:

• Creative Content Generation: It can automate the creation of diverse content, saving time and resources.
• Innovation: Generative AI can lead to novel ideas and designs that might not have been thought of otherwise.
• Personalization: It enables personalized content generation based on user preferences and historical data.
• Data Augmentation: Generative models can be used to augment datasets for training other AI models.
• Efficiency: It reduces the need for manual content creation and can speed up design and development processes.

7. Industries impacted by Generative AI:

Generative AI is making a significant impact across various industries, including:

• Entertainment: Generating music, visual effects, and immersive experiences in gaming.
• Art and Design: AI-driven art, design elements, and fashion.
• Healthcare: Medical image generation, drug discovery, and patient data analysis.
• Marketing and Advertising: Personalized content creation, ad design, and campaign optimization.
• Manufacturing and Product Design: Prototyping, CAD design, and creating innovative product concepts.
• Writing and Content Creation: Automated content generation for articles, reports, and social media posts.

8. Privacy concerns with Generative AI:

Yes, there are privacy concerns related to Generative AI, particularly when generating content. These concerns include:

• Data Privacy: Generative models often learn from large datasets, potentially containing sensitive information.
• Data Leakage: Generated content might inadvertently include information from the training data, raising privacy risks.
• Misuse: Malicious actors could misuse AI-generated content for unethical purposes, like deepfake creation.
• Ownership and Attribution: Determining ownership of AI-generated content and providing proper attribution can be challenging.

9. Measuring the value of implementing Generative AI solutions:

Businesses can measure the value of Generative AI by considering factors such as:

• Time and Cost Savings: Calculate how much time and money is saved by automating content generation.
• Quality and Innovation: Assess the creativity and novelty of the generated content compared to manual methods.
• User Engagement: Measure user response and engagement with AI-generated content.
• Personalization: Evaluate the impact of personalized content on user satisfaction and conversions.
• Efficiency Gains: Analyze improvements in design and development processes due to AI assistance.

10. Design considerations for user-friendly applications with Generative AI:

When developing user-friendly applications with Generative AI, consider the following:

• User Control: Provide users with options to influence and customize generated content.
• Transparency: Communicate when AI-generated content is being used to maintain trust.
• Ethical Use: Implement guidelines to prevent misuse of AI-generated content.
• Quality Control: Ensure that generated content meets quality standards and is relevant.
• Feedback Mechanisms: Incorporate user feedback to improve the quality of generated outputs.

11. Common misconceptions about Generative AI:

• Fully Autonomous Creativity: Some may think Generative AI operates independently, but it’s often guided by human input or data.
• Instant Mastery: Generative AI models require significant training and refinement to produce high-quality outputs.
• Lack of Human Input: Humans play a role in shaping and refining the generated content, especially in creative processes.

12. Examples of companies using Generative AI for creative content generation:

• Adobe: Adobe’s “Content-Aware Fill” uses AI to intelligently fill in missing parts of images.
• Runway ML: This platform provides tools for artists and creators to use Generative AI in various creative projects.
• Amper Music: Offers AI-generated music composition tools for musicians and content creators.

13. Future trends and advancements in Generative AI:

• Higher Fidelity: Improved models will produce more realistic and higher-quality content.
• Interactivity: Users may have more control over generated outputs in real-time.
• Cross-Domain Creativity: Models that can generate content across multiple domains, pushing the boundaries of creativity.
• Hybrid Approaches: Combining Generative AI with other AI techniques for even more powerful results.

14. Ethical considerations when using Generative AI for content or art:

• Plagiarism and Attribution: Ensuring proper attribution and avoiding plagiarism of existing works.
• Misinformation and Bias: AI-generated content can unintentionally propagate misinformation and biases present in training data.
• Ownership: Defining ownership of AI-generated content and creations can be challenging.

