Free tool · Google Ads · Runs in your browser
Bid strategy selector: what you are eligible for, and what it will cost you in control
Most bidding advice is a preference dressed as a recommendation. “Always use smart bidding” is wrong when your conversion tracking is broken, and the automation will not tell you, because optimising toward a wrong number confidently looks exactly like optimising well. Eligibility decides more than preference does. This tool checks what your account can actually support, names the strategies you are not ready for, and says plainly which control you are giving up.
Your account, honestly
Answer as your account is, not as you would like it to be. Every answer here changes what you are eligible for.
Tested means you have seen a real conversion arrive and attributed correctly.
Conversions recorded in Search, not clicks or sessions.
Value means a real number per conversion, not a placeholder.
The objective only. How much bid control you keep is the next question.
Headline
ReadyThe recommendation with its reasoning, the eligibility table for every strategy, and the CSV export unlock below.
Unlock the full recommendation
FreeGet the recommended strategy and why, an eligibility table for all seven documented strategies naming what each one requires, exactly what control you give up, the target and volume trade explained, and a CSV of the whole thing.
Your strategy
-Why this one
-
Every strategy, and whether you can use it
-Eligibility is a floor, not a ranking. A strategy you are eligible for is not automatically the right one, and a strategy you are not eligible for is not a failure — it is a prerequisite you have not met yet.
What to do first
-The target is a lever, not a promise
-
Export
How this decides
Short answer
Eligibility first, preference second, and the two are different questions. A bidding strategy you cannot use is not a choice you declined. Most bad bidding decisions are a strategy applied before its prerequisite was met.
The order below is fixed, and deliberately so. Each step can rule out options before preference is considered at all, because wanting a return target does not make your account able to run one.
- Can the account measure a conversion at all? No conversion tracking, and every conversion-based strategy is off the table regardless of goal. This is checked first because it invalidates everything after it.
- Is there enough conversion history to learn from? Smart Bidding needs history before it behaves. Thin data does not forbid it; it changes what you should ask it to do.
- Is there a value, or only a count? This decides whether value-based and return-based strategies are possible at all.
- What are you optimising for? The stated goal, checked against the above.
- How much control do you want to keep? The final filter, because it is the only one that is genuinely about preference.
Why control comes last
Control is a preference, and preferences are legitimate. But it is the only step that cannot rescue an account that is not ready, so applying it early produces exactly the outcome this tool exists to prevent: a campaign on a target strategy, optimising a misconfigured conversion, with nobody able to tell for several weeks.
There is also a version of this that runs the other way, and it is legitimate too. If you are buying data in a new market, deliberately buying information rather than efficiency, a return target is the wrong instrument and Manual CPC is the right one. The tool asks what you are optimising for so that choice can be made deliberately.
What this tool will not do
- Predict your cost per click, conversions or return. A bidding strategy is a set of rules, and a rule does not forecast a market.
- Recommend a specific target figure. Your margin is not a value this page can read.
- Set anything in your account. There is no connection, and there will not be one.
- Claim to know your conversion history, which is account data this page never sees.
The strategies, one at a time
These are the documented strategies. What each optimises for and what it requires are platform behaviour. Whether one is right for you is not.
Manual CPC
You set every bid. No conversion tracking required, no history required. In exchange, nobody is optimising: you are. The strongest argument for it is that it works when the thing you are optimising for is not a tracked conversion at all.
Maximise Clicks
Google sets bids to get the most clicks, optionally under a maximum CPC limit you set. No conversion data required. It optimises volume, so it can spend the whole budget on clicks that never convert unless you cap it and watch what it does with the cap.
Maximise Conversions
Google sets bids to get the most conversions within the budget. Requires conversion tracking and history. Reports the average cost afterwards rather than being given it in advance.
Target CPA
Maximise Conversions with a cost per acquisition you set as a constraint. The system bids more aggressively where it expects to hit the target, and more cautiously where it does not. The volume consequence is the whole point, and it is the part most people are surprised by.
Maximise Conversion Value
Maximises total conversion value rather than conversion count. Requires conversion values, and it is the right choice when a small number of high-value conversions matter more than a large number of cheap ones.
