Google x Splio: reinventing CRM in the AI era 

A Google partner for several years, Splio co-organized an event at the search engine's Paris offices on October 13th for an afternoon of discussions around a topic that is profoundly transforming digital commerce : conversational AI. Behind this term lie technologies such as chatbots, virtual assistants and natural language-based agents, capable of interacting in real time with customers in real time and supporting brands in a more personalized and effective relationships.

For this afternoon, three experts were invited to speak and highlight the impact of conversational AI on their technological vision (Google and Splio) and on the future use cases for CRM teams (David Bessis, founder of Tinyclues).

A look back at the inspiring discussions we had with our customers ⬇️

Google: an effective conversation is a targeted conversation

Clément Cardi, Product Manager at Google - Vertex AI Search for Commerce, opened the session by unveiling how Google integrates artificial intelligence into its engine to optimize product search.

Behind an apparently simple query, more than a dozen algorithms work in parallel. Their role: to understand immediate behavior (for example, the last three clicks, which are highly decisive), the long-term history, the brand constraints (positioning, branding...) and the customer expectations.

👉 Efficiency relies on guided freedom: letting the AI explore all possible combinations, while steering it with clear business signals.

A key insight concerns conversational search. In e-commerce, customers come to find a product. Tests conducted by Google have shown that overly open-ended conversations degrade the experience. Conversely, a targeted approach (focused on product search) streamlines the journey and increases relevance.

Finally, Clément reminded us that LLMs are not the one-size-fits-all solution. While these models knows how to "answer everything", they must be complemented by other specialized algorithms to ensure reliability and contextualize results. Google is therefore banking on a hybrid architecture: LLMs orchestrate, but precision comes from business models.

Tinyclues: AI-inside vs. Ai-first, changing engine and the experience

David Bessis, founder of Tinyclues and now Chairman of the Board at Splio, then took the floor to distinguish between two approaches:
Ai-inside: products, which add artificial intelligence to an existing structure;
AI-first: products, designed from the outset around artificial intelligence.

To illustrate this transition, he used a maritime analogy: in the 19th century, transport shifted from sailing boats to steamships. In between, hybrids combining both technologies emerged. A temporary solution, quickly superseded by the " engine-first" approach, which changed everything.
Similarly, truly AI-first products don't simply add AI features. They redefine the user experience.

👉 The UX often becomes radically simplified (as in ChatGPT) because the intelligence is carried by the engine, not by the complexity of the interface.

David also emphasized that these disruptions redefine the role of users. Just as the captain and sailors had to rethink their place on a steamship, professionals must adapt to new AI tools. This requires a cultural changes as much as a  technological one.
Finally, he warned against a natural bias of LLMs: their tendency to "agree" with the user to facilitate conversation. This fluidity can create an illusion of accuracy. Vigilance is therefore needed to avoid confusing a pleasant experience with a reliable answer. 

Splio: from AI decisioning to conversational CRM

Finally, to close the event, Antoine Parizot, CEO of Splio, shared the strategic vision that guides the evolution of our CRM platform. Namely, the need to put the customer back at the center of the debate, with artificial intelligence as the driving force.

Indeed, CRM managers today face two imperatives:

• Meeting customer expectations: personalized experience, automated cycles, relevant interactions.
• Serving business objectives: stock clearance, mass campaigns, commercial activation plans.

Splio provides a structured response to this challenge, with three levels:
1. Multichannel orchestration of commercial communication plans with the addition of new channels to its marketing automation platform, such as RCS and WhatsApp.
2. Unified integration of relational and commercial campaigns.
3. Predictive AI (Tinyclues) that anticipates which products to recommend, which promotions to offer and which channels to prioritize for each customer, while limiting marketing pressure.

This intelligence is not limited to CRM: it must fuel the brand's entire application ecosystem. Conversational agents, which are rapidly growing, are becoming a natural extension of these predictive capabilities.

Th discussions with customers at the event highlighted three key expectations:

• Ever more refined personalization,
• Tools that make teams more productive,
• And an ability to prove the value of marketing actions in a demanding economic context.

This Google x Splio event highlighted a shared conviction: the future of AI in commerce relies not only on the power of models, but on how they are integrated into targeted, hybrid and strategically designed experiences.

Google showed how AI can make product search smarter without sacrificing efficiency. Tinyclues reminded us that true breakthroughs come from AI-first products, which are profoundly changing the game. Finally, Splio demonstrated how this vision translates concretely into AI-first CRM, at the service of brands and their customers. They agreed that the integration of artificial intelligence depends on the alignment of vision, technology and usage.

To find out more about integrating artificial intelligence into Splio's CRM marketing platform, don't hesitate to contact our teams!