15. Applying Generative AI to enhance educational content:

• Interactive Learning Materials: Generative AI can create interactive simulations and visualizations to aid learning.
• Personalized Learning: AI can generate tailored educational content based on individual student needs.
• Language Learning: AI-powered language models can assist in generating language exercises and resources.

16. Concept of “generative design” within Generative AI:

Generative design is a process where AI generates numerous design alternatives based on a set of constraints and goals. This is often used in engineering and product design to explore innovative solutions that humans might not have considered.

17. Limitations and challenges of Generative AI:

• Data Dependence: Quality of generated content depends on the quality and diversity of training data.
• Control and Bias: Ensuring generated content adheres to desired styles and avoids biases is challenging.
• Complexity: Advanced Generative AI models require significant computational resources.
• Evaluation: Measuring the quality of generated content objectively is an ongoing challenge.

18. Inclusivity and accessibility in Generative AI:

• Diverse Training Data: Using diverse datasets can help prevent biased or exclusive outputs.
• User Input: Allowing users to provide guidance and feedback on generated content promotes inclusivity.
• Accessibility Features: Implementing features that assist users with disabilities ensures accessibility.

19. Advice for developers interested in working with Generative AI for creative purposes:

• Learn the Basics: Familiarize yourself with the fundamentals of machine learning and Generative AI.
• Experiment: Start with smaller projects to understand how different models and techniques work.
• Collaborate: Partner with domain experts to ensure the accuracy and relevance of generated content.
• Ethics Awareness: Be mindful of the ethical implications and biases that can arise in content generation.
• Stay Updated: The field evolves rapidly, so staying current with the latest research and tools is essential.

FOUNDEVER GENERATIVE AI POSITIONING

Generative AI – the next digital disruption

In the middle of the competitive shifts in our fast-paced industry, it is important that we also continue to pay attention to other disruptive trends that pose both potential risk and huge opportunity. One important trend that is very much top-of-mind for our clients, our business and our industry is the topic of generative AI and large language models, such as ChatGPT. We are already responding to many questions in this space and can expect that this will continue to grow in the months ahead.

Beyond the buzz of current media hype surrounding generative AI (which is following a similar trend to other recent topics including the metaverse), there remains a real and major step to be taken for these technologies to be capable of contextualizing a question then, answering in natural language, and then delivering a precise and up-to-date answer, or even taking one or more decisions to a customer’s account.

Our role at Foundever is to act as a solutions integrator. With the combination of our proven leadership in customer experience management, diverse digital solutions and technology and access to data from more than nine million conversations daily, we are among a small number of companies capable of building a real hybrid model between humans and machines to deliver a real CX strategy based on artificial intelligence. Delivering success is dependent on access to a lot of data – updated daily – along with proficiency in CRM tools and functions and underpinned by employees capable of training and supervising the machine. AI still requires some levels of human oversight. Importantly, the ability to decide by coordinating tools, the need for connection and empathy between brands and customers for the development of a lasting relationship; only humans can do this today.

Generative AI presents significant opportunities for our business. Automating repetitive tasks with no real added value enables us to refocus our people on the moments that matter in the customer journey – this will help to maximize the value of the talent in our teams while increasing engagement through more rewarding work.  And, as talent shortages impact us around the globe, AI also offers us a potential solution to one of our biggest challenges, addressing hiring concerns in the future by increasing the effectiveness and efficiency of available talent.

It is our vision that the next stage of digital transformation will be built around a hybrid model integrating both human and machine. For us, we see two main approaches for AI in CX:

1. The Co-Pilot model: 

In the first approach, we leverage AI to augment our agents by offering insights and suggestions and automating repetitive background tasks. This enables the agent to go faster, with more accurate data, focusing our people on taking care of the customer (to “value conversation” in line with our brand message). In this approach AI is integrated into front-line support solutions in the role of ‘Agent Assist’ delivering tasks such as summarization, completion of repetitive tasks, delivering reminders and proactively offering information to support the customer interaction. In back-office and business-support functions AI will drive efficiencies and deliver productivity improvements to maximize speed-to-delivery and cost-competitiveness.