Target ROAS
Maximise Conversion Value with a target return on ad spend as a constraint. Requires conversion values. Note that it optimises the ratio, which means it can quite rationally stop bidding on the small orders to protect the average.
Enhanced CPC
Google's older hybrid, which bids for clicks but raises the bid on clicks it predicts will convert. It sits between Manual CPC and the conversion-based strategies and is documented as being retired in favour of the others, so treat it as something to move away from rather than towards.
What each one gives up
Manual CPC → you give up optimisation
Maximise Clicks → you give up efficiency
Maximise Conversions → you give up cost predictability
Target CPA → you give up volume
Target ROAS → you give up volume and the small orders
That list is the useful part of the choice, because every strategy is a trade rather than an upgrade. The strategies that appear to be strict improvements are the ones that hide which thing they cost you.
The one that is usually wrong
Target ROAS on a small account is the most common misapplication in the category. The strategy optimises an average ratio, so with thin data it protects that average by declining the searches it cannot predict, and a new campaign produces no data at all. A brand campaign with a few high value orders is the classic case: the average is excellent and the volume is nearly gone.
Target CPA is not a promise
It is worth being blunt about this, because the term invites the opposite reading. A target does not describe what the campaign will achieve. It describes what the system should refuse to pay more than, on average, in exchange for volume. Set it tight and the average gets closer while the traffic goes away. Set it loose and the traffic comes back and the average gets worse. Both are the lever working, not the lever failing.
What a target actually does
A target is an input to the bidding system's decision about where to push. It does not set your average cost; it changes which auctions the system finds worth entering aggressively. Everything you have heard about targets follows from that one distinction.
Why a tight target reduces volume
Because the system declines to chase conversions it predicts will miss. Every search it judges unlikely to come in under the target is bid on conservatively or not at all. That is a large fraction of a campaign's opportunities, and removing it removes real conversions along with the expensive ones.
Why a loose target increases volume
The same mechanism in reverse. A target set above your actual cost is not a hard expectation at all; it is a permission slip, and the system takes it. Volume rises and the average settles wherever the market puts it.
How to set one that means something
- Derive it from your own margin, not from a competitor's or a blog's number. Break-even cost per conversion comes from your margin, and that is the arithmetic worth doing — see the ROAS calculator.
- Start a little looser than your break-even and tighten once the account can see volume. The reverse locks the campaign out of the auction.
- Change one thing at a time, and give it long enough to show. A target changed weekly is a target being used as a knob rather than a decision.
- Judge it against your own conversion lag, not a week.
The volume trade, concretely
Almost every surprise people have with a target is this trade, arriving unannounced.
Conversions fell after I added a target
That is the lever working. Fewer conversions at a better average cost is the intended outcome. If the average was the need, the target is doing its job. If volume was the need, the target is too tight.
The average rose after I loosened the target
Also the lever working. You bought volume and paid the price in efficiency, which is the trade rather than a failure.
Nothing happened at all
Usually one of three things: the campaign is too small to generate enough conversions for the system to have anything to learn from, the target is so loose it is not constraining anything, or something else in the account is limiting delivery — budget, schedule, targeting, or an ad that is not eligible to show.
The conversions went to the wrong campaigns
A target is applied per campaign. Consolidating several thin campaigns into one so the system has enough data is a common fix, and it is also how a good campaign starts funding a bad one. Check where the volume landed before treating it as an improvement.
So which position is right?
Short answer
Whichever one your margin can afford. That is a question about your unit economics, not about bidding, which is why no tool can answer it for you. The account audit scorecard will show you what you are currently paying and whether it clears the bar.
Data readiness, honestly
Short answer
Conversion-based bidding needs conversion tracking and a history of conversions. Without either, the automation is optimising toward nothing, and it will do so confidently rather than obviously.
The failure nobody sees
A misconfigured conversion action is the most expensive quiet fault in paid search. The campaign runs, spend goes out, clicks arrive, and the reporting shows activity. Smart Bidding then optimises toward whatever the broken action reports, efficiently and indefinitely, because from its point of view the tracking is working perfectly.
