Reinventing the Customer Experience with Artificial Intelligence
In a world where digital experience and artificial intelligence (AI) are redefining customer interactions, it is crucial to understand how brands navigate this constantly evolving landscape. To this end, Gary Roth, VP of Business Value Services at Contentsquare, and Laurent Nicolas-Guennoc, Chief Marketing Officer at Converteo and host of the podcast “Changement d’époque en cours” dedicated to AI, share their insights.
Have you been an “early adopter” of AI, or did it come later in your career?
Gary Roth: I wish I could say yes, but the real pioneers were in the 1950s with the perceptron, and then in the 1980s with essential algorithms. It was the 2000s, with the migration to the cloud and the explosion of data, that led to significant advancements, followed by the 2010s with the rise in computing power.
Laurent Nicolas-Guennoc: Like millions of connected citizens, I have been using Waze while driving or the power of DeepL to translate texts for several years now. AI has already begun to change our daily lives significantly. As for generative AI chatbots for text or images, to be honest, I jumped on the bandwagon like many others in late 2022, but since then, yes, these tools have become part of the solutions I use daily.
Complete the sentence: For me, AI is…
Gary: … probably one of the greatest technological revolutions of our lifetime, not to mention quantum computing, which could also be on the horizon. It is both the greatest opportunity we have encountered to date, but also potentially a threat to humanity.
Laurent: A set of general-purpose technologies whose immense transformative potential— and, in doing so, destabilization, or as Mustafa Suleyman writes, amplification of vulnerabilities— for society and the economy we are only just beginning to perceive, having identified only a tiny fraction of future purposes and applications. With the invention of electricity, we quickly invented the light bulb, but from there to imagining the TGV…
How do you see the adoption of AI in large companies?
Gary: The adoption of AI in large companies has been extremely rapid, unlike previous technological revolutions. Companies like L’Oréal and Coca-Cola have managed to create powerful AI-driven products in record time. However, one of the major challenges remains the cost of AI. In the long term, companies will need to assess not only the performance of AI technologies but also their cost and environmental impact.
Laurent: There are several parallel movements. The year 2024 is already marked by a rationalization of experiments, focusing on a few projects where a greater added value from AI algorithms is anticipated, on one hand, and a capacity to scale and integrate into the IT and cloud architecture of companies, on the other. It’s the end of playtime; we’re stopping the tinkering. Yet at the same time, new AI features, particularly generative ones, are being integrated into office suites (Microsoft, Google, of course) used by hundreds of millions of employees. A retreat on one side, a tidal wave on the other.
What are the main barriers to AI adoption in companies today?
Gary: The primary challenge is often resistance to change and a lack of specialized resources. A gradual approach, starting with small-scale actions, can help overcome these obstacles. It is crucial to begin with a real use case: an AI initiative must address a concrete business need, whether it’s adding value for customers, reducing costs, improving synergy with partners, or increasing revenue growth.
Laurent: The main barrier, in my opinion, is the actual utility of AI within the company and how to measure it. Is it really helpful for my tasks? Is it more reliable, faster? How can I ensure that and quantify it? Making the general management or employees believe that AI is the technological key that will solve all projects only leads to disappointment and can misdirect investments. On the contrary, it’s essential to maintain the ability to diagnose technological needs accurately and align the right tools with identified problems. AI is fantastic for many use cases, particularly within general management, marketing, sales, or product departments that we support; however, it can be disproportionate or simply unsuitable for others.
How do you help your clients overcome these challenges?
Gary: We assist our clients by offering plug-and-play solutions that minimize technological barriers. For example, our tag is compatible with 99% of the market without compatibility issues, simplifying integration. We also have a dedicated performance optimization team to ensure that our technology does not negatively impact website speed. Finally, we integrate our solutions into our clients’ existing tech stack, including Adobe Analytics, Splunk, Dynatrace, Salesforce, etc., to maximize added value.
