CRM Managers: How to prepare for the agentic era?

Until now, CRM managers structured customer journeys and planned campaigns in advance based on specific segmentations. Acquisition, conversion, and retention followed a relatively linear logic, centered on proprietary touchpoints.

This mechanism is disappearing with the rise of AI agents, conversational web, and smart interfaces, which are revolutionizing the way consumers discover, compare, and choose brands. Increasingly, purchasing decisions are being made outside of brand platforms.

The role of the CRM manager is therefore undergoing profound change in the age of agency. It is no longer just a matter of managing campaigns, but of becoming a conductor capable of guiding decisions on an ongoing basis and collaborating with AI, while retaining strategic control of customer relations.

Accept unpredictability

As the role of the CRM manager is reinvented, linear customer journeys are a thing of the past, and the proliferation of channels and signals is transforming uncertainty into a strategic challenge to be understood and exploited.

The end of linear customer journeys

Historically, CRM teams built tunnels structured around four key stages:

• Discovery
• Consideration
• Conversion
• Loyalty

These stages of the customer journey still exist, but they no longer necessarily take place entirely on the brand's website. Above all, the point of entry becomes a conversation. Consumers ask a conversational AI for advice, compare several brands without visiting their websites thanks to its recommendations, and proceed directly to checkout with a decision that is almost already made.

Interactions are also multimodal, contextual, and spontaneously triggered. A customer can switch from a voice search to a text chat, then to a mobile purchase without ever following a traditional path. For CRM managers, this change in usage means one thing: closed scenarios are no longer sufficient.

More channels, more signals, more uncertainty

Instant messaging, shopping assistants, generative recommendations... The points of contact between a brand and a consumer are multiplying. Platforms such as WhatsApp are already transactional spaces in their own right, where customer relationships are built even before the brand collects any identifiable data. A prospect can interact with AI assistants for several weeks before entering the brand's database.

The CRM then receives more signals... but controls them less. Because if customer loyalty remains the ultimate goal, it now begins much earlier and often outside the organization's ecosystem.

Faced with this rapid change, curiosity is becoming a key skill for CRM managers. It is necessary to be able to quickly test new interfaces, understand the logic of conversational agents, and closely observe emerging uses. When everything is in flux, the only constant is the ability to adapt. In 2026, this will necessarily involve a new relationship with AI.

From automation to co-decision: building a contract of trust with AI

Marketing marketing automation has always been based on programming rules. Today, AI is stepping in to make decisions. Predictive audiences, product recommendations, conversational analysis, or choosing the right moment to send a message... Systems are capable of identifying opportunities invisible to the human eye in order to anticipate the actions to be taken. Antoine Parizot, co-CEO of Splio, sumsitup: "The role of the CRM manager is changing profoundly: it is no longer just a matter of executing marketing scenarios, but of understanding how to collaborate with systems capable of making decisions faster than we can."

At certain brands such as Mademoiselle Bio and Air France, CRM teams are already observing that AI selects unexpected and highly effective segments. But the same questions often arise: are the audiences really more relevant? Do recommendations delegated to AI really increase conversion?

Some brands have begun to adopt a very pragmatic approach: systematically testing AI decisions against human approaches. This is particularly true of Air France, which generated more than 900 predictive audiences in 2025 and saw conversion rates double on its campaigns.

In this context, AI makes suggestions and the CRM manager makes the final decision, without ever being replaced. This is because the objective goes beyond immediate performance and is based on understanding customer behavior over time. AI can sometimes act quickly, without emotional or historical context. This is precisely where the CRM manager retains their value and becomes the long-term memory of the customer relationship and the guarantor of brand consistency.

What to delegate to AI... and what CRM managers should keep for themselves

On certain topics, the debate is already over. AI excels at:

• Complex segmentation
• Calculating the best time to send
• The product recommendation
• Detecting weak intentions

Cédric Gourgeon, E-commerce Director at Mademoiselle Bio, has experienced this firsthand. For the brand, the main challenge was targeting consumers who had made a single purchase but never placed another order. Predictive AI made it possible to consider complex scenarios for reaching them, by simultaneously analyzing hundreds of variables. After three months, the results of using AI are clear:

• 90% of revenue generated by email was achieved by targeting only 28% of the contact database.
• SMS campaigns targeting predictive audiences performed up to 16 times better than those based on manual segmentation.
• Product recommendations have increased click-through rates by 33%.

Despite these achievements, delegating everything to AI would be a mistake. CRM managers must remain in control of brand strategy, sensitive decisions, and data governance. AI can optimize open rates, but it cannot decide on editorial repositioning or define the relational promise.

Why CRM is becoming the core of acquisition in the agentic era

In the age of agents, CRM is the cornerstone of visibility and growth. It connects customer knowledge, algorithmic recommendations, and user relationships, where intelligent agents now orchestrate discovery and purchasing decisions.

Being recommended by an agent is becoming the new SEO

For more than twenty years, digital acquisition relied heavily on search engines. SEO, media buying campaigns, and content strategies all had the same goal: to capture the user's attention at the precise moment they expressed an intention.