2. AI Automation: 

In the second approach, AI is directly completing the tasks of an agent – either partially or fully automating the conversation, depending on the complexity of the process. Projections for full AI automation are expected to be 10-30% of volume within the next 5 years. Our agents, becoming AI supervisors, will increasingly answer the need for human agents to train the machine, validate decisions and supervise automated conversations.

For now, we know many of our clients remain cautious about the implications of this technology. Questions remain concerning security, privatization, access to data, the economic and energy model, not to mention the socio-economic consequences on all countries and their populations. Undoubtedly, the model will evolve in the months and years to come. As with all innovations we have experienced since founding our company – the rise of the internet, the creation of social media and the birth of artificial intelligence – emerging technologies like ChatGPT offer Foundever many opportunities to transform and move to the next level in our industry. 

Ready for the future

Our strategy remains unchanged. We must continue to refocus our team on creating value for our clients. We must continue to drive simplicity through how we work, how we are structured, and in our go-to-market approach across our services and solutions. We have already simplified our broader product range, integrating solutions together based on needs and use cases, or bringing them directly into our operational model – reducing our total go-to-market offer from almost 70 individual products to just 20. Importantly, this does not mean our capabilities are reduced – with a full and more integrated set of products and services integrated into our basic offer.

For us, we hold a unique position that enables us to offer both scale and agility to our clients. Regarding our size today, it is easier to transform an organization of 170,000 employees than +400,000, like Teleperformance and Concentrix. Additionally, our capacity to invest in digital capabilities is greater than our smaller competitors. 

2023 is the year that we complete our transformation – and there is still a lot of work for us to do this year. I need to ask each of you for your support as we continue to build a scalable, agile and digital-first leader in our industry.

Looking ahead to 2024 and beyond, we can expect the current environment to improve. As a result of the actions that we take now, next year will be a year for us to gain significant traction. As we transform our approach, our clients will also need to find new ways to drive cost efficiencies, scale to meet new customer needs and transform their approaches to adapt to a new digital world. Whether clients look to us to help them to outsource new lines of business previously delivered in-house or find ways to drive efficiencies through further offshoring of onshore work, we have all the capabilities to answer this need and create new value.

FRANCE SHOWROOM – GENERATIVE AI VOICEBOT

HARD QUESTIONS

Differentiating the IVR, the bot, and understanding how AI collaborates with each tool in building the best customer experience is a big challenge today in CX. Bot, IVR, AI, who’s who for the best results in CX right now?

Before making any recommendation, it is important to understand the current scenario of the client, the objectives and analyze very well the main reasons for contact. The voicebot with AI allows the consumer to express themselves in natural language and get straight to the point, without wasting time on menus with several options, offering a much more efficient and satisfying experience.

The voicebot can replace the IVR, being the gateway of the voice channel, solving from simple to the most complex issues in an automated way. It can also be an effective pre-qualification solution by verifying important information before moving on to a human attendant.

However, in cases where the customer prefers to use AI in a more conservative manner, the voicebot can be implemented to solve specific use cases and may be integrated with the IVR.

It is a great solution to reduce AHT and avoid service queues, improving the consumer experience.

What criteria and how should each feature be adopted and used when deciding on artificial intelligence in self-service?

In Foundever we like to use the race car metaphor to explain how Conversational AI solutions work: there’s no point in having a Ferrari without a good driver behind it. We use the most advanced technology, but it does not work alone, it is necessary to have a great team of experts and the right strategy for each client. Through a super detailed assessment, we understand the needs, challenges and opportunities of each project.  These are criteria such as contact volume, main reasons for contact, relevance of the channels to the target audience, ease of automation for those use cases, necessary integrations, processes, pain points… All this will be discovered in our assessment and will guide the decision for the best strategy.

How do you assess the challenges of being omnichannel—which still persist – on this journey?

