Build vs. buy is the wrong question for martech AI

Arguably the spiciest debate across martech today: build it or buy it?

It’s the debate that has fueled the SaaSpocalypse firestorm. And there’s no denying that AI is collapsing the cost of custom development. But is building sensible, a source of competitive advantage? Or a rabbit hole that burrows a team further away from their real source of value, the customer?

The answer, in typical consultant-speak: it depends.

So Frans Riemersma and I went looking for what “it depends” on.

In February, we surveyed 208 marketing and marketing ops leaders on how they’re using AI across roughly 70 specific use cases. For each one, we asked whether they’re using AI embedded in an existing SaaS tool, a new AI-native product, a homegrown AI solution they built themselves, or none of the above.

Here’s a sneak peek at just one of the categories that we analyzed:

For many use cases, the answer was “yes.” Respondents used multiple solutions in parallel. Baseline audience targeting in their MAP. Specialized segmentation through an AI-native tool. A custom agent analyzing proprietary first-party data. Three approaches, one use case, all running in parallel.

Stack strategy has shifted from “standardize on a suite” to “orchestrate a portfolio of capabilities.” It’s one of several patterns in the data that cut across every use case we studied. Two more worth previewing here:

The long-running martech debate — platform consolidation versus best-of-breed diversification — has a 2026 answer: neither. Instead, the stack is stratifying into layers with different competitive physics.

AI-native tools are largely winning creation. Copy ideation, pitch decks, visual production, competitive intelligence. Tasks where the primary input is a prompt and some brand context, and where model quality is the product.

Incumbent SaaS platforms, such as HubSpot and Salesforce, are largely holding on to orchestration. Lead scoring and routing, pipeline management, channel delivery, offer personalization, etc. Coordinated tasks where data-meets-execution inside a CRM, a MAP, or an ecommerce platform. These systems increasingly serve as infrastructure that other commercial and custom AI agents rely on.

The more interesting action is where those physics collide. Agentic customer platforms such as GrowthLoop, Hightouch, and Treasure Data Treasure AI — many of them evolved from the CDP space — are making a different bet: sit on the warehouse, reason over first-party data, orchestrate activation, and skip the legacy application layer entirely. Third layer, or redrawn lines between the first two? One of the open questions of 2026.

This one surprised us.

Conventional wisdom says B2C leads tech adoption: mobile-first, social-first, personalization-first. On AI, the data inverts that pattern. B2B shows broader adoption across more use cases, with consistently lower non-adoption rates.

Two plausible reasons. B2B teams are chronically understaffed relative to their content and operational demands, so AI arrived as a relief valve. And a decade of CRM, MAP, CDP, and revenue intelligence investment had already built natural docking stations for AI capabilities.

When B2C does adopt, it builds deeper. For B2C, the customer-facing AI output is the brand experience. Off-the-shelf tools get you to 80%. The remaining 20% — brand voice calibration, proprietary guardrails, custom data integration — is where differentiation lives. That 20% doesn’t get outsourced.

A test you can run on any B2C use case: can a bad AI output become a screenshot? The higher the public consequence of failure, the more B2C builds custom (or doesn’t build at all).

There are more cross-category findings we uncovered: a Trust Ladder, a Governance Gap, and a RAG-Everywhere pattern. Then each of the six top-level categories of the landscape gets its own section and its own build-vs-buy chart, so you can benchmark your org against B2B and B2C peers on any of the 70 use cases.

But AI use cases are just one chapter in our new State of Martech 2026 report.

The cover of this year’s report is a chrysalis. That messy middle stage where the caterpillar has dissolved but the butterfly hasn’t yet assembled. That’s where the market sits right now, and the updated 2026 Marketing Technology Landscape tells the story.

I won’t give the headline number away. That’s one of the reveals saved for #MartechDay.

But I’ll tell you this: after 15 years of tracking this market, the top-line figure tells you almost nothing on its own. What matters is what’s happening underneath. Products are churning in and out at historic rates. Whole subcategories are reinventing themselves while others get quietly consolidated, acqui-hired, or wound down.

To make sense of it all, we organized a simple 2x2 for reading the landscape: Growth, Renewal, Stability, Decay. Every subcategory lands somewhere on it based on who’s entering and who’s exiting.

Everything above is just a sliver of a preview. The report goes much, much deeper, and #MartechDay is where Frans and I bring the whole picture together.

Here’s what you get when you register (did I mention it’s free?):

The keynote. An hour of Frans and me walking through the findings, with additional material and live analysis that go above and beyond what’s published in the report.

7 candid sponsor conversations. After the keynote, Frans and I are hosting candid interviews with executives from each of our sponsors: GrowthLoop, Hightouch, Knak, MoEngage, Pega, Progress, and SAS. We dig into what they’re seeing in their own businesses and with their customers.

First access to the State of Martech 2026 report and the updated 2026 Marketing Technology Landscape. Registrants get both, ahead of the public release.

Can’t make it live? Register to watch at your convenience — the whole event will be available to you on-demand throughout May.

Don’t miss out — register now.

See you soon,

Scott