Eighty-six integrations, four that work

86 tools across 10 categories. That was the compatibility count on one growth-stage marketing site I reviewed. The hero line read integrates with your entire stack. Partner logos ran across three pages. I asked the head of partnerships how many of the 86 the engineering team supported, with docs maintained, edge cases handled, and support staff trained. He said four. The rest had public APIs and a plausible path to connection. No customer had ever finished an install without an engineer on the call.

I asked the head of sales how often buyers raised specific integrations during evaluation. Most of the time, he said, especially in mid-market and enterprise deals. The buyer would name a tool from the site, usually not one of the four, and ask how the integration worked. Sales had learned to hedge. We have a partnership. It's on the roadmap. We can build a custom connector during onboarding. The hedges were not lies. They were also not what the buyer wanted to hear. Late-stage deals died on that gap with real regularity, the same way a buyer walks when the integration nobody owned breaks after signature.

The pattern was consistent. The logo appears. The buyer reads it as they support this. The buyer asks for detail. Sales reveals the gap. The buyer decides the company exaggerates. Once that read lands in one place, it spreads. Credibility took the hit on the integration claim and on every claim next to it.

This is a common B2B positioning mistake, and it works against the goal it chases. The pull toward claiming broad compatibility is strong. Logo grids look impressive. Integrates with everything reads better than integrates with these four tools. Marketing gets graded on how the site looks, not on whether the look closes deals, which is a different problem from attracting the customers you want.

So claim only what is real. A real integration has documentation, dedicated maintenance, customer references, and an install that runs without engineering. Most companies have a dozen at most. A site listing those dozen with substance, case studies, technical docs, and named customers, beats a site listing eighty-six with none. Credibility compounds across the rest of the positioning.

The test is operational. Can a customer install from the docs alone, in under thirty minutes, without contacting support? If yes, the integration is real. If every install needs a setup call, a config ticket, or an engineering escalation, it's a partnership, not a product.

Partnerships are fine to claim, claimed honestly. We have a partnership with X differs from we integrate with X. Marketing conflates the two because the conflation looks better on the page. It costs credibility every time a buyer checks.

Cut the list to the integrations that survive a test call. Four, or ten, or twelve. Make those the best-documented, best-supported claims on the site. A buyer who tests one and finds it true stops testing the rest.