People spend more time choosing a Google Ads bidding strategy than on anything else in the account, and it is rarely the thing holding them back. Bidding decides how much you pay for a search. It does not decide whether that search was worth having, and it cannot improve the page people land on.
Get the signal and the page right and almost any sensible strategy works. Get them wrong and no strategy rescues you.
That said, there is a right order to move through them, and it depends on how much conversion data you have.
This is for you if you are choosing between the options and want to know which actually matters.
The three Google Ads bidding strategy options worth using
Most of the list can be ignored for lead generation. Three do the work.
Maximise Conversions gets you as many as your budget allows, with no cost target. It is the right starting point, because it builds the conversion history everything else depends on.
Target CPA aims for a cost per conversion you specify. It is where most established lead generation accounts should end up, once there is enough data for the target to mean something.
Manual CPC still has a place in small or unusual campaigns, where you know something the system cannot. It is a control instrument rather than a performance one.
Everything else, including impression share targets and click maximisation, optimises for something that is not enquiries. Those metrics feel like progress and are not.
Choosing a Google Ads bidding strategy by conversion volume
Volume is the deciding factor, not preference.
Below roughly 30 conversions a month, smart bidding has too little to learn from. Delivery swings, costs jump around, and the system never settles. In that situation, either consolidate campaigns so one of them is properly fed, or run Maximise Conversions and accept some variability.
Above that threshold, Target CPA generally outperforms, because it has enough history to predict with.
Consolidation is the underrated move here. Four thin campaigns each getting eight conversions a month will all perform badly. Merged into one getting thirty-two, the same budget produces better results, because the system finally has something to work with.
Setting a Google Ads bidding strategy target that is not a wish
The most common Target CPA mistake is setting the number you want rather than the number the account has achieved.
Here is what I actually do, and it runs against most of the advice you will read. Run Maximise Conversions first and let a real cost per conversion emerge over a month or two. Then set Target CPA 20% to 50% above that figure. Not at it, and not below it.
So if an account is genuinely producing leads at $50, I set the target somewhere between $60 and $75.
That sounds like giving money away and it does the opposite. A target set at or under the real figure leaves the system no room, so it throttles delivery to protect the number. You get an excellent cost per conversion attached to almost no volume. A target set slightly above lets it keep buying through the expensive days and average out across the month, which is what you actually want from it.
Set it far below anything the account has ever produced and delivery collapses altogether. The system will not buy clicks it does not believe can hit your target, so you get very few conversions at a lovely cost per conversion, and almost no business.
Tighten it from above later, once delivery is stable and you want to push. Never start there.
For outside context, WordStream’s 2026 study of 13,474 US search campaigns put the all-industry average cost per lead at $66.69 with an 8.18% conversion rate. If your target sits far below your industry’s typical figure, expect throttling rather than a bargain.
The signal matters more than the Google Ads bidding strategy
This is the part people skip and it decides everything downstream.
Whatever strategy you choose optimises towards your conversion action. If that action is a plain form fill, every strategy will efficiently find you form fillers. Target CPA will hit your target beautifully while your sales team wonders why nobody answers the phone.
Send closed-won data back from your CRM and the same strategy starts optimising towards customers instead. Nothing about the bidding changed. What changed is what it is aiming at.
If your CRM does not send closed-won data back to the ad platform, you are asking the algorithm to optimise blind and then blaming it for the result.
So the order is: fix the conversion definition, then choose the strategy. Doing it the other way round is how accounts end up with excellent numbers and disappointing businesses.
Give any Google Ads bidding strategy change time to settle
Every strategy change restarts a learning period. Expect one to two weeks of unstable delivery.
That instability is where most of the damage happens, because performance dips, somebody panics, and the change gets reversed before it ever produced a result. Then the same thing happens again three months later.
Decide before you switch how long you will leave it and what you will judge it on. Write both down. It sounds bureaucratic and it prevents the most common expensive mistake in the account.
Do not rebuild the account because performance dipped for four days.
Where manual bidding still earns its place
Three situations.
Genuinely low volume, where automation has nothing to learn from and manual control is at least predictable.
Brand campaigns, where you know exactly what you want to pay for your own name and do not need a model to work it out.
And when you know something the data cannot, such as a customer type you can no longer serve, or a seasonal shift that has not happened yet.
Outside those, manual bidding is mostly nostalgia. It does not scale with your attention, and the accounts that keep it usually keep it because someone once had a bad experience with automation on badly-defined conversions, which was really the conversions’ fault.
What a Google Ads bidding strategy cannot do for you
Worth stating plainly, because a lot of hope gets placed here.
It cannot make an irrelevant search relevant. That is what negatives and match types are for, and no amount of clever bidding turns a job-seeker into a customer.
It cannot improve your landing page. If two people in a hundred enquire, bidding decides what those hundred cost, not how many of them act.
It cannot fix a weak offer. The system finds people likely to do what others have done. If few people want what you are selling, it will efficiently find you the few.
And it cannot invent data. Below a certain conversion volume there is nothing to learn from, and switching strategies in that situation just changes which kind of instability you get.
Bidding is the last few percent. It is worth getting right, and it is worth getting right last.
What to change this week
Four steps.
Count your conversions over the last 30 days per campaign. That number decides which strategy is appropriate, before any preference does.
Check what your conversion action actually is. If it is a plain form fill, fixing that matters more than any bidding change.
If several campaigns are under thirty conversions, consolidate rather than leaving them all starved.
Then, if you switch, write down the date and what you will judge it on in two weeks.
The wider piece on judging lead cost covers what all this is aiming at, working out what a lead is worth gives you the target to aim for, fixing Performance Max lead quality covers the same signal problem in automated campaigns, and the landing page is usually the bigger lever anyway. If you would rather this were set up properly, Google Ads management covers it, and you can book a teardown.