
Our new Martech for 2027 report launches on December 1, the culmination of five months of research by Frans Riemersma and me. I dare say it may be our best work yet, explaining the shifting sands of marketing and martech as we see them for the next 12 months:
Changing the calculus of build vs. buy — what should you build?
The sudden explosion of “personal agents” disrupting go-to-market
Six pace layers of marketing — and how AI is widening their gaps
Organizational (not technical!) disconnects holding context back
Diverging views of context in the C-suite vs. marketing operations
Martech’s commercial migration from apps to infrastructure
Maintaining coherence as AI amps marketing’s volume and velocity
We invite you to join us on December 1 at 11am EST — or on-demand any time after that the rest of the month — for a very human discussion of these issues and how to navigate them in what is sure to be a wild year ahead.
If you sign up for our launch event, you’ll get first access to the full 100+ page Martech for 2027 report, as well as two exclusive bonus research briefs. (By registering, you’ll also be supporting our research, as the sponsors for our event enable us to provide this premium analysis to you and the rest of the marketing community at no cost.)
In today’s newsletter, I’d like to give you a taste of what we’ll be covering on the topic of coherence. If “context” was martech’s word-of-the-year in 2026, my bet is “coherence” will be the word of 2027.
Coherence is martech’s next frontier
[Our coherence layer framework] originated from a friendly argument over dinner.
An executive at one of the major martech companies was insisting that their built-in UI was a better experience for marketing and sales people doing their day-to-day work than using Claude or ChatGPT in a more freeform fashion.
We skeptically pushed back. Not that there aren’t good native UIs or good cases where it’s preferable to use them. But for more and more tasks, it’s faster and easier to turn to your AI assistant or agent. And AI’s generative UI capabilities are going to just keep getting better.
We were persuasive enough in making that case that the executive put down their fork, leaned over with a slightly worried expression, and asked: “So if the data is in the warehouse, and the UI is in the AI, what’s left?”
“Plenty!” we replied and began talking about all the work required to acquire, manage, and activate context — the hard work of context engineering that platforms such as theirs could productize. More importantly, we added, to do that across a business that’s wrangling an explosion of AI activities and agents, in a way that keeps everything and everyone aligned and coordinated, there’s a desperate need for mechanisms to maintain coherence.
And that’s a huge platform opportunity. A huge infrastructure platform opportunity.
It was their turn to be skeptical. “Is there really that much to coherence?”
Hell, yes. It’s the difference between using AI in single-player mode — and the chaos of many people across a business disparately using it in single-player mode — versus running things in a synchronized multi-player environment. Because business is absolutely a multi-player game.
However, an impromptu, hand-wavy explanation over tapas wasn’t going to do that justice. And we were sure more people than that one executive would have the same question. So we spent the past several months mapping out a coherence layer to explain it more systematically.
Some people might think of this as an orchestration layer. Or that these capabilities are really just part of a context layer. Or that this should be the role of an AI harness for marketing.
We’re fine with any of those.
But coherence is the force that binds the business together.
(Yes, that last sentence is a Star Wars riff. If you’ve read our previous reports, you know that we like to pair our very serious analysis with not-so-serious asides and pop references. Mostly constrained to the footnotes.)
A comprehensive coherence layer framework
The illustration at the top of this article is a high-level sketch of the framework we’ll then walk through in our report. In short, we believe there are five elements to coherence:
Context. This is the spine of the agentic marketing organization. For internal teams and external customers and partners, what is known and what is intended? Our aspirational golden context.
Control. Who may do what? Permissions and decision rights. Operating rules. Degrees of autonomy. Budgets and cost controls. Escalation paths. Ownership and accountability.
Coordination. How do all the pieces fit together? The air traffic control of marketing. The more moving parts there are — and with AI, it’s a swarm — the more critical this capability becomes.
Cadence. The clocks and calendars on which the business (and customers) operate. What happens when, how often, and in what sequence? How do activities at different speeds stay synchronized? What exceptions interrupt their normal rhythms?
Calibration. How do we learn and change? How do observations and experiments at the edge of the organization get executed, evaluated, and understood? How do those lessons feed back into the evolution of the organization at multiple levels?
Each of these five elements operates at six different “pace layers” of marketing context: company, brand, campaigns & programs, adapted plays, decisioning & optimization, and – on the very edge of the algorithmic organization — real-time activation.
This gives you a 5×6 matrix of coherence capabilities to consider, which we’ve filled out with concrete examples. Here’s a teaser of what that will look like:

We also map several of the most popular categories of martech products against this matrix — CRM, CDP, data clouds, etc. — to help illustrate where this matrix is being served today… and where there are significant gaps.
This coherence layer framework could rightfully be an entire report of its own. But it’s just one of the fascinating topics we explain in-depth in Martech for 2027.
Hopefully enough to whet your appetite to register for a copy and a keynote?
Thank you to our sponsors GrowthLoop, Hightouch, Progress, SAS, Snowflake, Treasure AI, and WebEngage. While they have had no editorial influence on the content of our report, their generous support is what makes it possible for us to do this research and provide it to the community at no charge.
There definitely won’t be a dull moment in martech in ’27,
Scott
P.S. If you’re in the Bay Area on December 3 and registered for Martech for 2027, I’d like to invite you to a special in-person networking and fireside chat event I’ll be hosting with Treasure AI’s chief product & growth officer Rafa Flores at their groovy Treasure House headquarters in Mountain View.
Bring your gnarliest questions from the newly released report, and we’ll dig into them over tasty beverages and snacks.




