Skip to Content
01
Data and AI

Imagining a new era of customer experience with generative AI

Empower the next level of CX engagement.

The remarkable recent acceleration of generative AI technology has captivated the imagination of business leaders worldwide. In fact, 93% of consumer products executives have made it their top boardroom priority.[1] They recognize its revolutionary potential to create substantial value and unlock previously unreachable levels of content efficiency, productivity, and customer personalization and engagement.

We’re entering new frontiers of customer experience and moving to an era of experience empowerment. We believe the generative AI is a tool that can not only enable efficiency and enhanced creativity, but it can significantly empower both customers and employees.

What is generative AI?

While AI algorithms predominantly analyze data to make simple predictions, generative AI has the capability to learn and reapply the properties

Opening minds to new possibilities

Broadly, we see the potential impact of generative AI across four key domains: commerce, service, sales, and marketing, but the potential goes well beyond this when we consider combining the touchpoints that make up important interactions with customers, resulting in orchestrated, personalized journeys.

Today’s chatbots are notorious for their bland, often inaccurate responses to user queries. Customers can immediately recognize they’re talking to a machine. The current state of chatbots results in customer frustration, misinformation, and missed opportunities in resolving problems. Customer support costs then go up as human intervention becomes a necessary element to mitigate chatbot limitations and shortcomings. Generative AI chatbots, on the other hand, have a more sophisticated understanding of intent and can build on context through conversations. The customer will detect a human-like, empathetic approach that is almost indistinguishable from interacting with an actual person.
 
The latest Capgemini Research Institute survey revealed that 83% of 800 organizations think these improved chatbots are the most relevant generative AI application, and 63% of retail organizations say they’re using generative AI to improve their current customer service.1 But these chatbots aren’t limited to just a customer support role. Morgan Stanley, a US financial services organization, is using GPT-4, the newest large language model, to power an internal chatbot that provides employees instant access to the company’s vast archive. They can query one platform for advice from multiple knowledge sources.

The quality of service a customer receives typically depends on the knowledge and accessibility of the agent they’re talking to, whose attention may be divided among multiple screens. A generative AI “co-pilot” can support the agent by suggesting the most probable answers to quickly address customer needs. It can even detect emotion in real time and offer recommendations based on a caller’s mood. The quality of coaching continuously improves by leveraging human feedback to reinforce models. And since the learning takes place during calls, not after, quality assurance levels increase as early as on the next call. Generative AI can also help complete the after-call work by generating the follow-up letter, communication, and one-day contract.
 
67% of organizations agree generative AI can improve customer service by providing automated and personalized support.1 Outreach, a leading sales-execution platform, recently introduced Smart Email Assist, which uses the technology to auto-generate accurate and relevant email copy based on patterns detected in prior buyer-seller conversations. In other implementations, the Salesforce-owned chat app Slack has integrated ChatGPT to deliver instant conversation summaries, provide research tools, draft messages, and find answers in relation to various projects or topics.

Bespoke solutions require in-depth knowledge and training. When B2B leads create complex product and service offers, they must pull content from disparate sources and tailor it to different industries, which can take months. Generative AI can considerably shorten the process, providing direct access to product/service expertise. It can generate initial versions of proposition/sales support collateral that align with the company’s business portfolio. Once the offer is complete, a generative AI suggestion platform can advise account execs on how to address client questions and provide the most relevant information.
 
On the B2C side, Stitch Fix, an online personal styling service, is using AI to recommend specific clothing to customers. The company is experimenting with DALL-E 2, an AI image generator, to visually represent its family of products based on color, fabric, style, or any other customer request. For example, if a customer wants a pair of high-rise red skinny jeans, DALL-E 2 will generate a composite image based on these qualities to aid an employee associate in finding a similar product in the company’s inventory.

Generative AI can support organizations with expedited content creation capabilities that include image, voice, text, and video generation. It can also improve marketing strategy with advanced data analysis and customer insights. Although we don’t believe generative AI will fully replace human creativity and expertise, it can save marketers valuable time, which they can channel into crafting more exceptional campaigns. After all, it’s much easier to slightly tweak an almost complete marketing asset than it is to build one from the ground up. The used vehicle retailer CarMax is using generative AI to create fast text summaries for its car research pages. In addition to being precise and engaging, the content is tailored to rank high in search-engine listings.
 
