Target ROAS vs maximize conversion value
Start with maximize conversion value and no target. Add a target only when the campaign is producing conversions steadily and you know the return your margin needs. A target set before there is data is a guess that quietly caps delivery, and the campaign will underspend while you wonder why nothing is happening.
The two strategies are the same engine with a different instruction. Without a target, the bidding is told to buy as much revenue as the daily budget allows, so the budget is your only limit. With a target, the bidding is told to buy revenue at a stated ratio to spend, so the target becomes the limit and the budget often goes unspent. Which limit you want depends on whether your constraint today is cash or profitability.
Choose
Target ROAS
when
- The campaign already gathers conversions every week with no empty days
- You know your break even return after product cost, shipping and payment fees
- Spend is at the level you want and the job now is protecting margin
- You are scaling a proven product line and need the return to hold as budget rises
- A category has to clear a specific return before finance will approve more spend
Choose
Maximize conversion value
when
- The campaign is new and has no conversion history to bid from
- You want the full daily budget spent so learning happens quickly
- You are testing a new product range and need volume before judging it
- Seasonal demand is short and holding out for a return would mean missing the window
- A target ROAS campaign has been underspending its budget for two weeks
| Criterion | Target ROAS | Maximize conversion value |
|---|---|---|
| Cost | Spend sits below budget whenever the target is hard to hit | Budget is spent in full, so daily cost is predictable |
| Setup effort | Needs a defensible target from your own margin maths | Set a daily budget and let it run |
| Time to signal | Two to three weeks after each target change before results settle | Faster, because volume accumulates from the first day |
| Control | You control profitability, the platform controls how much sells | You control how much is spent, the platform controls the return |
| Best for | Established lines with known economics | New campaigns, new products, short seasonal pushes |
| What can go wrong | A target set too high starves the campaign and it stops serving | Revenue grows while margin quietly disappears into low value orders |
| How we measure it | Spend against budget first, then return over a full purchase window | Cost per order and average order value, checked against break even weekly |
When both apply
Sequence them rather than choosing once. Run maximize conversion value until the campaign has a few weeks of steady conversions, read the return it settles at, then set the first target slightly below that level so delivery is not cut off.
After that, move the target in small steps and wait for the learning to settle before judging the change. Large jumps produce a campaign that stops spending, which reads like a bidding failure but is only an instruction that cannot be met at the current price of demand.
Questions
Related
Your plan is free.
A person reads it. You keep the plan. No logins, no call first.