Agentic Enterprise 2026: Strategy Roadmap

Reliability at scale is achieved through Multi-Agent Systems. Rather than relying on a single model, we now architect « societies of agents » where specialized units (e.g., Research, Analytics, Execution) collaborate through Agentic Design Patterns. [Andrew Ng / AI Fund, 2024]. This architecture ensures that the « Decision Intelligence » layer of the enterprise remains robust and hallucination-free. [Gartner, 2025]

For Yazan, the ultimate goal is Operational Sovereignty. Companies must move beyond « AI as a feature » to « AI as an Infrastructure. » By integrating these agents directly into the CRM and decision-making pipelines, firms can achieve a level of responsiveness that was previously impossible.

Over the past 12 months, we had significant shifts in how consumers are searching. Consumers are shifting towards clickless and agentic search, searching through LLMs. Searches are becoming faster and much more relevant as consumers add more context to their ask, as Claire quoted, consumers on average are now using 8 words even when searching in Google. Consumers are shifting quite rapidly: despite the fact research from Semrush showed AI Search was only roughly 5-6% of volumes last July, expectations is that it will be around 25% at the end of 2025 and reach a 50/50 split between traditional and agentic search within 4 years.

Besides the search, consumers and especially for some industries, are also starting to delegate to LLMs their agents for buying decisions, as seen with the deployment of InstantGPT in the US on the strategic partnership between ChatGPT and Instacart. Fabien used the word ‘we are starting the era of delegation’ to describe this shift. Agentic commerce is dramatically changing the marketing funnel: compressing Awareness to Consideration, reducing the number of brands in consideration, and making brand stories invisible to consumers. Also, while humans are loyal, bots are not. Companies would need to adapt their marketing strategy to win with both bots as well as humans.

Winning with LLMs is quite different than winning with SEO. Today, a lot of new tech companies are emerging in the GEO space (Generative Engine Optimization). Some of these technologies help brands scan their AI citation, identify the key sources LLMs pull information from, and create new content with GenAI to fit LLM queries, helping increase the likelihood of being cited by an LLM for key strategic questions.

Besides the need to win with algorithms, brands still would want to invest in their owned marketing emails, apps, and loyalty programs. These are the insurance policy against disintermediation and the loss of first-party data, which are becoming even more important in this new ecosystem.

Similarly, stores and physical engagement opportunities will still be important. Besides the fact that consumers won’t only shop online, stores are critical to drive awareness and loyalty. Companies like Decathlon have already been working for a few years on bringing agentic experience to the store. Now it’s mostly for internal work, such as replenishment of inventory, but eventually, it will also be for consumer-facing experiences.