Customer & Marketing
The year 2026 marks a breakthrough for marketing departments. Artificial intelligence (AI) technologies are no longer gadgets: they are being massively adopted by consumers and businesses alike. According to Microsoft’s AI Diffusion Report 2025, global AI adoption grew by 1.2 points in the second half of 2025, with around one in six people worldwide now using generative AI tools. However, the spread is uneven: 24.7% of the working population in the “Global North” uses AI tools, compared with just 14.1% in the “Global South”. This rapid, differentiated penetration is forcing marketing managers to think strategically about AI.
BearingPoint reminds us that AI “is no longer an experimental technology”; moving from siloed experiments to transformations at scale requires structural, cultural and technological change. Marketing directors must play a strategic leadership role: they define the AI roadmap, empower teams and guide ethical adoption.
This guide describes the key concepts of AI, the strategic priorities for marketing departments and the stages of a transformation roadmap. It is designed for CMOs, transformation directors and executives who want to accelerate the responsible adoption of AI in their organizations.
Discover our Digital Expertise
Generative AI tools have reached over 16% of the world’s population in just one year. Countries that invested early in digital infrastructure (United Arab Emirates, Singapore, Norway, Ireland, France and Spain) have usage rates in excess of 60%. The rise of generative AI is transforming information retrieval: AI engines such as ChatGPT, Copilot or Perplexity are becoming primary sources for certain queries. Studies show that AI systems reduce organic click-through rates in Google by an average of 34.5%, and require content to be optimized for generative engines (Generative Engine Optimization – GEO).
Organizations that are scaling up are making measurable gains. BearingPoint observes that by 2028, 80% of marketing campaigns will be hyper-personalized thanks to AI, and more than two-thirds of content will be generated by AI tools. In B2B sales, more than half of all transactions are expected to pass through conversational interfaces by 2028, and AI could reduce the time spent prospecting by 50%. According to Boston Consulting Group, AI agents already accounted for 17% of the value created by AI in 2025 and are expected to reach 29% in 2028; the most advanced companies devote 15% of their AI budget to these agents.
These figures show that there is a growing gap between the “future-built” companies investing in AI and the laggards. Marketing departments that delay risk losing ground in terms of responsiveness, personalization and loyalty.
AI encompasses several families of technologies. In marketing, they include :
Predictive AI: statistical or machine learning models that predict behavior (purchase probability, churn, lifetime value). It is based on the analysis of historical and real-time data.
Generative AI: algorithms capable of creating text, code, images or synthetic data. Large-scale language models (LLMs) use self-supervised learning and multimodal architectures. Their effectiveness is based on three elements: self-supervised learning, the ability to process multiple types of data, and the power of cloud infrastructures.
Agents or co-pilots: autonomous programs that orchestrate multiple AI models to perform complete tasks (article writing, reporting, campaign tracking). BCG notes that “future-built” companies already allocate 15% of their AI budget to agents, and a third use them in production.
AI is data-intensive. To obtain reliable results and avoid bias, it is imperative to have a solid, governed data foundation. The Afges article reminds us that digital transformation is not an IT project, but a question of strategy and governance. Management indicators need to be aligned with business objectives, not isolated in silos. On the other hand, 74% of companies plan to deploy AI agent networks in production within the next two years, and 36% plan to automate at least 10% of jobs, but 84% have not given any thought to the evolution of functions. These figures illustrate the urgent need for proactive governance.
AI also raises risks: data protection (73% of respondents to the Deloitte study cite this risk), intellectual property (50%), AI governance and control, consistency of answers (46%). An ethical framework and validation process are essential to maintain trust.
Marketing directors must embody this transformation. BearingPoint stresses that they must adopt a strategic leadership role, align business and IT departments around a common vision, and guide ethical adoption. They are at the heart of customer transformation: reinventing customer journeys with generative AI, adopting agile organizations combining human creativity and AI efficiency, building AI-ready databases and governance adapted to hyper-personalization.
AI makes it possible to create hyper-personalized experiences. Generative AI is capable of producing tailor-made content for each segment, or even each individual. However, the customer journey is becoming more complex: in 2026, Internet users may discover a brand via ChatGPT, compare prices in Perplexity and then access the site only at the last moment. The consequence is an apparent drop in traffic but an increase in the quality of leads: fewer visitors but more decision-makers. Marketing departments therefore need to rethink their funnels, prioritize conversion and value measurement over mere volume, and optimize their presence on AI engines.
