A campaign can reach 50 conversions and still be unready for target CPA. If those conversions are junk form fills or the CPA swings wildly each week, the threshold has only produced a larger pile of noise.
A lower-volume campaign may be more ready when its leads are real, its recent CPA is stable and its budget can support the target. That is why the familiar conversion-count rule should be treated as a warning light, not a switch.
The count is standing in for stability
Target CPA asks the bidding system to preserve a cost constraint while finding more conversions. That instruction becomes useful only when the recent account behaviour gives the constraint some meaning.
A campaign producing 20–25 conversions a month with a repeatable CPA is not the same as one producing the same total through bursts and droughts. The first may support a careful test. The second is likely to lose volume when an aggressive target is placed on top of an already unstable pattern.
This is the useful idea hidden inside the fixed 30- or 50-conversion advice. More data generally makes performance easier to judge. But the number cannot tell you whether the conversions were valuable, whether the weekly pattern was stable or whether the budget allowed the campaign to explore demand. Turning a rough comfort level into a law replaces judgment with arithmetic.
A target cannot create budget headroom
A small service business described a daily budget of £15.50, an average cost per conversion of £16.25 and a target CPA recently raised to £18. The campaign had begun producing daily enquiries, but the business could not simply multiply its spend to satisfy a generic budget rule.
That is the real constraint. When the daily budget is close to the cost of one conversion, the campaign has little room to produce several outcomes in a day. Raising the target does not remove the harder budget ceiling. Yet a universal demand to spend three or five times the target is not a business answer either; the economics may not support it.
The decision is whether the existing budget can generate enough comparable opportunities for a target to regulate. If it cannot, changing the bid strategy will not manufacture evidence. It may only reduce reach or move volatility somewhere else.
The conversion has to deserve optimization
A form fill is not automatically a useful business outcome. If bot submissions, job seekers or poor-fit enquiries are recorded as success, automated bidding receives a precise instruction to find more of the wrong thing.
Before debating Max Conversions versus target CPA, inspect what the recorded conversions became. Qualified leads, paid jobs and obvious junk should not remain indistinguishable. Even rough lead-quality feedback can improve the decision because it reveals whether the apparent CPA represents value or merely activity.
This is where a cleaner signal can matter more than a smarter bid setting. Target CPA cannot repair a conversion definition that rewards waste.
Make the switch a controlled constraint
A defensible target-CPA test starts with the campaign's observed reality, not the CPA the business wishes it had:
- Confirm that recent weekly conversion volume and CPA are reasonably stable.
- Verify that the conversion action represents acceptable lead quality.
- Set the initial target near the recent actual CPA rather than forcing an immediate efficiency gain.
- Leave the test unchanged for a defined observation window.
- Decide in advance how much lost volume or worse lead quality will trigger a rollback.
Some low-volume campaigns will still need more data before this test is sensible. Others can work below the popular thresholds, especially when the signal is consistent. That makes one universal floor indefensible.
The switch is ready when the campaign has something stable and valuable to constrain. A conversion count can suggest that moment. It cannot make the decision for you.