Google Ads bidding strategies: what each one is actually asking for.

Most bidding advice is a ranked list: use this strategy for that goal. It skips the thing that actually breaks accounts.

Every bid strategy in Google Ads does exactly what its name says. None of them malfunction. When bidding goes wrong, and we audit a lot of accounts where it has, the campaign was faithfully following a clear instruction that nobody meant to give.

So the useful question is not which strategy performs best. It is what each one is asking for, and whether that is what you actually want to buy.

The distinction that matters more than the list

Google defines Smart Bidding as the strategies that use auction-time bidding to optimize for conversions or conversion value. There are four: Target CPA, Target ROAS, Maximize conversions and Maximize conversion value.

Maximize clicks and Target impression share are automated. They adjust bids for you. They are not Smart Bidding, and they do no auction-time bidding toward your conversions at all.

That gap is where the most expensive mistakes live. A campaign can be fully automated, show a conversion goal in the interface, report on leads, and still be buying nothing but traffic. We wrote about that specific failure in your campaign says it's optimizing for leads.

Two housekeeping notes on the current interface, because both changed recently and both trip people up:

• Enhanced CPC is gone. Google removed it for Search and Display from the week of March 31, 2025. Campaigns that were not migrated fell back to Manual CPC. If you have inherited an account nobody has touched since, that is what it is running now, whatever the history says.
• Target CPA and Target ROAS look standalone again. Since June 2026 they appear as their own strategies in the interface rather than as targets inside Maximize conversions and Maximize conversion value. Google is explicit that this is a labeling change only. "Maximize conversions with a Target CPA" and "Target CPA" are the same mechanism.

What each strategy is actually asking for

• Maximize clicks: "Here is my daily budget. Get me as many clicks as you can." You are controlling quantity, not quality.
• Manual CPC: "Do not give me clicks above this price." A maximum CPC is also a quality cap, because it excludes you from the auctions where the better buyers are.
• Target impression share: "Put me on the page." Genuinely useful for defending brand terms. It says nothing whatsoever about outcomes.
• Maximize conversions: "Here is what a conversion is. Go and get as many as you can within budget."
• Maximize conversions with a Target CPA: "...at roughly this average cost."
• Maximize conversion value: "Get me the most total value you can."
• Maximize conversion value with a Target ROAS: "...at roughly this return."

Read down that list and something becomes obvious. The first three describe what you are willing to pay. The last four require you to have already defined what you are trying to buy.

That is the real work, and it happens before bidding. Choosing a conversion strategy forces the question most accounts skip: what actually counts? A form fill, a booked meeting, a qualified opportunity in the CRM? Pick the stage that matters commercially and send that back to Google as the conversion goal. The bid strategy is downstream of that decision.

Where we start: Maximize conversions, from day one

The standard advice for a new campaign is to start on Maximize clicks, build conversion history, then switch. Google's own help community suggests roughly 15 conversions in 30 days as a baseline before moving to conversion bidding.

We do not do it that way, and we have yet to regret it.

The plan assumes the data you build on Maximize clicks is the data a conversion strategy needs. It is not, quite. Maximize clicks buys traffic selected to produce clicks cheaply, so the conversions you record come from a traffic mix chosen for cheapness and volume rather than intent. You are not building a clean picture of who converts. You are building a picture of who converts among the people cheap clicks brought you.

Meanwhile the formative early window, when the campaign is forming its habits, goes to optimizing toward an outcome you do not care about. Running Maximize clicks first does not build conversion data. It builds click data. The fuller argument is in why we start every campaign on Maximize conversions.

The honest counter-argument. With very little conversion history a conversion strategy has less to model on, so early delivery can be slow, expensive and volatile. If your budget is small enough that two bad weeks matter, that volatility is a real cost. Some good practitioners would rather buy cheap traffic, gather signal and switch. That is defensible, and the volume threshold is a judgment call rather than a rule.

Our answer is that a strategy pointed at the right outcome and learning slowly beats one pointed at the wrong outcome and performing beautifully. But expect a slower, bumpier start, and budget for the learning period rather than panicking in week one.

