Google Ads Target CPA: the conversion threshold that matters

Google Ads Target CPA promises to manage your bids for you, based on your conversion history. The problem isn’t the promise, it’s when you switch it on. Google gives a precise threshold for judging whether the algorithm has learnt enough, and that threshold is counted in conversions, not in days on the calendar.

How Target CPA really adjusts your bid

Target CPA is one of the automated bid strategies, also known as Smart Bidding. At every auction, the algorithm sets a different amount depending on the device, the time of day, the location and the account’s conversion history. You no longer set a bid per keyword: you set a cost-per-conversion goal for the whole campaign.

Google spells it out in its documentation on Target CPA: the strategy can start without any conversion history at all. Nothing technically stops you from switching it on, even on a brand-new account. But being able to start and being able to judge the result are two different things, and that’s where the conversion threshold comes in.

The budget also follows a different logic from manual bidding. Google can spend up to twice your average daily budget on some days, to make up for a quieter day elsewhere in the month. A manual bid limit left active directly holds back that optimisation: the two settings don’t combine without losing the point of Target CPA.

Who can switch it on now, and who should wait

Two quick checks are enough to tell which side you’re on.

You can switch it on if:

• your conversion tracking has been set up and reliable for several weeks;
• your campaign already gets around thirty conversions a month or more;
• your budget can absorb a swing in daily spend without straining your cash flow;
• no manual bid limit has been left active by mistake on your ad groups.

You should wait if:

• your account has only just set up its conversion tracking;
• your campaign gets fewer than ten or so conversions a month;
• your budget is already tight, with no room for a day of higher spend;
• you have doubts about how reliable the counted conversions are, duplicates and internal tests included.

How many conversions a month you really need

Google is explicit about one figure only: to evaluate a Target CPA strategy, it recommends measuring performance over the last 30 days, with at least 30 conversions in that period. Below that, you’re judging on too little data to tell whether the algorithm is learning properly or just going through a bad patch.

The intermediate tiers in this table are a working assumption, not a classification published by Google: the only official figure remains the threshold of 30 conversions over a rolling 30 days.

Conversion tracking matters twice over here. A badly configured GA4 property that sends the same sale twice, as a key event and as a Google Ads conversion, inflates that figure without a single extra sale existing. My guide to GA4 key events explains how to check that every conversion counted in Google Ads matches a real sale, before you rely on it to change your bidding.

The checklist before switching on Target CPA

1. Check that conversion tracking has been in place for at least thirty days without interruption.
2. Count the real conversions over the last thirty days, excluding duplicates and internal tests.
3. Check that no manual bid limit has been left active on the ad groups concerned.
4. Plan a budget that can absorb doubled spend on the odd day.
5. Set a CPA target close to your current cost per conversion, not an ambitious figure from the start.
6. Leave it running untouched during the evaluation period, then review the following thirty days before adjusting.

What I think

Target CPA is neither a magic formula nor a trap. It’s a tool that needs one specific raw material: clean conversions, in sufficient numbers. The threshold of 30 conversions over 30 days that Google gives deserves to be taken at its word, not rounded off to “a few weeks” as you often read.

My view: the real question is never “should I switch on Target CPA”, it’s “does my account have the data for it today”. On an account still running below the threshold, I prefer tight manual bidding and work on conversion volume before handing the decision over to the algorithm.

Going automated too early is often repeating the mistake I see with Performance Max: believing the tool makes up for a lack of data rather than needing it.

And if the real issue is the time this monitoring takes you every month, let’s talk about it: I can help you put a figure on it.

Frequently asked questions

Can Target CPA run without any conversion history?

Technically, yes. Google states that the strategy can start without any history. But the algorithm then has no data to adjust its bids, so the first few days are trial and error. Conversion tracking needs to be in place and reliable before you switch anything on, history or not.

What happens if I switch on Target CPA with fewer than 30 conversions a month?

The algorithm keeps working, but with less data to learn from. Bids vary more from one week to the next, and a bad three-day run can be enough to skew your reading. It isn’t forbidden, it’s just harder to judge properly.

Can Target CPA spend more than my daily budget?

Yes, occasionally. Google can spend up to twice your average daily budget on some days, to make up for a quieter day elsewhere in the month. The monthly total stays capped, but the day-to-day variation can catch out a tight cash flow.

Is a manual bid limit compatible with Target CPA?

Technically, yes, but it directly caps what the algorithm can do. A limit set too low means switching on an automated strategy while stopping it from raising the bid, precisely when that moment would matter most.

Should I wait a specific number of days before judging Target CPA?

No, and it’s a common confusion. Google thinks in terms of conversions over a rolling 30 days, not weeks elapsed. A high-volume account can reach that threshold in ten days, a smaller account can take two months.