Scott Brinker and Frans Riemersma published the State of Martech 2026 last week. If you work in this space, read it. If you don’t, here’s what it says and why I think most CMOs are about to misread it.
By Jason Simon
The headline number is 0.79%. That’s how much the martech landscape grew in the last 12 months, after fifteen years of compounding at 34.5%. The instinct is to call this maturity. Saturation. The bubble finally popping.
That’s the wrong read.
Underneath the 0.79% net number, 1,488 new products entered the landscape and 1,367 exited. About 9% of the entire stack turned over in a single year. The market didn’t stop moving. It started metabolizing. Brinker’s word for it is renewal. New ideas arriving even as the old guard exits. The casualties skew small. 80% of exits had fewer than 50 employees, 71% under $10M in revenue. The narrative writes itself. Small fish being eaten by the big platforms. Tidy. Reassuring if you bet on the big platforms.
It’s also incomplete.
The legacy platforms that built this category, the email, campaign, and customer-data systems most enterprises have been running for a decade or more, are not all going to be there in their current form. Some are being absorbed into broader suites. Some are being pruned by parent companies that no longer see martech as core. Some are sitting in maintenance mode while their roadmaps quietly thin out. Consolidation among the survivors is the next phase of this story, and it doesn’t show up in landscape exit counts. It shows up in renewal conversations, in roadmap calls that get rescheduled, in product teams that stop hiring. The platforms most enterprises actually run on are not a fixed point. They’re moving too, just on a different clock than the small-cap exits Brinker is counting.
I have a particular lens on this. I ran tech sales at Oracle’s marketing cloud organization. Before that, Epsilon. Two of the bigger seats on the platform side of this industry over the last fifteen years. You learn something specific in those seats. You see, very clearly, the gap between what a platform can do on the day of signature and what it’s actually doing 18 months later. The gap is rarely the platform itself. It’s the activation work nobody scoped, nobody owned, and nobody measured. I sold against that gap for years without naming it. Moving to lead growth at a consultancy has clarified what it actually is.
It’s the work between the contract and the outcome. And the data in this report says the entire industry is now sitting in that gap.
The most striking data point, for me, is this one. 91% of marketers are using AI to produce content. Brinker reports 103 of 163 respondents, almost two-thirds, are doing nothing to verify the AI content they’re generating. 63% publish AI-optimized content for search and answer engines. 13.6% measure whether they’re getting included. The work is happening. The measurement isn’t. Brinker names this pattern across every category. SAS reports that only 8% of organizations have full confidence in their AI governance readiness. Production has decoupled from proof.
Not proofing the AI-generated work is a fool’s errand. The teams that allow this to continue are going to get hurt by it. Hallucinated claims in customer emails. Brand voice violations at scale. Compliance exposure nobody catches until legal does. The cost of cleaning up a problem you shipped to a million inboxes is higher than the cost of catching it before it sent. Every team I see treating verification as somebody else’s job is one bad campaign away from finding out whose job it actually is.
Brinker’s framing for this is the cleanest I’ve seen. He calls it the difference between AI dissolving a production constraint and revealing a context constraint hiding behind it. The bottleneck isn’t whether you can generate the email, the offer, the journey, the segment. The bottleneck is whether the right context, your products, your margin rules, your customer history, your brand voice, reaches the decision moment when it matters. He has a line in there I keep coming back to. A brilliant customer insight sitting untouched in a data warehouse is not an asset. It’s inventory that’s spoiling.
The temptation reading this report is to treat it as a buying guide. The market is reorganizing around AI; therefore buy the AI thing. The data argues the opposite. Content Marketing as a category posted 176 product removals against 139 additions. Largest net negative in the landscape. The first wave of AI content tools is being absorbed, acqui-hired, or quietly wound down. Brinker’s diagnosis: most of the customers who would have bought a standalone AI content tool already had one, attached to a platform they were already paying for.
Buying your way out is not always the move. Activating what you have is. And for the customers running on legacy platforms that are themselves consolidating, proving the value of what you have is the leverage you’ll wish you’d built before the renewal conversation.
Brinker names two disciplines for the activation work. Value engineering, which is figuring out where the revenue actually lives. Which customers, which journeys, which moments. And context engineering, which is making that value accessible at the moment of decision, by connecting the data, the content, the rules, and the systems that already exist. He puts it bluntly. Value engineering without context engineering is strategy without execution. Context engineering without value engineering is plumbing without purpose. Neither is a product. Both are work.
The honest answer most enterprises would give if you asked them what they’re getting back from their martech is that nobody can quantify it. McKinsey reported in 2025 that zero of 50+ Fortune 500 CMOs they interviewed could quantify martech ROI. Zero. Brinker’s 2026 data is the same finding from a different angle. The work to close it doesn’t look like buying. It looks like 90 days of unglamorous activation. Pulling the right data through the right channel into the right decision, then proving the lift. The companies pulling ahead are the ones doing this now, inside platforms they already own.
Brinker ends the analytical section with a line worth holding onto. AI is a commodity. Context is differentiation.
The chrysalis isn’t where you stop. It’s where you decide whether you metamorphose or dissolve. The CMOs I see winning right now aren’t waiting for a platform to package context for them. They’re not waiting to find out which of their current vendors will still be standalone in two years. They’re treating context as the work, not the wait.
This is the work my team at Solvenna does every day. We sit with CMOs and CIOs running enterprise stacks, and we close the gap between what those platforms could do and what they’re actually doing. The pattern Brinker quantifies, production decoupled from proof, is what we measure and fix. Our QuickLift engagements are 90 days, no new platform purchases, measurable revenue proof in the systems already in place. The new reality Brinker is describing isn’t a future state. It’s the work in front of every team running a real stack right now. Activate, measure, prove, repeat.
If your stack is showing, and Brinker’s data says most stacks are, that’s the work.
