How to Control ROAS in Performance Max Without tROAS or tCPA
You can hold a Performance Max campaign to a ROAS goal without ever setting a tROAS or tCPA — by running it on Maximize Conversion Value with no target and using the daily budget as your throttle. It sounds backwards, but it solves the single biggest frustration with target bidding: the targets that are supposed to control efficiency quietly control your volume, and cap your ability to scale.
Why a target throttles you
A tROAS (or tCPA) doesn't just steer toward efficiency — it decides how much of your audience Google will pursue. To hit the number, the algorithm takes the path of least resistance: it concentrates spend on the handful of products, audiences, and searches that are already converting at or above your goal, and it drops nearly everything else. You get your ROAS, but at the cost of reach.
Set a high target on an account that isn't ready, and spend collapses — Google would rather under-spend than miss the goal. That's efficiency working exactly as designed, and scaling dying as a side effect.
So a target is a great tool for protecting a proven campaign's efficiency, and a terrible tool for growing one. The mistake is using it for both.
The alternative: budget as the throttle
Instead of asking Google for a return and letting it shrink to hit it, run the campaign with no target — Maximize Conversion Value (or Maximize Conversions) — so it spends the full budget and explores the whole opportunity, not just the safe wins. Then you control efficiency the old-fashioned way: with the daily budget.
Budget becomes your ROAS dial. At a given spend the campaign settles into an actual ROAS. If that's comfortably above your goal, you have room — raise the budget in steps. If it's below, pull the budget back. You're adjusting the throttle, not fighting the algorithm's volume controls.
How to monitor and adjust
- Give it the learning phase. A new campaign with no conversion history needs time (often 60–90 days of data) before its numbers mean anything. Don't judge or choke it in week one.
- Read the ROAS the budget produces. Let it run at a set daily budget and watch the ROAS it naturally returns.
- Step the budget, not the target. Above goal with stable performance? Increase budget ~10–20% and let it re-stabilize. Below goal? Decrease. Move in steps and give each change time to settle before the next adjustment.
Why this bites hardest when you're chasing new customers
The trap is sharpest when your goal is new customers. A target — whether it's a tCPA on Maximize Conversions or a tROAS on Maximize Conversion Value — pushes Google toward the cheapest conversions it can find, and the cheapest conversions are almost always people already close to buying: returning visitors, cart abandoners, and people searching your brand name. So the campaign quietly harvests demand you already had instead of going out to find buyers who've never heard of you. If growth means net-new customers, a target actively works against you. Turn it off, let the budget fund discovery, and where the platform offers it, switch on new-customer acquisition goals or value rules so the spend is aimed at fresh demand rather than the easy wins.
Feed it data it can actually use
The no-target approach only works if the campaign can measure value accurately. Get conversion tracking clean, send real conversion values (not just counts), and give it strong assets and a healthy product feed. The model can only scale what it can measure — garbage data plus no target is just fast spend.
An advanced twist: let Shopping and PMax share the work
A more advanced structure has a Standard Shopping campaign carries the majority of the budget as the primary driver of high-intent search, and a smaller PMax campaign runs alongside it to "bolt on" the fast, cheap, easy wins across the rest of Google's inventory. Because the two share traffic rather than feed one another, the blended cost-per-click drops and total efficient spend rises. It's an elegant way to scale without handing the whole account to a black box.
So when should you use a target?
Once a campaign has consistent conversion data and your goal shifts from growth to protecting margin, a tROAS is the right tool — apply it after the data exists, to hold a proven line. The rule of thumb: no target while you're scaling and learning; a target once you're defending. Using a target to scale is the thing to avoid.
Turn the method into a repeatable audit.
The Paid Digital Audit Kit reads any Google or Meta account through the funnel and includes the KPI Forecast Calculator to model what a budget change should return — so budget-as-throttle becomes a plan, not a guess.
Get the Paid Digital Audit Kit — $49Frequently asked questions
Can you control ROAS in Performance Max without a target ROAS?
Yes. You can steer Performance Max with budget, conversion values, audience signals, and campaign structure instead of a rigid target ROAS, which is useful when you are scaling or acquiring new customers.
Why avoid target ROAS when scaling?
A high target ROAS tells the algorithm to chase efficiency, which often caps volume and favors existing customers — counterproductive when the goal is growth and new-customer acquisition.
How do you push Performance Max toward new customers?
Use the new-customer acquisition goal, feed accurate conversion values, segment by margin, and control spend and structure rather than over-constraining with a target ROAS.