This is why the first question in this tool is not about goals. It is about whether the number being optimised is real.
Partial tracking is the same failure with extra steps
If only some conversions are recorded, the system is not merely missing conversions, it is optimising for a biased sample. The recorded ones may be the easy ones: the ones that happen online, immediately, without a phone call. That is a pattern, and the automation will find it.
What "enough history" means
Google requires conversion tracking and a conversion history for the conversion-based strategies to work, but it does not publish a specific number of conversions that unlocks them. The bands in this tool are practical guidance, labelled as such, because presenting a rule Google does not publish would be the exact failure this project is built against.
- None. Conversion-based strategies will not run. Start with a counting strategy or manual bids and fix tracking first.
- Thin. Expect unreliable behaviour. Use Maximise Conversions without a target, or Maximise Clicks, and let the account learn before constraining it.
- Workable. A conversion count target becomes defensible. A return target is still thin, because it needs value on both sides of the ratio.
- Comfortable. Constrained strategies are viable. Still watch volume, because the constraint is what causes it.
Conversion value, specifically
Value tracking is what separates a count-based account from a value-based one, and it is the prerequisite for Maximise Conversion Value and Target ROAS. It is also where lazy setup does the most damage.
Every conversion valued at the same placeholder number makes a value-based strategy identical to a count-based one, so it looks like it is working while providing no advantage. A value that does not reflect margin — revenue rather than profit, say — will optimise confidently toward the wrong thing.
Choosing a value
- Use a real figure: what the order is worth to you, not what the product retails for.
- Where margins differ sharply, the value is what tells the system which orders to chase.
- If you cannot assign a meaningful value, use a count-based strategy. It is the honest choice, not a lesser one.
The order to fix things in
- Confirm conversion tracking works, with a real test conversion you watched arrive.
- Confirm it records the conversions that matter, not only the easy ones.
- Only then choose a conversion-based strategy.
- Only then add a target.
Every step skipped here is a step the bidding system will happily automate on your behalf, at speed, in the wrong direction.
Traps worth naming
- Target ROAS on a small account. The most common misapplication in the category. The strategy protects an average ratio by declining the searches it cannot predict, and a new campaign produces no data to predict from.
- Automation over broken tracking. The most expensive one. It fails quietly and it fails consistently.
- Reading a target as a forecast. It is a constraint, and the volume is the price of it.
- Changing the target weekly. A target is a decision. Adjusted constantly it is a knob, and the system never settles.
- Consolidating thin campaigns to “fix” learning, and discovering the good campaign is now funding the bad one.
- Placeholder conversion values. A value-based strategy on uniform fake values is a count-based strategy with extra steps.
- Judging a change over a week, against a system that learns on a longer clock than the person watching it.
- Treating a stated preference as an analysis. A list of “best practices” is not a diagnosis of your account.
This advice has a shelf life
Google's bidding landscape has changed repeatedly, towards automation, and it will change again. What this tool gives you is a way of deciding that survives those changes: a prerequisite check, a named trade, and a refusal to forecast.
The specific strategy names, their requirements and their availability are true of the documented platform as it stands. Check your own account, which is authoritative, over this page. If they disagree, your account is right and this page is out of date.
What is a fact and what is judgement here
Platform behaviour: which strategies exist, what each optimises for, which require conversion tracking, which require conversion values, and that a conversion history is required for the conversion-based strategies. Judgement: the order of the eligibility checks, the guidance on which strategy suits which amount of control, the history bands, and the advice to loosen a target before tightening it. The page labels which is which, and the tool never presents the second as the first.
What this tool will not do
- Tell you your account's real conversion history, because it never sees your account.
- Recommend a target CPA or target ROAS figure, because that is your margin, not its arithmetic.
- Predict clicks, conversions, cost or return. A strategy is a rule; a rule is not a forecast.
- Change anything in your account.
What happens to your answers
Short answer
Nothing leaves your browser. Five dropdowns describing your own account are not transmitted. Unlocking sends your name and email so the copy can be emailed to you.
