1. Generative AI: accelerating content creation and adaptation

Generative AI opens up new possibilities for marketing teams. It makes it possible to produce content more quickly, adapt a message for different channels and streamline campaign preparation.

In a CRM context, it can notably help to:

 

• Generate several variations of an email subject line or message.
• Adapt content for different segments, levels of familiarity or stages in the customer journey.
• Adapt an email campaign into push notifications, SMS messages or other customer communication formats.
• Prepare a first draft of a campaign based on a marketing brief.

• Rephrase content to adjust its tone, length or degree of personalisation.

   

These uses help reduce the time spent on certain production tasks and make it easier to explore different creative approaches.

However, producing more content does not guarantee a better customer experience. Message relevance, consistency with the brand and alignment with the recipient’s expectations remain essential. Human involvement therefore continues to play a central role in validation, creative direction and the final quality of campaigns.

The aim is not simply to produce faster, but to produce more relevantly.

 

2. Predictive AI: better anticipating customer behaviour

AI can also help marketing teams make better use of their data to anticipate behaviour and identify opportunities for action.

Using available data and predictive models, it becomes possible to estimate certain probabilities: risk of disengagement, likelihood to buy, potential interest in an offer or probability of reactivation.

In practical terms, these approaches can make it possible to:

 

• Identify customers whose engagement is declining and who may need a retention initiative.
• Identify contacts most likely to respond to an offer.
• Detect reactivation opportunities among inactive customers.
• Refine campaign audiences according to the objectives being pursued.

• Prioritise marketing activities according to customers’ potential value or expected behaviour.

   

The aim is to move beyond an approach based solely on static criteria—age, purchase history or the date of the last interaction—and take greater account of observed behaviour and changes over time.

However, this ability to anticipate depends on the quality, availability and relevance of the data. A predictive model cannot compensate for insufficient customer knowledge: it relies on it.

AI performance depends as much on data quality as on the relevance of the decisions that the data enables us to make.

 

3. Decision-making AI: orchestrating more relevant journeys

Beyond content creation and prediction, AI can help optimise marketing decisions: which message should be sent, to which customer, through which channel and at what time?

The aim is to make customer journeys more dynamic by taking account of behaviour, recent interactions and campaign results.

For example, an activation strategy may seek to:

 

• Adjust campaign send times according to engagement habits.
• Compare several message variations to identify which best meet a given objective.
• Adapt a customer’s journey based on their most recent interactions.
• Adjust marketing pressure to limit unnecessary communications.

• Direct a customer towards a more relevant action or piece of content based on their profile and behaviour.

   

These mechanisms can rely on business rules, predictive models or more advanced optimisation capabilities. Their value lies not in technological sophistication alone, but in their ability to tangibly improve the experience and results.

This requires defining the right objectives, appropriate performance indicators and the rules governing decisions. Effective personalisation must remain consistent with the brand promise and respect customers’ preferences.

The right scenario is not the one that multiplies interactions, but the one that triggers the most relevant action.

 

4. Agentic AI: towards better-supported CRM teams

A new stage is emerging with agentic AI. Beyond generating responses or content, some systems can chain together several tasks to achieve an objective, analyse a situation and suggest actions.

In CRM activities, these capabilities can notably help to:

 

• Analyse campaign results and highlight the main lessons learned.
• Identify friction points in a customer journey.
• Prepare a segmentation or an initial proposal for a customer journey scenario.
• Suggest optimisation opportunities based on observed performance.

• Support teams in preparing, checking or improving their campaigns.

   

The level of autonomy of these systems varies depending on the solutions, use cases and framework in place. It is therefore essential to distinguish between assistance that makes recommendations and a system authorised to execute actions directly.

For marketing teams, the challenge is to determine which tasks can be delegated to these systems, what checks are needed and which responsibilities must remain with people.

Agentic AI does not eliminate the need for CRM expertise. On the contrary, it may make that expertise even more important: teams need to know how to define objectives, govern actions, check results and assess recommendations.

The more capable the tools become of taking action, the more decisive the strategic and operational framework becomes.

 

5. From technology to impact: the key role of CRM expertise

AI offers many possibilities. But its value depends on how it is integrated into marketing strategy and teams’ practices.

Deploying an AI feature is not enough to transform a customer relationship. It is necessary to identify use cases that address a genuine business challenge, have the required data, design relevant journeys and measure the impact on performance.

This is where CRM expertise comes into its own.

At Degetel, our approach is to bring together relationship strategy, customer insight, creativity and technological expertise to turn AI’s potential into concrete action.

This notably involves:

• Identifying the right use cases, starting from business challenges, customer expectations and marketing objectives.
• Structuring customer insight by working on data quality, segmentation and the signals useful for activation.
• Designing relevant customer journeys by defining the right messages, channels and moments.
• Integrating AI capabilities into existing systems, taking account of tools, processes and team skills.
• Governing and evaluating use cases while preserving brand consistency, the quality of interactions and control over decisions.

• Measuring results to check that initiatives create value and improve scenarios over time.

   

The objective is not to add AI everywhere. It is to identify where it can genuinely improve team efficiency, the relevance of interactions and the performance of marketing activities.

 

 

From idea to impact. Continuously.

 

By Caroline Flausino, Director of Degetel’s Agency division, specializing in customer strategy and activation.