Google Ads Bidding Strategies: Maximize Conversions, Target ROAS and When to Use Each

Google Ads · October 26, 2026

Bidding strategy is the setting in Google Ads that most often gets changed for the wrong reason. Someone sees a bad week, switches strategies, and the account spends the next month relearning. This guide covers how we think about the main options, what data they need, and how to tell whether a change actually helped.

Manual versus automated bidding

With manual bidding you set the maximum cost per click yourself, by keyword or ad group. It gives you control and a clear cause and effect: raise a bid, see what happens. It also takes constant attention, and it cannot react to the signals Google sees at auction time, such as device, location, time of day and the user's recent behavior.

Automated bidding hands those decisions to Google's models. You tell the system what to optimize for, and it sets a bid for each individual auction. For most accounts with a working conversion setup, automated strategies outperform hand-tuned bids on anything but the smallest budgets. The catch is that they are only as good as the signal you feed them.

That is why we treat conversion tracking as step zero. Our guide on Google Ads conversion tracking covers how to audit that before you touch bidding.

Manual bidding still has a place for brand new accounts with no conversion history, and for very low volume campaigns where the data will never be enough for a model to learn from.

The main automated strategies

Google's names and options change over time, so check the current options in Google Ads before committing to a plan. The core ideas have been stable for years.

Maximize Conversions tries to get the most conversions it can for your budget. It has no built-in view of what a conversion is worth, so it treats a five dollar lead and a five hundred dollar order the same.

Target CPA (cost per action) tells the system the average cost you are willing to pay per conversion and asks it to find as many as possible at that price. It suits lead generation, where each conversion is worth roughly the same.

Maximize Conversion Value is the value-based sibling of Maximize Conversions. It tries to get the highest total value within budget.

Target ROAS (return on ad spend) asks the system to hit an average return, for example three dollars of revenue for every dollar spent. It suits ecommerce and any account where conversions carry different values.

Conversion value versus conversion count

This is the decision that matters most. If every conversion is worth about the same, a count-based strategy such as Maximize Conversions or Target CPA is fine. If order values vary a lot, a count-based strategy will happily chase cheap, low value orders because they are easy to find.

For a store selling products from fifteen dollars to four hundred dollars, a value-based strategy is almost always the right call, provided the values you pass back to Google are accurate. That means the revenue figure should match what your store reports, and ideally you have decided whether to pass revenue or profit. Passing profit or margin-adjusted value is more work, but it keeps the system from favoring high revenue products that barely make money.

Data thresholds

Automated strategies learn from your conversion history, so they need enough of it. Google publishes guidance on how many conversions a campaign should have recorded before switching to Target CPA or Target ROAS, and that guidance has changed over the years, so check Google's current recommendation rather than trusting a number from a blog post, including this one.

What we can say in general terms:

• A campaign with only a handful of conversions a month gives the model very little to learn from, and results will swing wildly.
• If volume is thin, consolidate campaigns so the data pools in one place.
• Start with Maximize Conversions or Maximize Conversion Value, and add a target later once results are stable.

Setting a target from margin

A target should come from your economics, not from a number that feels reasonable. Here is the process we use.

1. Work out the gross margin on the product or product group, after cost of goods, shipping and payment fees.
2. The break-even ROAS is one divided by that margin. A 40 percent margin means a break-even ROAS of 2.5, because you need two and a half dollars of revenue to cover each dollar of ad spend.
3. Set your target above break-even by whatever cushion you need for overhead and profit. If the goal is growth, you may accept a lower return for a period. If the goal is cash, push it higher.
4. For lead generation, work backward from what a closed customer is worth and what share of leads close.

The same logic underpins Amazon work, and we cover the marketplace version in our guide to break-even ACoS and target ACoS.

A common mistake is an aggressive target on day one, which makes the system throttle spend. Start close to recent actual performance, then tighten gradually.

Learning periods and not changing things too often

After you change a bidding strategy or its target, Google's system enters a learning period while it recalibrates. Performance is often noisy during that time. The exact length varies by account, so look at the status shown in Google Ads rather than assuming a number of days.

Our rules of thumb:

• Make one meaningful change at a time. If you change the strategy, the budget and the ads on the same day, you will not know which one mattered.
• Wait until the learning status clears before judging.
• Adjust targets in modest steps rather than large jumps.
• Avoid reacting to single days, since conversion lag makes recent days look worse than they will end up.

Portfolio strategies

A portfolio bid strategy applies one strategy across several campaigns, so the system pools their data and optimizes toward a shared goal. It helps when you have several small campaigns that each have too little data on their own, but similar economics.

It is a poor fit when campaigns have different margins or goals. Branded and non-branded campaigns should usually have separate targets, because brand traffic converts far more cheaply and would mask weak performance elsewhere.

Guardrails

Automation needs boundaries, because it will spend whatever it is allowed to chase the goal you set.

• Budgets. Set a daily budget you are comfortable with, and watch for campaigns that are limited by budget while hitting target, which is a sign you could scale.
• Negative keywords. Automated bidding does not know which queries are irrelevant to your business. Keep the exclusion list current, as covered in our guide to negative keywords.

How to judge a strategy change

Before changing anything, write down the baseline: cost per conversion or ROAS, conversion volume, spend, and revenue over a comparable period. Then decide what success looks like and how long you will wait.

When you evaluate, compare like with like. Account for seasonality and promotions, and look at business outcomes, not only platform numbers. If revenue in your store did not move while Google Ads reported more conversions, something is off. Where possible, use Google's built-in experiments to run the new strategy against the old one on a split of traffic, which is far cleaner than comparing before and after.

Want help with your account?

We manage Google Ads alongside Amazon PPC and Shopify work, and we set bidding targets from real margin data rather than guesses. If you would like a second set of eyes on your current strategy, we are happy to take a look.

Talk to us about Google Ads.