There is no Ads account connection on this page. That is a deliberate limit: a bidding tool that reads your account is a bidding tool that can change it, and the interesting failure modes of automated bidding are not ones to build a tool around.
Details worth stating plainly
- No account data is requested, transmitted or stored, in either mode.
- No cookies are set by this tool and no third-party script is loaded.
- Unlocking stores a short-lived local token so you are not asked twice on the same device.
- The emailed copy is the one place this recommendation exists outside your machine. Ask and we will delete it.
Last reviewed 2026.
Bidding strategy questions, answered properly
Will Google hit my target CPA?
A target CPA is a constraint the bidding system works towards, not a promise about your account. Google uses it to decide where bidding more is worth the risk, so a target that is set too tight makes the system bid more cautiously and stop competing for the searches it thinks it cannot win. The result is closer to your target and less traffic. There is no setting of a target that makes a campaign's average cost equal to it, and the average over a short period can be well away from it in either direction.
How much conversion history does Smart Bidding need?
Google requires conversion tracking and a history of conversions before the conversion-based strategies will work properly, but it does not publish a single number of conversions that unlocks them. As a practical guide, the conversion-based strategies behave unreliably on very thin data and settle once there is enough volume to learn from. This tool treats the history bands as practical guidance and labels them as such, because presenting a rule Google does not publish would be exactly the kind of invented certainty this page exists to avoid.
Should I use Maximize Conversions or a target CPA?
Maximize Conversions spends the budget for the largest number of conversions it can find at the cost it finds them, and reports the average cost afterwards. A target CPA adds a constraint: the system will only bid aggressively where it expects to hit that cost, which usually means less total volume. Add a target when the average cost matters more than the count and you have a figure in mind. Leave it off when you would rather have the volume, or when you do not yet have a trustworthy cost figure to anchor to.
Is manual CPC still a valid choice?
Yes, and it is often the correct one. Manual CPC is the right answer when conversion tracking is missing or unreliable, because an automated strategy will confidently optimise toward whatever your tracking reports, and if that is wrong the error is amplified rather than caught. It is also the right answer when you are optimising for something conversion tracking cannot represent, such as a phone call that nobody has logged, or a new market where you are deliberately buying data. It is a worse answer when tracking is solid and volume is the priority, because you will leave bid management on the table.
Why did my conversions drop when I added a target?
That is the target doing its job rather than failing. A target makes the bidding system more selective, because it will not chase a conversion it predicts will cost more than the target. Fewer conversions at a better average cost is the intended trade, and the right response depends on your goal: if the average cost is what you needed, the target is working. If volume was the need, the target is too tight, and loosening it or removing it will return volume at a higher average cost. Neither outcome is a fault; they are the same lever set to different positions.
Sources and method
What comes from the documented platform
The set of bidding strategies and what each optimises for: Manual CPC, Maximise Clicks, Maximize Conversions, Target CPA, Maximize Conversion Value, Target ROAS, and Enhanced CPC as the older hybrid Google documents as being retired. That the conversion-based strategies require conversion tracking and a conversion history. That Maximize Conversion Value and Target ROAS additionally require conversion values. That Maximise Clicks does not require conversion data. That a target functions as a constraint on bidding behaviour rather than a set outcome.
That a conversion history is required for Smart Bidding, and that Google does not publish a specific conversion count that unlocks the conversion-based strategies.
What is judgement, and whose
Documented platform behaviour: the strategy names, what each optimises for, which require conversion tracking, which require conversion values, and the requirement for a conversion history.
Judgement, and marked as such on the page: the order of the eligibility checks, the guidance on which strategy fits which amount of control, the conversion-history bands, the advice to loosen a target before tightening it, and the framing of a target as a lever rather than a promise. The history bands in particular are practical guidance, not a published threshold, and the page says so where they appear.
Not asserted anywhere: any predicted clicks, conversions, cost per click or return; any recommended target figure; any Ad Rank; any claim about the specific contents of an account this page has not seen.
Last reviewed 2026. Google's bidding options change over time — your account is authoritative, and if it disagrees with this page, your account is right.