Laurent: By being agnostic to market technologies, the only way to resist the hype is to never lose sight of their needs. As a consulting firm for large companies, we must take the time to understand the existing technological environment and assess the maturity level of both the systems and, more importantly, the teams and organizations. AI, like any major new technology, impacts human organizations: for example, there is no uniform answer to the question of whether to appoint a “Chief AI Officer.” Only a thorough immersion in the company culture allows us to determine if such an appointment is relevant and, if so, at what level to position it, with what roadmap and resources.
Which AI applications offer the best return on investment, in your opinion?
Gary: It really depends on your maturity and resources, including your team and budget. For some, immersive technologies allow customers to see products from all angles. Others benefit from AI-powered search to quickly connect the right products with the right customers. Try-on technologies, like Google’s impressive demonstration, could be revolutionary if they work well.
Laurent: We see strong productivity gains in automating content production optimized for search engines (SEO) with AI, as we did with L’Oréal, handling document bases that are impossible to process manually. We also achieved very good results with the Groupe Seb by adding previously nearly unusable information to product sheets, thanks to AI, which was stored in unstructured sources like PDFs. The dynamic creation of contextualized visuals in advertising environments, as well as conversational assistants for sales support on eCommerce sites, are also cases where we are very optimistic about the ROI of artificial intelligence systems, based on the initial confidential projects we are advancing with major players.
What are the use cases of AI that you believe are most accessible and effective for medium-sized enterprises?
Gary: Session-based interfaces and large language models (LLMs) only cost around twenty euros per month. These powerful models are versatile and can be used by anyone in an organization.
Laurent: The answer to this question is likely to change quickly given the advancements in systems and the needs of the business, of course, but I would say that starting to use a large language model (LLM) to produce more relevant multimedia content for its blog, website, social media, and newsletters is a first step that many teams can take even without technical expertise. AI can also save a lot of time in searching for internal documents (knowledge management), summarizing meetings, translating, and transcribing text to audio, video, and vice versa from audio to text.
What is your vision for the online store of the future? How do you think shopping experiences will evolve with AI and immersive technologies?
Gary: I believe there won’t be a single type of online store in the future, but rather a multitude of experiences tailored to different types of customers. For example, my parents will probably continue to prefer a simple eCommerce site, while my niece, who spends her days on Roblox, will be more attracted to immersive 3D stores. My personal dream is to shop in 3D in my living room, in a virtual department store by the sea, with next-day delivery by Amazon. AI and immersive technologies will transform these experiences, making online shopping more interactive and personalized.
Laurent: The success of the store of the future will, I believe, always depend on the ability to put the consumer in a position of trust. Trust in the product, of course, trust in the recommendation, in the advice, and in the information surrounding that product. In 2050, why would I buy this t-shirt? Because an app tells me that my stock of t-shirts is running low and that I need to renew a piece. Then, I can try on the model virtually and see right away that it goes well with the color of my eyes. Once reassured about the style, I can click to access the entire supply chain, including scope 3, showing me where the raw materials come from, their quality, and where the transformation takes place. I truly need this t-shirt, I am reassured that it suits me, and I consider its environmental impact to be sustainable: I trust it, I buy.
How do you see the evolution of the customer experience in the coming years, particularly with advancements in AI and machine learning?
Gary: I believe the future of customer experience will be characterized by more intuitive and engaging journeys thanks to AI. This technology will enable a better understanding of customer needs in real-time and predict their behaviors. This will lead to more personalized and frictionless experiences. Systems will automatically detect and resolve user frustration points, significantly improving customer satisfaction. Additionally, AI will facilitate the integration of user feedback, making products and services more responsive to consumer expectations.
Laurent: I completely agree with Gary, and I would add that one of the most transformative factors in customer experience, enabled by AI systems, lies in voice and audio signal processing. Recent advancements have been as phenomenal as those in image processing, although less publicized. This is a fundamental change: I believe we will interact much less with screens and much more with our voices in the future. Searching for and comparing products, asking for details in multiple languages, and possibly even negotiating prices will involve natural language interactions interpreted by AI algorithms with error rates that will increasingly approach zero. Let’s not forget that with over 270 million speakers, French is the fifth most spoken language in the world. While we are far from English or Chinese, large AI models have access to vast amounts of content to make French one of the best-understood languages by artificial intelligence.