Target ROAS in Google Ads: When and How to Use It

Target ROAS in Google Ads is the most-asked-about topic Chris gets questions on. Every day, someone asks when to set it, what number to use, and why their campaign fell apart after they turned it on.

The short answer: target ROAS is a tool for campaigns that already have a lot of data and are out of the learning phase. Set it too early and you will choke your campaign. Set it too high and you will cause the same problem. Get it right and it gives your account more stability while you scale.

Here is how it actually works.

Never set target ROAS at the start

This is the first and most important rule. When a campaign is new, the algorithm has no idea where conversions come from. It has not seen enough purchases to know which products, audiences, or times of day convert well for your store.

If you set a target ROAS before that data exists, the algorithm tries to hit a number it cannot see a path to. So what happens? It stops spending. Your daily budget barely gets used. You might get a few clicks and almost no sales. The campaign stalls out.

The beginning of a campaign is about one thing: gathering data. You do not touch the target ROAS. You let it run. You let it learn. Patience here is not optional - it is the whole strategy.

Why 30 to 50 conversions is not enough

Google recommends you have at least 30 to 50 conversions in the last 30 days before setting a target ROAS. That sounds reasonable. It is also misleading.

Here is the problem. That recommendation looks at the account as a whole. It does not look at what those conversions are actually spread across.

Think about a shoe store with over 600 product variants. If the account has 50 purchases in a month, that sounds OK on the surface. But when you break it down by product, you might have one or two purchases per variant, sometimes less. Some variants have never converted at all.

The algorithm needs to learn at the product level, not just the account level. It needs to know which specific products convert, which sizes or colours sell, which queries drive actual sales for each item. If the data is spread thinly across hundreds of variants, the algorithm is still essentially guessing.

Chris makes this point directly:

When you have that many product variants and you think about 30 to 50 conversions - if you actually look at it on a product level, how many conversions do I have per product? It's not a lot of data.

The real amount of data you need before setting target ROAS is much higher than Google's baseline. There is no fixed number because it depends on how many products you have and how conversions are distributed across them. The key question to ask is: do I have enough purchases per product for the algorithm to understand what is working?

This is the same reason some larger stores with high revenue never quite reach a consistent level of performance. If they have too many products and conversions come from everywhere, they never fully exit the learning phase.

Check out the full guide to scaling Google Ads for ecommerce in 2026 for more on how the learning phase fits into a scaling strategy.

When you actually use it - stability while scaling

Once you are out of the learning phase and have solid conversion data, there are two situations where target ROAS earns its place.

The first is when you are actively scaling and your ROAS is getting noisy.

When you push more budget into a campaign and things are going well, revenue climbs. But ROAS can bounce around a lot. Some days it is great. Some days it barely breaks even. That inconsistency is not necessarily a problem - it is just what aggressive scaling looks like. But it can be hard to manage.

This is where target ROAS helps. It does not change your goal. It gives the algorithm a floor to work around. Instead of wild swings, you get more stability.

Here is how Chris approaches it. Say the overall store goal is a 300% ROAS. In the ad account, because no tracking setup captures every conversion, the visible ROAS might look closer to 280%. If the account is set to target 280%, it will push hard to hit that and might sacrifice volume to get there.

Instead, Chris sets it lower - something like 250% while scaling. That gives the campaign room to spend freely, test more products and placements, and scale without the algorithm tightening too hard. And often, when the settings are right and there are winning products, the actual ROAS comes in above the target anyway.

The key mindset shift: do not set target ROAS because you want to squeeze the most profit out of the account right now. Set it to bring stability to a campaign that is already working. If everything is going well without it, there is no reason to add it. Let it run.

You can read more about how ROAS fits into your profit picture in the ROAS vs POAS guide for ecommerce.

Using target ROAS to push efficiency on a flat account

The second scenario is different. The account is established. Revenue is consistent but not growing. Margins are thin or just barely acceptable. You want to push profitability higher without cutting too much volume.

Here, you can use target ROAS to send the algorithm a signal.

The approach: set it slightly above your current average. Not dramatically higher - just a step up. Say your account is sitting at around 190% ROAS and you want to push toward 200%. You set the target there. Then you wait. You watch revenue over the next one to two weeks. You check whether conversion volume drops significantly or stays steady.

If the algorithm can get there, it will. The account has enough data to understand what it needs to do. Revenue might dip slightly as it optimises for efficiency over volume - that is expected. The trade-off is intentional.

If it cannot get there, you will know because conversions will fall fast and spend will start to drop. That means the target you set is too ambitious for where the account is right now. Pull it back. Find the level the algorithm can actually hit.

The key warning: never jump too far above where you are. If you are at 190% and you set it to 300%, you get the same problem as setting it too early on a new campaign. The algorithm throttles. Spend drops. And now your stable, profitable account is in trouble.

The downward spiral to avoid

There is a trap that Chris sees again and again. Someone sets target ROAS too high. The campaign throttles. Spend drops. Performance suffers. So they remove the target ROAS to give it freedom again.

But removing it after a throttle does not instantly fix things. The campaign now needs time to readjust. It starts spending broadly again, which can push ROAS down even further while it relearns. You end up in a worse position than where you started.

The lesson: prevention is much easier than recovery. Set it conservatively. Give it room. Watch what happens. Nudge it slowly.

If you ever do get stuck in that spiral, give the campaign real time to recover - days, not hours. Do not make changes based on one bad day. Emotional adjustments based on short-term data make things worse. The algorithm needs stability to perform, and constant changes remove that stability.

The right mental model

Target ROAS is not a magic tool. It does not override the fundamentals. You still need great products. You still need a strong store with solid conversion rates. You still need good tracking.

What it does is help the algorithm stay within a range you are comfortable with - once it already has the data to work with. Think of it as a guardrail for a campaign that already works. It keeps results steady once the data is there.

Watch the full breakdown here:

If you want to see how a full campaign structure fits together and where target ROAS sits in a real scaling strategy, see how we work with clients. Or if you want proof it works at scale, have a look at the results.

Frequently asked questions

When should I set target ROAS in Google Ads?

Not at the start. You need to be out of the learning phase first and have a lot of conversion data. The algorithm can only hit a ROAS target reliably when it already has enough purchase history to know where those conversions come from. Set it too early and your spend drops off and your campaign stalls.

How many conversions do I need before setting target ROAS?

Google recommends 30 to 50 conversions in the last 30 days, but that is not nearly enough in practice. What matters is how conversions are spread across your products. If you have hundreds of product variants and only 50 purchases total, most products have almost no data. You need enough data at the product level, not just the account level.

What happens if I set target ROAS too early?

Your campaign starts to throttle. It cannot spend properly because the algorithm does not know how to hit the target yet. Spend drops, performance suffers, and you get stuck in a downward spiral. Removing the target ROAS after that point causes its own problems because the campaign then needs time to readjust.

How should I use target ROAS when scaling?

Set it lower than your actual target to give the campaign freedom. If you want a 300% ROAS overall but only see 280% in the ad account, Chris suggests setting it to something like 250% while scaling. That adds stability without choking volume. The campaign often over-delivers beyond the target on good days.

Can I use target ROAS to push profitability on a flat account?

Yes, but carefully. If an account is running consistently but not growing, you can nudge the target ROAS slightly higher to signal to the algorithm that you want more efficient conversions. Set it just above your current average, wait one to two weeks, watch what happens to revenue, then nudge again. Never jump to a number far above your current performance or you will throttle the campaign.