This model is evolving: whereas yesterday internet users would type a query into Google to compare several options, today they ask an agent directly which brand to choose, which product to favor, or which service best suits their situation.

However, in order to formulate their responses, agents simultaneously analyze a multitude of signals: a brand's editorial consistency, the quality and regularity of its content, customer reviews, overall online perception, and consistency of positioning across different channels.

In other words, visibility no longer relies solely on technical optimizations or well-placed keywords. It depends on how a brand is understood by artificial intelligence models. Thomas Skowronski, EVP at Jellyfish, notes: "We are seeing a return to branding."

When the customer returns to the brand

Even though brands no longer have complete control over the customer journey with the intermediation of assistants, they still retain one decisive advantage: the moment of conversion. Discovery and comparison take place via a conversational agent, but the final purchase is still mostly made in proprietary environments (website, app, or customer area).

This step is a strategic point where CRM comes into its own. The earlier the login occurs in the journey, the richer the first-party data becomes. Each interaction becomes an opportunity to rebuild a direct link with the customer, understand their preferences, and nurture lasting knowledge that external platforms do not possess. In an environment where entry points are multiplying and are partially beyond the control of brands, this ability to recreate a proprietary link becomes a key lever for acquisition and retention.

Acquisition, retention, and CRM are now one and the same

Little by little, the agentic era is blurring the line between acquisition and retention.

A satisfied customer shares their experience, leaves a review, or spontaneously recommends a brand. These signals feed into conversational models. Agents, in turn, give greater prominence to companies that are perceived as reliable or consistent. Loyalty then produces an indirect, almost organic acquisition effect.

CRM is no longer just a tool for customer retention or sales promotion. It has become a driver of overall growth and a central source of reputational and relational signals, capable of influencing discovery, recommendation, and conversion in an ecosystem where relational trust now carries as much weight as marketing visibility.

5 practical steps to transition to the agentic era

With the emergence of intelligent agents capable of searching, comparing, and recommending on behalf of users, brands must rethink their operational fundamentals: ensuring data quality, communicating effectively with customers, and fundamentally transforming their practices.

1. Invest in CRM data quality

No artificial intelligence can produce relevant recommendations based on inconsistent or fragmented information. Poorly structured data can also distort the interpretation of the brand by AI agents.

A clean database not only allows for more personalized interactions, but alsofeeds these models. In concrete terms, companies must consolidate their databases by deduplicating customer profiles, standardizing formats, enriching behavioral attributes, and updating transactional information.

2. Transform your platforms into information hubs

Long considered transactional showcases, websites and mobile applications are now becoming hubs for data collection and learning. The more a brand enriches its product listings, FAQs, blog articles, metadata, structured data, and highlights its product reviews, the more it will be recommended.

Cécile Laloy, Sales Director Digital Natives at Google Cloud, affirms: " You need to invest now in your website and/or app: these are the sources that feed chatbots. It is necessary to provide them with comprehensive, rich, high-quality data that is well-structured and open to APIs. "

3. Test the prediction rather than segment manually

Traditional marketing segmentation is often based on human assumptions (age, location, purchase history, etc.). Today, these approaches are reaching their limits in the face of the complexity of customer journeys.

Predictive models can identify weak signals invisible to the human eye: browsing patterns, interaction timing, or atypical behavioral combinations. In many cases, audiences generated by a predictive model reveal rapid gains in conversion or engagement.

4. Implement a conversational strategy

As interfaces evolve, conversation is becoming a major entry point in customer relations. Instant messaging, chat commerce, and application-integrated assistants are no longer just support tools: they now structure a growing part of the purchasing process.

An effective conversational strategy requires understanding the key moments when dialogue adds value: the inspiration phase, hesitation before purchase, post-order support, or personalized recommendations. When well orchestrated, these exchanges become a continuous source of qualitative data that can enrich customer knowledge far beyond simple transactions.

5. Train CRM teams to collaborate with AI

Finally, agentic transformation is above all cultural. Tools are evolving rapidly, but their effectiveness depends on teams' ability to understand and manage them. Knowing how to interpret an algorithmic recommendation or challenge a model is becoming as important a skill as building a relational scenario.

CRM managers must become capableof balancing automation and human intervention, immediate performance and long-term brand consistency. As Antoine Parizot, Co-CEO of Splio, sums it up, " We will no longer be able to predict everything. The role of the CRM manager is to decide what to delegate to AI and what to keep: challenging the proposed audiences, checking the analyses, and ensuring that the recommendations really serve performance and the brand. "

The CRM manager or the conductor of AI

The agentic era is redefining the importance of CRM. As AI agents become involved in discovery, recommendation, and purchasing, the role of the CRM manager increasingly consists of coordinating an ecosystem where humans and intelligent systems collaborate continuously.

In this new context, accepting a degree of unpredictability becomes essential. The challenge is no longer to control everything, but to build a relationship of trust with AI: delegating large-scale analysis and execution to it, while retaining control over data, customer relationships, and strategic decisions.

More than just an operational expert, the CRM manager becomes a conductor. They ensure consistency, memory, and vision for the brand, while AI provides speed, power, and adaptability.