North American Media Experts
Maximize Conversions, Target CPA, or Target ROAS? A step-by-step framework to pick the right Google Ads bidding strategy in 2026 and set targets that work.
9 min read
If you have ever stared at the Google Ads bidding menu and wondered whether Maximize Conversions, Target CPA, or Target ROAS is the right call, you are not alone. The bid strategy you choose quietly decides how every dollar in your account gets spent, which auctions you win, and whether your cost per lead climbs or falls next month. Pick wrong and you can burn through budget chasing cheap clicks that never convert. Pick right and Google's machine learning does the heavy lifting for you, evaluating thousands of contextual signals per auction that no human could weigh in real time.
This guide gives you a practical, step-by-step Google Ads bidding strategy decision framework for 2026. By the end you will be able to look at any campaign and know exactly which Smart Bidding strategy to use, what target to set, how long to leave it alone, and when to switch. No theory for theory's sake. Just the decision logic our team uses on real accounts.
A bid strategy is not a setting you flip once and forget. It is the objective function Google optimizes against. Maximize Conversions tells the algorithm to spend the full budget and bring back as many conversions as possible. Target CPA adds a cost constraint: get conversions, but try to hold the average cost per acquisition near a number you choose. Target ROAS changes the goal entirely, optimizing for conversion value rather than conversion count, which matters enormously for ecommerce where a $400 order and a $40 order are not equal.
The mistake most advertisers make is choosing a strategy based on what they want the outcome to be, rather than what their data can actually support. Smart Bidding is a prediction engine. Starve it of data and it guesses. Feed it clean conversion signals at sufficient volume and it consistently outperforms manual bidding. So the framework below starts not with the strategy menu, but with your goal and your data.
Every bid strategy maps to one of three commercial goals. Be honest about which one you are actually optimizing for, because picking the wrong goal is how accounts end up with a great-looking CPA and a shrinking bank balance.
Volume at a fixed budget. You have a set daily spend and you want the most conversions that money can buy, with no hard cost ceiling. This is a Maximize Conversions goal.
Efficiency at a target cost. You know what a lead or sale is worth to your business and you need the average cost to stay near that figure. This is a Target CPA goal.
Profitability by value. Your conversions have different revenue values and you care about return, not raw count. This is a Target ROAS goal. If you cannot pass accurate revenue data back to Google, you are not ready for this one yet.
Write your goal in one sentence before you open the campaign settings. If you cannot, that is a sign your paid search measurement foundation needs work first.
This is the single most important gate in the framework, and it is the one advertisers skip most often. Smart Bidding needs a baseline of conversions to learn from. The widely cited thresholds, which match what we see in practice, are roughly:
Pull your last 30 days of conversions at the campaign level and compare against these numbers. If a campaign is sitting at eight conversions a month, no clever target setting will save it. The honest move is to consolidate thin campaigns, fix tracking, or run Maximize Conversions until volume builds. Running a quick Google Ads account audit first will surface broken conversion actions and duplicate pixels that quietly poison the data Smart Bidding learns from.
With your goal and your conversion volume in hand, the choice becomes mechanical. Here is the decision logic strategy by strategy.
Choose Maximize Conversions when a campaign is new, conversion data is thin, or you simply want to extract the most volume from a fixed budget. It is the natural starting line. The trade-off to understand: Maximize Conversions will spend your entire daily budget every day. It has no cost ceiling, so if your goal is efficiency rather than raw volume, you will eventually want to graduate to Target CPA. Many experienced managers deliberately run Maximize Conversions first precisely to gather the 30 to 50 conversions Target CPA needs, then switch.
Choose Target CPA once a campaign clears roughly 30 conversions a month and you have a defensible cost-per-acquisition number tied to your unit economics. This is the best long-term home for most lead generation campaigns. The constraint it adds is the whole point: instead of spending everything, it pulls back in expensive auctions and leans in where conversions look likely, holding your average cost near target. If your CPA target is realistic, this is where stable, scalable performance lives.
Choose Target ROAS when conversions carry different values and you can pass accurate revenue back to Google. It is the default for ecommerce and any business where a $1,200 deal and a $120 deal should not be bid on identically. The prerequisite is non-negotiable: clean value tracking plus enough volume, typically 50 or more value-carrying conversions in 30 days. Without trustworthy revenue data, Target ROAS optimizes toward a number it cannot see, and results wander.
