June 22, 2026 · 6 min read
Why We Switched Attribution Vendors Twice in One Year
Two vendor switches in twelve months looks like bad decision-making from the outside. From the inside, each switch was the correct call given what we knew at the time.
Switching attribution vendors once in a year is a normal, if annoying, part of MarTech work. Switching twice makes a client's finance team start asking pointed questions in QBRs. Both switches were justified, and the story of why is more useful than either vendor's feature list.
Switch one: the vendor that couldn't see mobile app conversions
The first vendor was excellent on web attribution and treated mobile app conversions as an afterthought — a basic SDK integration with none of the cross-device stitching the web side had. The client's revenue was shifting toward app purchases, and the attribution data was increasingly describing a business that no longer matched reality. We switched to a vendor with genuinely strong mobile deep-link attribution.
A vendor doesn't have to be bad to become the wrong choice. It just has to stop matching where your business is actually moving.
Switch two: the vendor whose pricing model punished growth
The second vendor solved mobile attribution well, but priced on a per-tracked-user basis that scaled faster than the client's revenue did. A 40% increase in active users — genuinely good news for the business — came with a pricing tier jump that ate most of the margin gain. We switched again, to a vendor with flat-rate pricing at the client's volume tier, accepting a slightly less polished mobile feature set as the tradeoff.
- Attribution vendor evaluation needs a 'does this scale with us' pricing test, not just a feature bake-off, run against a realistic 18-month growth projection.
- Each switch cost roughly three weeks of parallel tracking to validate the new vendor's numbers against the old before cutting over.
- We now insist on a pricing model walkthrough at 2x current volume before signing any attribution contract.
- Two switches in a year is a sign the underlying business was changing faster than the first vendor choice anticipated, not that the vendor choices themselves were careless.
If I'm being fully honest, the deeper lesson is that we should have asked harder questions about growth trajectory before the first vendor pick, which would have let us skip straight to a vendor that handled both mobile and pricing at scale. But you don't always know what you don't know until the business has already outgrown the first answer.