Having a good omnichannel is a constant process, not a one-off job. Consumers expect seamless and consistent experiences across all communication channels. For this we must evaluate how well the company maintains consistency, context and quality in different and simultaneous channels. But there are several challenges that many companies face, such as sharing information between different systems with a lack of integration and difficulty synchronizing data.

Conversational AI can be a great alternative for consistency and standardization of communication and can be available in multiple channels. In addition to being easily integrated with CRM systems and other data sources.

Where is there a greater need for new technologies for self-service: text or voice? Or is it growth together?

It is a joint and constant evolution. Innovations such as Generative AI, for example, are already impacting all self-service technologies, whether by voice or chat, both by contributing to accelerate the creation of these bots, and to evolve the level of conversations with consumers. But it is always important to analyze what works best for the needs of each client, building the ideal and customized journey. Foundever is constantly seeking relevant innovations, but at the same time always in a very careful and responsible way.

Where do we go in CX when we think of AI and the roles of Bots, IVRs in this evolution?

We believe that it will be increasingly important to have a well-structured and planned knowledge base that can be consulted by generative AI and transformed into humanized conversational responses in different channels. In parallel, customized experiences that can identify the history of that consumer will become increasingly important.

In general, we are moving towards a more automated and efficient service, so AI is a great ally to raise the level of the consumer experience.

Objectively, what have these last few years taught us about digital fulfillment? What were the main challenges? And which ones still remain? 

Digital transformation was already present in our processes and was even more accelerated during the pandemic. We are active in this field, as solution creators and early adopters of market innovations, helping to create a new era for Customer Experience. 

Before the pandemic, being automated was a sign that your brand was one step ahead. In 2022, this is no longer the case, as automation has become one of the basic needs of every company. Providing end-to-end customer self-service has become a determining factor in delivering an agile, resolute consumer experience, resulting in an increase in positive brand sentiment. The method today is the problem-solving preference of 35% of all consumers and 42% of millennials and Gen Z.  

As 79%* of contact center leaders plan to invest in better AI technologies over the next 2 years, brands that aren’t planning to keep up with this pace will end up falling behind. The data proves it: By next year, companies using AI will increase their customer satisfaction rates by 25% and save about 2.5 billion hours. 

With the evolution of technology, it is possible to use platforms with advanced artificial intelligence in various channels to be where consumers are and provide what they need. The biggest challenge today is to build a personalized journey, with intelligent UX (user experience) planning that can solve, if not all, of the user’s issues efficiently.

What can be enhanced in terms of technologies, Indicators/KPIs and Interaction and experience with customers? 

With the maturity we have developed in the digital fulfillment market, we know that all this is connected. Technology doesn’t work on its own, without the UX expertise behind it. A good user experience strategy (able to understand users and respond assertively) achieves its maximum performance when put into practice with a robust technology behind it. And the success will depend on the quality of these two deliveries, aligned with the diverse needs of the client, which may be in constant motion.

That’s why at Foundever we specialize in end-to-end customer delivery, which merges the most advanced in Conversational AI with the advice of a team of experts. Another crucial point is the personalization of interactions. Our commitment to projects that speak the language of the audience, with persona and tone of voice built through a lot of quantitative and qualitative research, make the difference in each response, creating a far superior connection and engagement.

The KPIs that measure the efficiency of digital solutions will be defined according to the objectives of the business and the product, mixed with the objectives of the consumers. We position ourselves as partners of our customers at all times of the process, to promote the best experience and effective results. In quantitative analysis we have in the portfolio 98% of understanding of the bot (for a Brazilian retail company), 94% of customer satisfaction (for a North American retail company) and US$ 113k of cost savings in just the first month of the project (for a North American beauty retailer). In qualitative evaluations, we have added numerous awards in terms of innovation, global expansion, best customer experience, best case of products and services, among others.

Where should companies “target” their digital fulfillment strategies going forward? 

