The short answer: on a small account, start with Maximize Conversions once your conversion tracking is verified, then consider Target CPA after the campaign has built a stable record, as a rough guide around 30 or more conversions in the previous 30 days. Target CPA gives you cost control, but it needs data; Maximize Conversions needs less, but it will spend your whole budget whether the day is good or bad.
These thresholds are practical rules of thumb we use, not published limits. Google’s own guidance changes over time, and results vary by industry. Treat the numbers below as starting positions and check them against what your account actually does.
What is the difference between the two strategies?
Maximize Conversions tells Google to get as many conversions as possible from your daily budget. It does not know what a conversion is worth to you or how much you can afford to pay; it just spends the budget in the way it expects to produce the most conversions.
Target CPA (cost per acquisition) tells Google to aim for an average cost per conversion at your chosen level. It will bid higher for searches it expects to convert and lower for others, and it may not spend the whole budget if it cannot find conversions at your target price.
In current Google Ads, a target CPA can also be applied as an optional setting on a Maximize Conversions strategy, which blurs the line between the two. The principle stays the same: one pushes volume, the other adds a price ceiling.
Which one should a new campaign start with?
For a new campaign, there are three realistic choices:
- Manual CPC: you set bids yourself. It gives control but needs weekly attention, and it is slow if you have many keywords.
- Maximize Clicks: useful only for a short period to gather data, and risky because it does not care about quality.
- Maximize Conversions: usually the best default once conversions are tracked correctly.
Before any of them, check your tracking. A bidding strategy based on faulty conversions is worse than none, and we describe a quick test in our 20-minute conversion tracking check.
How much data does Target CPA need?
Automated bidding works by learning patterns in who converts. With few conversions there is little to learn from, so results swing. Our working guide is roughly 30 conversions in the last 30 days at campaign level before switching to Target CPA, and more if your conversion rate is volatile. Below that level, a target often produces erratic delivery.
If your campaign sits well below that, there are two options:
- Stay on Maximize Conversions and give it time.
- Combine similar campaigns so that data is pooled, instead of splitting a small budget over five campaigns that each learn slowly. This is the single most common fix we apply to small accounts.
How do you set the target?
Do not copy an industry benchmark. Work backwards from your own numbers:
- Work out the most you can pay for one customer or lead and still make a profit. That is your break-even CPA. (If you sell products, our guide to break-even ROAS shows the calculation.)
- Look at your actual CPA over the last 30 days with Maximize Conversions running.
- Set the first target at or slightly below your actual CPA, not far below. If you currently pay €40 per lead, a first target of €36 to €40 is reasonable; €15 is not.
A target far below your actual cost does not make Google find cheaper customers. Usually it makes the system stop bidding on most auctions, so impressions and conversions both fall.
What happens when the target is too aggressive?
The typical symptoms are a sharp drop in impressions, a budget that is only half spent, and “Limited by target” style notes in the campaign status. The campaign is not broken; it simply cannot find enough auctions at that price. The remedy is to raise the target, not to raise the budget.
Our rule: if spend falls below about 70% of the daily budget for a week after a target change, the target is probably too tight. Raise it by 10–20% and wait.
How do you move from Maximize Conversions to Target CPA?
- Confirm the campaign has enough conversions and tracking is clean.
- Read the average CPA of the last 30 days.
- Switch the strategy and set the target close to that average.
- Do not change budgets, ads or keywords in the same week, so you can see what the change did.
- Leave it for at least one to two weeks. Short windows mislead, particularly if your sales cycle is longer than a few days.
- Reduce the target gradually, in steps of 10–20%, only if performance is stable.
What about manual CPC?
Manual CPC is not obsolete. For tiny accounts with a handful of keywords, a low daily budget and a single-person oversight routine, it gives you a clear view of what each keyword costs. It also avoids the unpredictability of learning periods. The cost is time: you need to check bids at least weekly and use the search terms report to prune. We describe a lightweight routine in our guide to the weekly search terms review.
What if conversions are rare, like one or two a month?
For long-sales-cycle businesses, such as B2B services, a lead that becomes a customer after three months, or high-ticket work, the real conversion is too rare to bid on. Two approaches help:
- Track a micro-conversion close to the sale, such as a qualified call or a booked meeting, and optimise toward that.
- Upload offline conversions from your CRM so that Google learns which leads became customers. Our article on call tracking and offline conversion import covers the basics.
Where does Target ROAS fit in?
Target ROAS suits accounts with reliable purchase values and enough volume. It is a step beyond Target CPA, and it is not the right choice for lead generation unless lead values are meaningfully different. We explain the risks in our Target ROAS guide.
A simple decision path
Verify tracking. If you have fewer than about 15 conversions a month, start with Maximize Conversions and consolidate campaigns. If you pass roughly 30 a month, test Target CPA near your actual cost. If your sales cycle is long, feed better conversion signals before touching strategies. Whichever you choose, change one thing at a time and give it time. Most bidding “failures” we investigate are really impatience or messy tracking, not the strategy itself.
Mistakes that make smart bidding look worse than it is
Several habits sabotage automated strategies. Editing the target every few days prevents the system from settling; give each change a full one to two weeks. Splitting a small budget across many campaigns starves each of data. Pausing and re-enabling campaigns resets some learning. Running with a daily budget lower than roughly ten times your target CPA leaves the system little room to explore, because a single click can consume a large share of the day’s money. That ratio is our own working rule, not an official requirement, but when we see a campaign with a €20 daily budget and a €60 target, it is rarely delivering smoothly.
Finally, remember that bidding strategy is only one lever. Poor ad relevance, a weak landing page or loose keyword matching will limit results under any strategy. If your search terms report shows irrelevant queries, fix that first; the strategy cannot repair what you feed it.