Yesterday marked another state of the union for the world of MarTech. Frans Riemersma and Scott Brinker gave another presentation about what's happening in MarTech, with a focus on where we could be heading in 2026. And yes, it's still all about AI, and its impact on MarTech. I noticed a few things during the event that I want to share with you as my key take aways:
Kinda, but not really. Only 30% of companies are actually replacing SaaS tools with AI. In contrast, 85% is adding extra functionality with AI. AI is mainly extending functionality rather than replacing what SaaS tools do.
Also, the largest source of new agents in use comes from large-scale MarTech SaaS platforms already in use, such as CRMs and CDPs.
But what about vibe coding?
Yes, there is that, and it's definitely working, but it has its limits. Vibe coding mainly works when it's used to replace design or front-end dev work. Making your own landing page or simple web app, for instance.
But, right now, vibe coding can't be used to develop your own enterprise software. And where does the vast majority of MarTech SaaS spend come from? Indeed, enterprise software.
The landscape of MarTech agents can be divided into three categories:
The most considerable category here is still the first one, with an average of 3,5 of these agents being used. Followed by the agents for marketing interactions, second, and the Agents for customers, third.
However, the prediction is that this order will flip. Increasing awareness and the rise of MCPs (model context protocols that enable easy integration with other tools) could significantly boost customer adoption of agents.
The most significant use case for AI in MarTech remains content production. This is the only use case for which more than 50% of companies use AI. These companies also see a shift happening from "web-based content" to "web-based and optimized for LLM's content".
McKinsey has predicted that as much as 50% of web traffic before GenAI could be replaced by LLM chats and agents browsing the web.
Therefore, it is interesting to see that, from the same set of companies, which were asked what types of AI tools they use to deal with this. Over 60% said they were using tools to create LLM-optimised content, but less than 20% said they were actually measuring LLM inclusion in their content.
When asked what challenges companies faced in effectively implementing AI tools, these were the top 3:
I don't know about you, but these sound highly familiar. They're the same challenges that pop up every time a new MarTech wave hits.