We know that today brands are losing about 4.7 trillion dollars every year due to bad experiences, as people refuse to consume again from companies that have offered a bad experience.  

Therefore, it is no use wanting to “surf” the new wave of the market, if the delivery is not effectively generating results for your customer. For the solution to add value to consumers, it needs to be relevant, understand their needs, and grant a resolution. And be efficient, offering resolutions quickly and easily, without the need for additional contacts. In addition to the high resolution performance, it must be attractive, with the personalized journey using the brand’s voice consistently.

It is no wonder that Foundever has developed a product focused on this, Boost Your Bot, a consultancy created precisely to analyze the problem of existing bots in order to create a roadmap of improvements to achieve this proposed ideal.

So, we would say that the strategy is to have an extremely prepared team, with analytical, technological and creative skills, to surprise consumers, increase the value of brands and raise the level of self-service in the market.

Will companies prioritize investments on CX with the rumors of recession on the air?

It’s hard to have the visibility of where to invest in difficult times, but I believe that companies that are truly customer driven know that this just has to be one of their priorities. That we can’t simply afford not to care about the customer experience, when we know that brands are losing some US$ 4,7 trillion because of bad customer experiences. It’s an investment to be made, focusing on the very near future. So yes, I believe that companies that put the customer first will find space to invest in CX.

What are the biggest challenges to achieve digital transformation?

It is really an ongoing, never-ending effort, that mixes a lot of aspects. It’s a mix of adopting cutting-edge technology with having the best CX experts to think the right strategies for the customers, with relying on data and analytics to understand better these needs, with promoting the most personalized and human-like experiences, with being a partner to the client in every step of the way. All this together, with a continuous improvement focus, will help achieve better customer satisfaction.

DIFFERENTIALS OF FOUNDEVER BOTS

There are a number of factors, some quite simple, others more complex, that can lead to a customer feeling “dead-end,” “dissatisfied,” or “lost” in a chatbot experience. And we know there are a lot of people who feel that way, because there’s a lot of poorly done bot out there, isn’t there?

And this is very serious. Brands are losing about $4.7 trillion every year due to bad experiences because consumers refuse to consume again from companies that have offered a bad experience.

It is no wonder that Foundever has developed a product focused on this, Boost Your Bot, a consultancy created precisely to analyze the problem of existing bots and propose a roadmap of improvements and recommendations.

But what to do to not have to reach this level? Like I said, it’s a lot of factors.

To begin with, you need to always focus on the consumer experience and keep in mind that the main reason people use chatbots is to quickly and simply solve their reasons for contact.

And that’s not just lip service. It’s asking yourself: what is the customer looking for? How can this chatbot be useful for him? What is the scope of our bot, and how can we best apply it? You can’t want to embrace the world and propose a million features in your bot if they won’t be applied well.

There are several other more technical or less technical factors that build this good experience. Some of them:

– State-of-the-art technology

Select a platform that uses artificial intelligence to create a humanized and fluid conversation. Today the market is full of platforms, with different capabilities and that use their own machine learning. Low-code platforms that enable content management without relying on complex developments. But we know that there are many differences between them, so you have to trust your partner. Because the performance of the bot will not depend only on the team that implements it, but on the potential of the technology. A car with a 1.0 engine even having its best driver, will not arrive at the speed of a 2.0 car.

– Transparency is essential

The user even knowing that he is talking to a bot, should always have the possibility of being transferred to a human at any time of the journey, always in a simple way. Even if he doesn’t request it, the bot should be able to identify that he’s “lost” or at a “dead end,” and offer that transfer option to him. If by chance it is outside of human service hours, the bot can give other options such as an email to contact, or ask it to leave the phone number for a human agent to contact as soon as possible. The bot even if it is built to solve 100% of the contact reasons, must prioritize the will of the user.

