1. AI agents are becoming productive – but only with ‘human in the loop’
2026 is the year when AI agents will transition from pilot projects to everyday use. This does not just mean ‘AI writes texts’, but rather an agent that takes on tasks along a process chain, for example:
Find target group → Vary content → Run campaign → Test → Improve
For decision-makers, the question is not so much ‘What can AI do?’ but rather: How can it remain controllable, brand-compliant, data protection-compliant – and how can its benefits be measured? This is precisely why human-in-the-loop (human approval at the right points) is becoming the standard. HBR clearly shows that trust is a limiting factor here, based on current trust data. (HBR, 2025)
What this means in concrete terms:
- Approvals: What can go live automatically?
- Quality assurance: Tone, brand fit, facts, errors
- Data protection & compliance: Consent, target group logic, channel rules
- Emergency rules: Stop/failover when data or results are uncertain
What will become more important in 2026:
- AI agents for clearly defined process steps (instead of ‘AI everywhere’)
- Human-in-the-loop as standard: approval, quality, compliance
- Guardrails for brand, data protection, frequency, risk
- Measurement of efficiency, impact and risks
2. The right message at the right moment – instead of more content!
The next level of personalisation is not ‘more campaigns’ but more context. In other words, status, behaviour, timing, channel rules and preferences must come together to make communication truly relevant.
Scott Brinker calls this ‘context engineering’. In practice, this means above all: better decisions instead of more content. Who decides when which customer receives which message – and clearly explains why?
What will become more important in 2026:
- From campaign thinking to situation thinking: What does the customer need right now – at this moment and on this channel?
- Bundle signals instead of evaluating them separately: Consolidate behaviour and status from email, apps, the web, CRM and service so that all channels work on the same basis.
- AI only optimises if the basis is right: AI can improve timing, content and frequency – but only if data, consent and definitions are consistent.
3. Breaking down data silos, improving data quality: the foundation for scalable AI
Many companies invest in new tools – and wonder why personalisation, cross-channel orchestration and AI still don’t scale properly. The reason is often not a lack of technology, but data silos and inconsistent foundations: teams work with different data sets and different definitions – and in the end, there is no reliable overall logic.
Typical symptoms include:
- Different status definitions (e.g. ‘active’, “inactive”, ‘migrated’)
- Fragmented consent and preference logic
- Unclear responsibility for data quality
- No uniform ‘customer truth’ for marketing, service and sales
The more AI intervenes in processes, the greater the impact: AI amplifies good data – and bad data. In addition, the governance risk is growing: 54% would use AI tools even if they were not authorised (‘shadow AI’) – at the same time, only 36% feel sufficiently trained. (BCG, 2025). Gartner also shows that a lack of foundation is a real scaling killer: according to forecasts, over 40% of agentic AI projects will be discontinued by the end of 2027 – due to costs, unclear business value or insufficient risk controls, among other reasons. (Gartner, 2025)
What will become more important in 2026:
- Standardising data: consents, preferences, status, interests, contact frequency
- Clarifying responsibility: professional and technical
- Defining measurement logic: success criteria and verification
4. Composable Martech: Modular Tech Stack – integration becomes a success factor
In 2026, companies will buy fewer ‘features’ and more connectivity. A modular tech stack is attractive because new requirements – especially AI components – can be integrated more quickly without having to replace everything.
But modularity is not a sure-fire success. Those who think modularly need integration as a core competence:
- clear API standards
- clean data models
- governance via tools and data flows
- an operating model for orchestration
In the context of AI agents in particular, integration determines whether agents can truly act – or only ‘analyse’.
What will become more important in 2026:
- Integration beats features: systems must be connectable
- APIs, data models and standards are a must
- Modularity requires governance (rules, responsibility, transparency)
5. AI consumer agents are changing access to customers
In 2026, AI consumer agents will increasingly take over the pre-selection of information, offers and news for end customers. Users will no longer see every marketing message directly, but only what their agent considers relevant, trustworthy or useful.
For marketers, this is a clear shift: relevance is partly decided before the message is delivered. Therefore, content, offers and signals must be prepared in such a way that they are correctly understood and classified by agent-based filters.
What will become more important in 2026:
- structured, machine-readable content (clear offers, metadata, value propositions)
- First-party data, preferences and intent signals as a basis
- Close integration of CRM, content and messaging systems for consistent signals
- Systems that evaluate and prioritise relevance – not just send content
Conclusion
In 2026, the winners will be those who set up martech as a complete system: using AI sensibly, keeping data clean and truly bringing channels together. AI agents can speed up processes – but only with clear guidelines and responsibility. Context determines relevance. Data quality and integration determine whether the whole thing scales. And AI consumer agents are also shifting access to customers: structure, clarity and consistent signals are becoming more important than ‘even more campaigns’.
What marketers should do now
- Make AI agents productive – clear use cases, guidelines, approvals
- Prioritise data quality – break down silos, maintain consent and preferences consistently
- Think context instead of campaigns – orchestrate customer moments
- Make integration mandatory – APIs, standards, data models, governance
- Optimise for AI consumer agents – provide content, offers and signals in a structured way
Sources:
-
-
SAS (2025): Marketers and AI: Navigating New Depths (Coleman Parkes Research)
https://www.sas.com/content/dam/sasdam/documents/20250124/marketers-and-ai-navigating-new-depths.pdf -
McKinsey (2025): The state of AI in 2025: Agents, innovation, and transformation
https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai -
BCG (2025): Companies Must Go Beyond AI Adoption to Realize Its Full Potential
https://www.bcg.com/press/26june2025-beyond-ai-adoption-full-potential -
Gartner (2025): Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027
https://www.gartner.com/en/newsroom/press-releases/2025-06-25-gartner-predicts-over-40-percent-of-agentic-ai-projects-will-be-canceled-by-end-of-2027 -
Harvard Business Review (2025): Workers Don’t Trust AI. Here’s How Companies Can Change That
https://hbr.org/2025/11/workers-dont-trust-ai-heres-how-companies-can-change-that
-
Whitepaper: The future of marketing with AI consumer agents

In our white paper: The future of marketing with AI consumer agents, we provide you with, among other things
- exciting examples from B2B and e-commerce of the impact of AI consumer agents on the customer journey.
- concrete instructions so that you can adapt your marketing to the new conditions in the best possible way and remain visible.
To download the white paper, simply fill out the form.


