Ask ten media buyers what a good ROAS is and most of them will say 4x.

It's a nice round number. It's also meaningless until you know one thing about the business.

Its margin.

The short version

A good ROAS is one that sits above your break-even ROAS by enough to hit the profit you want.

ROAS = revenue from ads ÷ ad spend

Break-even ROAS = 1 ÷ contribution margin

Your contribution margin is what's left of each order after the costs that come with it: product, shipping, payment fees. If 40% of each order is left, your break-even ROAS is 1 ÷ 0.40 = 2.5x.

At 3x, you're making money. At 2x, you're paying to sell. Same ad account, same ROAS, completely different verdict depending on the margin. You can work yours out in a minute with the break-even ROAS calculator.

The three ROAS lines

Every account I run has three numbers drawn on the wall. I call them the three ROAS lines.

1. The floor. Your break-even ROAS. Below it, every sale loses money.

2. The target. The ROAS that leaves the profit you actually want.

Target ROAS = 1 ÷ (contribution margin − profit margin you want)

With a 40% contribution margin and a goal of keeping 15% of revenue as profit, that's 1 ÷ 0.25 = 4x.

3. The ceiling. A ROAS so far above target that you're almost certainly spending too little. You're only reaching people who were going to buy anyway.

A good ROAS lives between the target and the ceiling. Not as high as possible.

What the benchmarks say

If you want a reference point, the most useful one I know is Triple Whale's Meta benchmarks. They cover more than 40,000 ecommerce brands from August 2025 to July 2026.

The median Meta ROAS in that data: 1.88x. A year earlier it was 1.86x. Median order value was $73.36 and median cost per purchase was $38.99.

Read that again. Half of those brands get less than $1.88 back for every dollar on Meta, by Meta's own count.

For a brand with a 40% margin, that's below break-even on the first order. Those brands are betting on repeat purchases, or on sales Meta helped with but didn't get credit for.

The median isn't a goal. It's proof that most accounts don't make money on the first order alone.

So know which game you're playing. If customers come back and buy again, a first-order ROAS below break-even can be a smart bet. If they don't, it's a slow leak. I break down where the other numbers come from in how much Facebook ads cost.

Why a high ROAS can be bad news

ROAS almost always falls as spend rises. The first dollars reach the easiest buyers. Every extra dollar reaches someone a little less ready.

That's why a 10x ROAS on $50 a day usually means you're only reaching people who already know you. It looks brilliant. It's tiny.

Here's the math that matters:

Profit from ads = ad spend × (ROAS × contribution margin − 1)

Take a brand with a 40% margin.

  • $1,000 a day at 5x ROAS: $1,000 × (2.0 − 1) = $1,000 profit.
  • $3,000 a day at 3.5x ROAS: $3,000 × (1.4 − 1) = $1,200 profit.

The second account has a worse ROAS. It also makes more money and wins three times the customers.

You can't bank a ratio. You bank profit dollars.

A good ROAS depends on the stage

The right target also changes with what the business is trying to do right now.

Launching. You don't know yet which ads, offers and audiences work. Running close to break-even is fine for a while. You're buying information, and the first customers.

Growing. You know what works and you want more of it. Aim for the target line and push spend up until ROAS starts sliding toward it.

Harvesting. Growth matters less than cash this quarter. Pull spend back to the most efficient campaigns. ROAS goes up, revenue goes down, and that's the point.

None of these is wrong. Problems start when nobody decided which one the business is in, so the ROAS target drifts week to week with whoever is most worried.

Pick the stage first. The ROAS target follows from it.

What moves your number

Who is counting. Meta's ROAS is Meta's opinion of what it caused. Check it against total revenue with MER, which no platform can inflate.

Prospecting vs retargeting. Retargeting always shows a higher ROAS because those people were already close to buying. Never compare the two directly.

First order vs lifetime. If customers reorder, the first-order ROAS undersells the ad. Only count repeat revenue you've actually seen, not what you hope for.

Discounts and returns. A sale that gets refunded still shows up as revenue in Ads Manager.

How to raise ROAS without strangling growth

The short list is below. I go lever by lever, with a worked example, in how to increase ROAS on Meta.

1. Fix the margin side. Bundles, better pricing and higher order values lower your break-even. That's free ROAS.

2. Make better ads. Creative moves ROAS more than any setting. Start with hook rate, then feed the account new concepts every week.

3. Fix the page. If the landing page converts badly, no ad can save it.

4. Cut losers faster. Money spent on dead ads drags the whole account down. My kill rules are built for this.

5. Fix your tracking. If Meta can't see your sales, it can't find more buyers like them. The Conversions API closes most of that gap.

Pool rule 006Screenshots aren't results. P&L or it didn't happen.

Quick answers

Is 2x ROAS good? Only if your contribution margin is above 50%. Below that, 2x loses money on every order.

Is 4x ROAS good? For most ecommerce brands with margins above about 30%, yes. At a 25% margin, 4x is only break-even.

What is the average ROAS on Facebook ads? The median across 40,000+ ecommerce brands was 1.88x from August 2025 to July 2026, according to Triple Whale.

What's the difference between ROAS and ROI? ROAS is revenue divided by ad spend. ROI is profit divided by total cost. A campaign can have a great ROAS and a terrible ROI.

Is a good ROAS different on Google Ads? The math is the same. Search ROAS often looks higher because people are already looking for you, especially on brand searches.

The one-line version

A good ROAS is your break-even plus the profit you want. Anyone else's number is about their business, not yours.