Scott Brinker has counted marketing technology products every year since 2011. His first chart had about 150 logos on it. By 2020 there were roughly 8,000, which most people assumed was the top. This year's count is 15,505, up just 0.79% on last year.
In other words, the market has stopped growing. Brinker's own headline declares peak martech, though he adds "maybe" in brackets.
That flat total hides a lot of churn. 1,488 products joined the chart this year and 1,367 left it, meaning roughly one product in ten turned over. And the ones that left weren't failed start-ups. More than half were founded in the 2010s, most had between one and fifty staff, and the largest group were turning over between one and ten million dollars. As Brinker puts it, they "found enough traction to become real businesses, but not enough to become inevitable ones."
If you've ever bought software from a company like that, you know what happens next. The product works fine, and then one day the renewal letter comes from someone else.
Where the Growth Went
Within that flat total, though, a few categories grew fast, and surprisingly they're the old ones. Content management and web experience grew 21.4%, more than any other category. Ecommerce platforms grew 19.9%. Analytics, integration and governance tools all grew too.
That's not a coincidence. AI puts new demands on the systems that hold your content, your products and your data. The website, which people spent a decade calling a solved problem, is back at the centre, because it now has to serve machines as well as people. (The next piece in this series is about that.)
Three Writers, One Direction
Three writers I read every week have arrived at the same conclusion by different routes.
Brinker's argument this year is that martech is moving from apps to infrastructure. For fifteen years, buying martech meant buying applications: a screen, a workflow, a login. Now that AI lets a marketing team build its own tools and agents on top of a platform, the value of the platform moves from the screen to the plumbing underneath it: data, identity, orchestration and governance. His advice to buyers is blunt: "Buy infrastructure. Build tailored apps and agents that run on top of it." As Jason Lemkin of SaaStr put it, in a line Brinker quotes approvingly, "The CRM decision is no longer a CRM decision. It's an AI infrastructure decision."
In July, Tony Byrne of Real Story Group published MarTech 2030, with a foreword by Brinker. Byrne has spent longer than most of us telling enterprises that they built their stacks "outward rather than downward": the money went into the engagement tier (email, web, loyalty, commerce), while the foundation underneath, which he calls a three-legged stool of customer data, content and decisioning, stayed thin. In his eras framework, most large organisations are stuck somewhere between MarTech 2.0, today's channel-by-channel stacks, and 3.0, a shared foundation of unified customer data, structured content, decision rules and governance. Only once you have that foundation do you get to build MarTech 4.0: experiences assembled from customer context and intent, where the website, the email and the app still matter, but no longer organise the work.
Juan Mendoza, in The Martech Weekly, has tracked the same shift from the vendor side. His 2026 essays follow the data warehouse slowly swallowing the customer data platform. Databricks launched a marketer-facing CDP with no platform fee. Salesforce agreed to buy Contentful, and then Braze, in the same year. And Gartner, as The Martech Weekly reported, now expects 80% of new enterprise CDP deployments to be embedded in or composable with a broader data platform by 2030. As The Martech Weekly summed up the consolidation: "If the CDP was built to remember and the CEP was built to execute, both now want to be the platform that decides."
Three different vantage points, all pointing to the same conclusion: the next decade of digital will be decided at the foundation, and the applications on top will come and go.
The market stopped growing, but the decisions didn't get smaller. They just moved down a layer, where they're harder to see and harder to reverse.
What 2030 Looks Like From Here
If you put those three trends together, you can see reasonably clearly what 2030 looks like.
First, there will be fewer vendors at the bottom. Brinker expects the shift to infrastructure to drive more consolidation, because "you just don't need 15,000 infrastructure options." That concentrates power. His warning on the Braze deal applies to every platform you depend on: "Everybody wants consolidation until the rent comes due."
Second, there will be more small things on top. The average organisation already runs 305 applications, up from 275 a year earlier according to Zylo's 2026 index. Increasingly, the new ones won't be bought off the shelf. They'll be built in-house, or by a partner, on top of your own data.
Third, channels will still exist, but they won't organise the work. Byrne's post-channel era doesn't mean the end of the website. It means the website becomes just one place where content, data and decision rules meet, alongside email, the app, the contact centre and, increasingly, somebody else's AI assistant.
And finally, there will be agents on both sides of the table. Yours will do more of the work, and your customer's will do more of the research and, before long, some of the buying.
Three Decisions to Make Now
You don't need a transformation programme to prepare for this. You just need to make three decisions deliberately, rather than letting them happen by default.
Which platform is the system of record for your customers?
Not which CRM, but which underlying data platform the CRM, the CDP, the email tool and the website all read from. If the answer is "several", fixing that is your first project.
Where does your content live in structured form?
A website is a way of viewing content, not a place to store it. If the only copy of your product information, your prices and your claims is trapped inside page templates, every AI initiative you fund will be working from scraps.
Where are decisions made, and who owns the rules?
You have to decide which customer gets which offer, how often, in which channel and why. Every platform vendor and most agencies would love to own that moment for you. In the fourth piece in this series, I'll argue it should be yours.
Behind all three sits a fourth question, which I covered in The Three Clocks: which of your suppliers will still be there to maintain any of this in 2030?
The Short Version
The martech market has plateaued at around 15,500 products, with heavy churn underneath. The growth is in content, commerce and data, which are the systems AI depends on. Brinker, Byrne and The Martech Weekly all agree that value is moving out of applications and down into the foundation beneath them. So decide where your customer data, your content and your decision rules live before you buy your next application, because the applications will keep changing, and the foundation won't.