Creative content creation typically requires expertise from agencies with specialist design tools. In a pioneering proof of concept, Capgemini has designed an AI campaign builder in which marketers can take control and create campaigns themselves. We imagined this tool in the hands of an automotive marketing department: first they select a car as the focal point for their campaign, then the features to highlight (safety, performance, space, etc.), the target audience (working professionals, parents/families, sports enthusiasts, etc.), and lastly the platform (Facebook, Instagram, Twitter, etc.) on which the campaign will run.
 
With this input, the tool generates a theme and combines images and messages while filtering everything through the company’s brand guidelines for consistency and cohesive representation. It provides several initial options for the marketer to closely examine and select from. With the asset nearly finished, a creative team can then make the final touchups and deploy the campaign in just 3-4 weeks – not the usual 2-4 months.

Clearly, generative AI can be a potent content, sales and marketing tool – and customer experience is one of the biggest areas where this technology can make a significant impact. But, as with any new frontier, there are risks. Organizations must navigate new complexities, including intellectual property risks and responsible and ethical usage, and prepare for the possibility of data leakage and irrelevant or biased outputs. An obvious risk is presented when generative AI is provided the whole internet as its data resource, meaning it draws on both safe, reliable data as well as potentially misleading or copyrighted information. That’s why delineated boundaries must be defined around relevant data sets to exclude false or misleading information and increase the quality and safety of AI-generated content. Such guardrails and other guidance are also needed to protect more intangible aspects, such as brand identity and reputation.
 
Despite these hazards, 40% of organizations have already created dedicated teams and budgets for generative AI,but most still haven’t considered how important the next step is: choosing the right advisor and solution partner. Although generative AI can create significant standalone value, it is only truly revolutionary when combined with existing capabilities. An experienced and reliable technology partner can identify the areas within the organization where its integration can bring the greatest benefits to transform the customer experience across the whole customer life cycle. They can provide the innovation, the transparency of data source and usage, and the type of features and experiences that will step-change how organizations engage with their customers, at scale.

Capgemini’s reference architecture, at a glance

Built on a strong generative-AI foundation that provides security, privacy protection, and scale, Capgemini’s robust architecture approach can bring CX use cases to life for any business domain.

True personalization through the lens of generative AI

Perhaps generative AI’s greatest capability is the hyper-personalization possibilities. Customers deal with multiple, fragmented touchpoints and inconsistent personalization at every turn. Just consider all the interactions involved in planning a trip abroad. There’s the transportation (buying tickets, securing taxis, arranging transfers), the accommodation, and everything else in between such as planning activities, making dining reservations, and managing local travel logistics. With so many interdependent elements, one disruption can have a ripple effect on the whole itinerary. Wouldn’t life be easier if someone (or something) helped manage all this? Although still a bit futuristic, we’re drawing closer to an age where generative AI, in conjunction with workflow and execution, will consolidate multiple touchpoints and act as a personal assistant for customers.

Suppose you’re on your way to the airport but find yourself stuck in heavy traffic. Not knowing if you’ll catch your flight, you open the airport’s app and inquire about available options. Generative AI then quickly assesses various factors such as your airport arrival time and if there’s a chance of a flight delay. Using voice interaction, it suggests personalized actions it can do on your behalf like prepare your shopping in advance, reserve a convenient short-term parking spot, or arrange fast-track service that allows you to speed through airport check-in.

The assistant then goes beyond merely providing recommendations. It connects the necessary workflows of separate touchpoints and coordinates the execution of the suggested actions. This may mean that if you don’t make your flight, the virtual assistant can seamlessly rebook airline tickets, change accommodation dates, make new restaurant reservations – and even send the letter of complaint and compensation claim to the airline.

Today, large consumer products brands simply aren’t equipped to provide each individual customer with accurate, consistent, yet always personalized, contextual content. Generative AI can make what was once unfeasible attainable. The visionary concept behind the 30-year-old groundbreaking book, The one-to-one future: building relationships one customer at a time,can finally be embraced and scaled in all its glory.

