What to check before you touch anything
- What's the campaign's current status — is it already flagged "Limited by budget" or still "Learning"? Lowering the target on a campaign that's already constrained is riskier — the effects will stack.
- Do you have enough conversion history (ideally a stable 30+ conversions over 30 days) for the system to actually have a statistical basis to relearn after the change?
- Are there seasonal factors in play right now (peak or slow season) that could skew your read on the result? Better not to test on an unusual stretch of the calendar.
- What share of your conversion volume comes from "marginal" segments by cost (geo, device, time of day)? Those are usually the first to shrink when you tighten the target.
- Are there other changes planned at the same time (budget, keywords, creative)? You want to isolate the effect of the tCPA change to actually see what it did.
Possible approaches
- The standard playbook, both from Google and common practice, is to move the target in steps of no more than 10–15% at a time, with 1–2 weeks between changes to let things settle.
- Instead of lowering the target for the whole campaign at once, you can test it through a Campaign Experiment on a slice of traffic — comparing volume and CPA before rolling it out account-wide.
- It often helps to widen your data pool first (raise the budget, remove other constraints) — that way the system lowers CPA by getting sharper at picking the right audience within a bigger pool, not by simply cutting volume.
- Some businesses time their tCPA tightening to a period of organic demand growth (a seasonal peak) — conversion volume is already rising on its own, so a tighter target is less painful to absorb.
- If a chunk of your conversions are low-quality and don't move the needle for the business, it can be more effective to clean up your conversion actions or audience targeting first — tightening the target then becomes a natural side effect of cleaner data, rather than a standalone risk.