– End-to-end and omnichannel resolution

We know that it is not always possible, but it is very important to try to solve the entire journey by the bot. Do not keep sending the customer to other channels, where he will have to restart the whole flow. This leaves anyone lost and dissatisfied. At Foundever, we have an average retention of 85%, reaching 96% on some projects, compared to an average of 60% market retention. We know the importance of this, of trying to solve and help the user in the bot itself.

And if the customer uses more than one channel throughout their journey, let the information be omnichannel and integrated across channels.

-Integrations

The more the bot is integrated with other systems, such as CRM, where we can already identify that user and know what he needs at that moment, the less time we take from the user, and the more satisfied he will leave at the end of the experience. It’s conversational memory, a trend that only has to grow.

A very cool example of integrations that we have is that of Havaianas. After a few months of implementing the chatbot, we were able to identify that the main reason for contact was the exchange of products, and this went through a long approval process. From there, we developed an Image Analysis solution, where people can send photos of the sandal they want to change, and the chatbot automatically identifies the problem using deep learning. The evaluation is automatic, which changed the exchange approval deadline from 10 days to 15 seconds! This also shows how chatbots can solve even complex problems, not just simple and repetitive tasks.

UX and Conversational Design

– Mix of buttons and open typing field. Often, for ease of navigation, a button-driven flow is created. But even if the strategy is to guide the first customer interactions by buttons, with the main reasons for contact, it is important to always offer the open field of typing, because the customer issue may be outside of these resolutions.

– The conversational design should be customized according to the target of the channel based on an investigative process that leads exactly to the user’s response. That’s why it’s important who is behind building the bot: experts in UX, in UX Writing, in CX, in data analysis, in bot training. All this will make a difference in the experience that reaches the consumer.

– UX Writing. The information should always be easy, simple, enjoyable to read. No drowning the user with information when he just arrived. No big blocks of text that are hard to read. No trying to push products or services when it is not what the user is looking for at that moment. That’s why the “conversation starter” is so important, or the way the bot presents itself to the customer: from there it can already win or lose that customer. Introducing rich media also helps to make this communication lighter, more fluid and more engaging: images, videos, gifs, audios, image carousels…

-Persona. Here also comes the importance of a well-made persona, which talks in a humanized way, which translates the personality of the brand and connects with the consumer of that brand in a genuine way. Because then, even if the chatbot didn’t understand a specific question, the way it found to help can make a difference between the person leaving very upset or seeing it as a small mishap. We have a chatbot in France where this happens: it is the bot of Parc Asterix, one of the largest French amusement parks, which is inspired by the world of the character Asterix. For this chatbot, we created a character with all the face of this universe, it is called “Atonservix”, which means “at your service” in the language “Asterix”. He is so cute and friendly, that even when he does not understand something, people usually give a super high grade to the service. It’s small details that can make all the difference.

– Going beyond the persona, it has the voice and tones. It’s not choosing a single “tone of voice” for your bot, because no one has only one tone. I talk to my wife one way, to my co-workers another, to my best friend another way! The chatbot too: it has only one voice, which represents that persona, but can have different tones for different reasons. If he identifies that the consumer is very frustrated or upset with the brand, he should go for a more serious tone and consistent with the situation.

– Avoid being repetitive. Even if the customer has to go through the same flow several times, or if they are lost, try to do this using different ways of speaking, this can make the experience more enjoyable. Look again at the importance of the UX Writing team!

– Continuous improvement

– Tests, reports, training, data, always! It’s no use thinking that the bot was created and ready, you don’t need to update it anymore. Like any technology product, it needs to be constantly improved and updated. Create dashboards to track performance on a daily basis, see what didn’t work, what they didn’t understand, when your user abandoned the flow. Always train, update the knowledge base, create new ways to ask and answer. Only then will the bot improve over time.

– Ask for feedback. Have you done all this and yet your user doesn’t seem satisfied? Understand why. Ask for feedback and analyze this data to invest in improving satisfaction with your bot’s experience.