AI requires close collaboration between marketing, sales, data and IT. BearingPoint stresses the need to adopt agile organizational models that combine human creativity and AI efficiency. This means creating multidisciplinary teams, experimenting rapidly and fostering a test-and-learn culture. Marketing departments must also anticipate the arrival of AI agents: in 2025, agents already accounted for 17% of the value generated by AI.
Data is at the heart of AI. Generative AI is sensitive to data quality, structuring and governance. It’s essential to invest in data hygiene, source mapping, RGPD compliance and governance (roles, access rights, validation processes). The priority is to move from “proof of concept” to industrialization: Eulidia reminds us that organizations need to consolidate their data foundation, acculturate teams and move from POC to industrialized uses.
AI use cases are multiplying: document automation, business co-pilots, assisted software development, advanced analysis and anomaly detection. But these innovations come with risks: hallucinations, bias, security, AI Act compliance and platform dependency. An ethical framework and validation procedures need to be put in place before they go into production.
Traditional indicators (traffic, sessions, visit duration) are less relevant in the age of AI engines. ELLEVATE shows that by 2026, traffic curves will be falling, but leads and sales will remain stable. The key KPIs become: number of inbound leads, sales influenced and conversions by channel. The central question is now: “How many decisions have we triggered? Marketing managers therefore need to set up attribution models, calculate customer lifetime value and assess the direct impact of AI on results.
Before any initiative is taken, the maturity of the organization (data, skills, processes) needs to be assessed. AI audits, inspired by BearingPoint methods, can be used to take stock and identify gaps to be filled. A maturity grid can include data quality, tool integration, data culture, governance and use of AI.
AI transformation must meet clear business objectives: sales growth, cost optimization, improved customer experience, product innovation. Afges reminds us that digital transformation must be driven by corporate culture and values. Management committees must set priorities and arbitrate investments on the basis of expected returns.
It’s tempting to multiply proofs of concept, but to create value, we need to target use cases where AI brings a competitive advantage: hyper-personalization of campaigns, product recommendations, dynamic pricing, chatbots, content generation, paid ad optimization or predictive churn analysis. The JUPDLC article suggests identifying areas where AI can offer added value, such as data analysis, message personalization or automation of recurring tasks.
AI success depends on data. Databases need to be cleansed and enriched, a data lake or data warehouse created, and collection pipelines and strict governance put in place. The selection of technologies (open source LLM vs. proprietary models, marketing automation platforms, agent orchestration tools) must take account of business needs and security constraints. AI agents, which account for a growing share of value, must be integrated into existing architecture.
Once the use cases have been identified, it is advisable to launch pilot projects on a limited scale. The aim is to test models, measure ROI and identify any ethical or legal risks. The results will be used to convince stakeholders and refine the roadmap.
Moving from POC to industrialization requires investment in infrastructure and skills. BearingPoint stresses that the success of an AI transformation depends on the ability to deploy technology on a large scale, with a clear strategic vision and cross-functional coordination. This means defining development standards, automating ML/LLM pipelines (MLOps/LLMOps), ensuring security and compliance, and integrating AI into existing processes.
Acculturation is essential to avoid rejection. Afges insists on employee training and on the importance of supporting teams in the appropriation of tools. According to JUPDLC, the human aspect must not be neglected: creativity, empathy and human intelligence remain indispensable.
AI transformation is not a one-off project: it requires continuous improvement. Marketing departments need to define business-aligned KPIs (lead scoring, lifetime value, conversion rate), regularly analyze performance and adjust models. They must also anticipate regulatory (European IA Act) and technological developments (multimodality, autonomous agents, generative AI). The Eulidia report emphasizes that generative AI is becoming an impact multiplier, and that it needs to be integrated into a solid data strategy to become a genuine lever for transformation.
Marketing departments are at a historic turning point. Generative AI and autonomous agents are transforming search, customer journeys and value creation. Companies that invest today in data governance, team training and priority use cases will enjoy a sustainable competitive advantage. Conversely, those who are content to experiment risk seeing their gap with the leaders widen. The time has come to move from opportunistic AI to strategic AI: build a vision, establish a roadmap and orchestrate collaboration between humans and machines to create value at the customer’s service.