Even before the rise of generative models, artificial intelligence (AI) played a key role in e-commerce. Initially, AI enabled us to better understand online customer behavior by processing massive volumes of browsing data that were impossible to analyze manually. Its field of action quickly broadened from descriptive analysis, measuring, observing, and understanding to concrete actions, such as personalizing and hyper-personalizing shopping experiences.
However, the most significant turning point in recent years has been AI's ability to understand natural language.
Understanding meaning beyond words: a turning point for e-commerce thanks to AI
Even before ChatGPT, AI had already begun exploring the field of semantics. Technologies such as machine translation demonstrated the ability to interpret meaning beyond individual words.
Starting in 2019, with the arrival of LLM (Large Language Models) like Google's BERT, machines began to contextualize each word in its environment. Language became modelable and mathematically comprehensible on a large scale. It is no longer viewed as a simple sequence of characters but rather as a network of linked concepts carrying nuances, contexts, and meanings. This breakthrough has opened the way to powerful new uses.
Sensefuel recognized this potential early on. As of 2019, we have integrated this approach into our e-commerce search engine, based on Deep Learning. This choice was driven by our conviction that to respond precisely to consumer expectations, we must understand the intention behind the words. This is the very basis of our name, Sensefuel - the fuel of common sense.
The evolution of AI models has been meteoric. Driven by scientific progress and massive investments, particularly at OpenAI and throughout the tech ecosystem, this race has produced models of an unprecedented scale. These models are fed by massive datasets and require massive computing capacities, and the associated costs are equally staggering.
And the results are in: a glass ceiling has been broken. For the first time, artificial intelligence gives the impression of understanding. Not because it thinks, but because it calculates, with phenomenal speed and depth, the probabilities of linking words, ideas and intentions. She completes sentences with disconcerting naturalness. This explains the feeling of being understood, or even anticipated, that many people describe today. The anthropomorphic effect is very real.
Don't miss our expert opinion: The impact of generative IA in e-commerce
What does this mean for e-commerce? Absolutely everything. The days of rigid keywords are gone. Internet users now express themselves freely by formulating long, nuanced queries and expect equally natural responses. Search becomes conversational. This is more than just a technical advancement; it reshuffles the deck in terms of customer experience.
GenAI, or the return of the salesperson to e-commerce
This transformation is part of a familiar trajectory. In the physical retail sector, we have moved from the local store, based on proximity, to more standardized formats, focused on efficiency and self-service, without ever totally abandoning the presence of a salesperson to advise or reassure if necessary.
E-commerce has followed a similar path. Initially conceived as a free-access platform focused on offers and prices, e-commerce has gradually adopted merchandising strategies, such as recommendations, filters, and highlights.
Today, a new phase has begun. Thanks to GenAI and conversational interfaces, the online experience can finally offer personalized, on-demand advice. From understanding a need to helping you choose to proposing a more suitable alternative, this is the return of the salesperson in digital form. Our Sensefuel Conversation solution is all about reintroducing support at the heart of the shopping experience.
And adoption is accelerating. ChatGPT, Perplexity, Google's Search Generative Experience... These interfaces define a new consumer reflex: formulating a request in natural language and obtaining a relevant response. In just a few months, what was once an innovation has become the new standard. Users of e-commerce sites also expect this standard.
Generative artificial intelligence represents a significant opportunity for e-tailers. However, it also poses a risk: losing direct connection with customers and becoming disintermediate.
Without AI mastery, brands risk becoming invisible
In an ecosystem increasingly driven by artificial intelligence, the danger is no longer of being less visible, but of becoming invisible. If a third party (marketplace, search engine, conversational assistant) captures the customer relationship, e-retailers risk becoming nothing more than a logistical link. The end customer buys their products without even knowing who they're buying from. And conversely, the brand no longer knows who its buyers are, or how to activate them. It loses all autonomy in the commercial relationship, to the benefit of an intermediary who concentrates most of the value.
The parallel with the restaurant industry is telling. On delivery platforms, customers order a dish, not a restaurant. If the algorithm suggests another supplier, the customer doesn't even notice. They consume the product without knowing where it came from.
The lesson for brands is clear: integrating AI means regaining control over the customer experience. It preserves their identity, value, and ability to differentiate themselves.
Regaining control of the brand experience
Faced with this new reality, brands must once again become destinations in their own right. This means offering an experience on their e-commerce site that meets the standards now set by AI-driven interfaces: fluidity and engagement. The goal is to make people want to come and, more importantly, stay.
While AI establishes new standards, it also signals a shift in the paradigm. We have transitioned from an era of assistance to an era of freedom. Consumers want the freedom to choose how they want to do things: independently or with the support of quality advice.
Experience is becoming a pillar of performance, as important as product and price. And if we were to rethink the 4Ps of marketing - product, price, promotion, place - we'd have to add an E.
E for Experience.