Making it work in CX

While generative technologies can help us create useful and contextual content, they still require a holistic framework to be used by enterprises to improve customer experience. At a high level, any enterprise will need four key elements to adopt generative AI in CX (in addition to standard elements like data, algorithms, integrations):

  1. Business use cases: While there are many use cases imagined for generative AI in CX, it’s important to understand the feasibility and value each will bring. An enterprise will require a refined strategy to select the right use cases that will deliver tangible outcomes (applicable to their business).
  2. CX orchestration: Generative content can be used to create a more engaging and personalized customer experience. However, it is important to carefully orchestrate this content in order to ensure that it is consistent with the brand’s values (tone, voice), target audience and overall CX goals. By carefully considering these factors, businesses can use generative content to create a more cohesive and memorable customer experience.
  3. Guardrails: A powerful layer of CX guardrails (brand guidelines, core values, vision of brand etc.) need to be applied to prompts and inputs, and most importantly, the security of models (scope of data and usage). By putting guardrails in place, businesses can ensure that generative AI is used in a responsible and ethical way. This can help to protect the brand, the customers, and the data.
  4. Adoption methodology: Generative solutions cannot be seen in isolation as they become part of existing work been done by team in CX space (marketing, sales, service or commerce). Enterprises need to have an adoption methodology that ensures all elements of technology, people and process are fine tuned to embrace changes brought in by adoption of generative technologies.

A strategic approach for controlled impact

Even though full maturity of generative AI isn’t expected for another 2-5 years, 70% of global organizations have already started exploring the technology’s probable future.[1] This has regulators scrambling to create guidance and restrictions around its use. As a first of its kind – before the fantasy of AI became reality – the European Parliament has put together a draft law, the AI Act, set to be released later this year. More regulations will undoubtedly soon follow.

Of the organizations that have kick-started their AI experimental journey, most haven’t considered the implications these regulations will have on their final creations. They could be forced back to the drawing board, increasing costs and delaying progress. This is where a skillful advisor can be most beneficial. They’ll know what to expect and can provide foresight to avoid the common pitfalls, especially if they’ve successfully overcome the challenges of previous technological evolutions. Ideas will be fast-tracked, efforts will be minimized, and the transformative value of generative AI will permeate across any organization ready to spark unprecedented change to customer experience.

[1] https://www.gartner.com/en/newsroom/press-releases/2023-05-03-gartner-poll-finds-45-percent-of-executives-say-chatgpt-has-prompted-an-increase-in-ai-investment

[2] https://www.capgemini.com/insights/research-library/generative-ai-in-organizations/

For more details, contact :

Alex Smith-Bingham

Executive Vice President, Group Offer Lead for Customer Experience; Digital Customer Experience Lead for UK
“Customer Experience covers all the support and help our clients need between them and their customers. This will range from changing their purpose, their propositions, new capabilities in sales/service/marketing and commerce, immersive experiences, new operating models, and new ways of working and ecosystems. We harness our global capability in strategic innovation (frog), business consulting, DCX solutions, Insight & Data and run operations in technology and business services.”

Darshan Shankavaram

Executive Vice President, Digital Customer Experience Global Practice Leader
I have close to 30 years of domain experience, with more than ten years within Digital and Mobile. I have led product concept-to-sell, business development, pre-sales, solutioning and technical implementation of CX transformation programs.

Steve Hewett

Head of Customer Transformation, frog, Capgemini Invent UK
Steve specializes in the digital transformation of ‘retailing’ – he is leading our offer development for how generative AI will impact the e2e CX of our clients and their customers – from how it will help to set new customer experience strategies & develop new propositions to how it will transform digital marketing, omni- commerce, store experience & operations, customer service, and CRM & Loyalty.

Naresh Khanduri

Expert in Innovation, Strategy

Mark Oost

Global Offer Leader, AI Analytics & Data Science
Prior to joining Capgemini, Mark was the CTO of AI and Analytics at Sogeti Global, where he developed the AI portfolio and strategy. Before that, he worked as a Practice Lead for Data Science and AI at Sogeti Netherlands, where he started the Data Science team, and as a Lead Data Scientist at Teradata and Experian. Throughout his career, Mark has had the opportunity to work with clients from various markets around the world and has used AI, deep learning, and machine learning technologies to solve complex problems.