Apple's App Tracking Transparency prompt is aimed squarely at consumer apps, and the loudest complaints are coming from e-commerce and mobile gaming. But if you run paid social for a B2B software company, the fallout reaches you too — and the useful response is not a new tool. It's finally admitting how much of your attribution was always an educated guess.
What actually breaks
The mechanism most paid social reporting depends on is a durable identifier that follows someone from an ad impression to a conversion days later. Restrict that identifier and the platform stops being able to close the loop for a growing share of your audience. What you see instead is modelled conversions, longer reporting delays, and shorter attribution windows. For B2C with a same-session purchase, that's painful. For B2B, where the gap between first ad view and closed deal is measured in months, the window was never long enough to capture the truth in the first place.
The uncomfortable part
Most B2B attribution was already fiction. A ninety-day enterprise sales cycle involving six people, three of whom never click an ad, cannot be meaningfully resolved by a last-touch model with a seven-day window. What ATT changes is not the accuracy of your reporting — it's your ability to keep pretending. In one account I reviewed this spring, paid social was credited with a low single-digit share of pipeline by the platform, and named unprompted by roughly a third of new customers when someone actually asked them.
- Attribution windows were always shorter than B2B buying cycles
- Committee members who influence the decision rarely click anything
- Platform-reported conversions are marketing for the platform, not for you
- Modelled data is not the same as measured data, however confident the dashboard looks
What to measure instead
Move up a level. Stop asking which ad produced which lead and start asking whether spending money on a channel makes the overall number move. Hold spend flat for a period, then cut it entirely in one region or segment, and watch what happens to inbound volume and self-reported source. It's cruder than a dashboard and considerably more honest. Pair it with a mandatory free-text 'how did you hear about us' on every form, and review the answers monthly as a team.
Server-side and consent, briefly
Yes, you should implement server-side conversion tracking, and yes, you should get your consent banner working properly rather than pretending the problem is someone else's. Both will recover some signal. Neither will restore the illusion of precision, and I'd argue you shouldn't want it back. Plan your budget on the assumption that a meaningful chunk of your influence will always be invisible, and you'll make better decisions than a team with a beautiful dashboard and misplaced confidence.
The teams that will handle this well are the ones who were already suspicious of their attribution model. The ones in trouble are those who built a budget case on numbers that were never as solid as the decimal places suggested.
Want to apply this to your business?
Book a free 30-minute strategy call and we'll diagnose your specific growth blockers together.
Book a Free Call