If you are deciding how this fits alongside automated campaign types like Performance Max, our Performance Max structure framework walks through how value-based bidding behaves inside those campaigns.
A correct strategy with a bad target still fails. The most common error is setting an aggressive target on day one that chokes off traffic before the algorithm can learn.
For Target CPA, set your initial target at or slightly above your actual average CPA from the last 30 days. A starting point about 10 to 20 percent higher than recent reality is safe. This gives the algorithm room to find conversions without immediately strangling delivery. Once it stabilizes, tighten the target in small steps.
For Target ROAS, anchor to the return you have genuinely achieved recently, not the return you wish you had. Setting tROAS at 600 percent when the account has historically delivered 300 percent simply tells Google to stop bidding. Start near your trailing performance and ratchet up gradually.
Whatever you set, change it in small increments. Which leads to the most misunderstood part of bidding.
When you launch or significantly change a Smart Bidding strategy, the campaign enters a learning period while Google recalibrates. In 2026 this typically runs two to six weeks, depending on volume. Campaigns generating 30 or more conversions a week exit faster; low-volume campaigns can take the full six weeks or longer.
During learning, performance is noisy and not representative. The discipline is to leave the campaign alone. Reacting to a bad three-day stretch by yanking the target is the fastest way to never escape learning at all, because several common actions reset the clock:
Plan changes in advance so you are not resetting learning every week. If your paid search budget is tight, give the algorithm a stable runway rather than starving and feeding it in cycles.
Smart Bidding removes most manual levers, but it leaves you a small, powerful toolkit. Knowing which tool to deploy for which purpose is what separates a managed account from one left on autopilot.
Seasonality adjustments. When you have a short, predictable spike, such as a 1 to 7 day promotion, tell the algorithm in advance to expect a temporary jump in conversion rate so it does not under-bid through your best window. The key nuance most people get wrong: size the adjustment to how much CPCs need to rise to capture the extra demand, not how much conversion rate changes. If conversion rates historically double during a sale but you only need CPCs about 25 to 30 percent higher to win the incremental impressions, a +25 to +30 percent adjustment is far more accurate than +100 percent. Use these for short events only; beyond about 14 days they lose their value.
Data exclusions. This is an emergency tool, not a routine one. If your conversion tracking broke for a defined window, say a tag was accidentally removed during a site update, a pixel double-fired, or a server outage blocked form submissions, exclude that window so Smart Bidding does not learn from corrupted data. Never use data exclusions to paper over normal performance dips.
Portfolio bid strategies. When several campaigns share the same goal and CPA or ROAS target, a portfolio strategy lets them pool conversion data and optimize together. This is especially useful for grouping lower-volume campaigns that individually struggle to clear the learning threshold but collectively have enough signal.
The healthiest accounts change bid strategies rarely and deliberately. Here is the cadence we recommend.
Switch from Maximize Conversions to Target CPA once a campaign has accumulated 30 to 50 conversions and you want to introduce cost control. This is the most common and most valuable graduation in the framework.
Switch from Target CPA to Target ROAS only when you have reliable conversion values and your business goal shifts from cost-per-lead to return. Do not make this jump on a campaign that lacks clean value data.
Leave it alone when a campaign is inside its learning period, when you are reacting to a few days of noise, or when the underlying problem is actually creative or landing page quality rather than bidding. No bid strategy can fix weak ads or a page that does not convert. If results are disappointing, audit the fundamentals before blaming the algorithm.
Choosing a Google Ads bidding strategy in 2026 comes down to a short, repeatable sequence: define the goal in one sentence, confirm you have the conversion volume the strategy requires, match Maximize Conversions, Target CPA, or Target ROAS to that situation, set a realistic target 10 to 20 percent off recent reality, respect the two-to-six-week learning period, and use seasonality adjustments, data exclusions, and portfolio strategies as precision tools rather than blunt instruments. Switch deliberately, document your changes, and resist the urge to tinker mid-learning.
Get this right and Smart Bidding becomes a genuine force multiplier instead of a black box you are nervous about. Get it wrong and you will keep paying for the algorithm's confusion. The good news is that the decision is far more systematic than the bidding menu makes it look.
If you would rather have an experienced team build the structure, set the targets, and manage the learning periods for you, our paid search management and broader digital advertising services are built around exactly this kind of disciplined, data-led optimization. Request a free quote and we will review your current bidding setup and show you where the easy wins are.