A target is an instruction, not a guardrail

This is the most common and most expensive misunderstanding we find in B2B accounts.

Founders set a strict target CPA to stay disciplined. It sounds responsible. It is the opposite. A target CPA is not a budget control. It is an instruction to Smart Bidding, and the instruction a low one gives is blunt: find me the cheapest conversions you can.

Smart Bidding obeys. It goes and finds the cheapest conversions on the table, which means small buyers, low intent, the people who convert without costing much. For a while the dashboard looks excellent. Then it stalls, because the cheap traffic runs out and the higher-value buyers you actually want sit above your ceiling, where the algorithm was never allowed to bid for them. Google has a name for that state, "Limited by bid strategy", and its own guidance is to raise the target.

In B2B the distance is enormous. A $50 target trains Google to optimize for the $50 lead, not the account that closes at $40,000. You end up with a full pipeline of the wrong people and a cost per lead that looks fantastic, right up until sales tells you none of them close. We made this case at length in a low target CPA isn't saving you money.

What changes on August 17, 2026

Google is changing how target-based bid strategies behave, and a large number of accounts will feel it.

From August 17, 2026, target-based strategies will perform more consistently toward the target you set, including when a campaign is limited by budget. Today it is common for a budget-limited campaign to quietly overachieve, hitting a $5 CPA against a $10 target. After the change, that campaign moves toward the $10 you actually asked for.

It applies to Search, Shopping, Performance Max, Demand Gen, Display, Hotel and Travel campaigns. App campaigns, Video reach and Video view campaigns are excluded. Google's Bid Target Adjustment Tool has been available since July 6, 2026, and it will suggest lowering targets to match recent performance.

Our take, and it is not the obvious one. The tool's default suggestion is to lower your targets so the numbers stay roughly where they have been. Before accepting that, work out why the campaign was overachieving.

• If it was overachieving because it found genuinely good buyers cheaply, lower the target and lock the gain in. That is a real win and you should keep it.
• If it was overachieving because the target was set once, long ago, and the campaign has been buying the cheapest conversions available ever since, then lowering it re-anchors you to exactly the trap described above. You would be using a platform change to make a bad instruction permanent.

The honest check is not the CPA column. It is what happened to those conversions downstream. Pull the last two quarters of leads out of the CRM and look at how they closed. If the cheap ones did not close, the overachievement was never a win, and the right move is to leave the target alone or raise it and let the campaign compete for better buyers.

Do this before August 17 rather than after. Afterwards you will be diagnosing a performance shift and a strategy question at the same time, and the two are hard to separate once they are tangled.

The 60-second audit

Worth running on any account you have inherited.

• Open the campaign, go to Settings, read the bid strategy. Not the conversion goal column, the strategy itself. They are two different settings and nothing forces them to agree.
• If it says Maximize clicks or Manual CPC, the campaign is buying traffic whatever the goal column claims.
• If it carries a target, ask when that number was last set, by whom, and against what assumption.
• Check the shape of the data. Healthy clicks, fine CTR, conversions flat or arriving from obviously wrong-fit people.

That last pattern gets misdiagnosed constantly. It looks like a landing page problem, so teams rewrite the page, change the form, test new headlines. The page was never the issue. The campaign was never told to find buyers.

Before you change anything

Two prerequisites, both routinely skipped.

Conversion tracking has to be trustworthy first. Pointing a conversion strategy at a badly defined conversion action just makes the wrong thing happen faster.

And every strategy change restarts the learning period. The campaign recalibrates, performance gets noisy, and the numbers during that window are not a verdict on the decision. Changing strategy again mid-learning is how accounts end up permanently unstable. Make the change, then leave it alone long enough to read it.

The end state: optimize to value, not volume

For most B2B accounts the destination is not a better cost target. It is not optimizing to cost at all.

Move to Maximize conversion value and feed it real values. Pass back-end value, pipeline or closed-won, into the account so that "cheap" stops meaning "good." The algorithm is only ever as smart as the goal you hand it.

You are not buying conversions. You are buying customers.