If you type in “what is a good ROAS for Google Ads” into a search bar, you’ll usually find a benchmark or rule of thumb. It can be a useful reference point, but it rarely tells you what a good ROAS looks like for your specific business.

According to Optmyzr’s Q2 2026 Benchmark Report, the median ecommerce advertiser posted a 557.17% ROAS this quarter, drawn from more than 20,000 managed Google Ads accounts. If you treat that as your target, you’ll either chase a return your margins can’t support, or leave money on the table because your business could profitably spend at half that ratio. A benchmark only tells you where the market sits, but it doesn’t tell you where your business should sit.

Source: Optmyzr Q2 2026 Benchmark Report. Absolute figures are Q2 2026 medians of per-account values; ROAS reflects e-commerce accounts only.

Download the report here.

The number that actually moved this quarter

The broader Q2 picture was fairly stable. CTR rose 8.62% year-over-year, while CVR increased 3.59%. ROAS, however, moved in the opposite direction.

Median ROAS fell 3.75% quarter-over-quarter, breaking a run of three consecutive quarterly gains of 1.44%, 1.76%, and 1.86%. At the same time, CPC was up 5.56% year-over-year.

Frederick Vallaeys, CEO and co-founder of Optmyzr, connects those trends in the report to changes in the search results themselves.

“As AI Overviews take up more space, paid inventory becomes tighter. Advertisers may end up paying more for each click even as the ads that do appear earn stronger engagement.”

That is why the 3.75% decline in ROAS is more useful than treating the 557.17% median as a target. The median tells you where advertisers landed in Q2. But the quarter-over-quarter change tells you how that performance is moving.

Why “good” changes depending on who’s asking

Account size is one reason a “good” ROAS can look very different from one advertiser to another. In Q2 2026, Mid-Market accounts had the highest median ROAS at 674.87%, followed by Enterprise at 570.08% and SMB at 511.32%. The differences by industry were even larger. Sports reached a median ROAS of 753.64% and Style & Fashion 711.70%, while Real Estate came in much lower at 185.66%. At the same time, Real Estate saw the strongest year-over-year growth in ROAS of any of the eight industries in the report, up 44.35%.

That lower number does not necessarily mean real estate advertisers are performing poorly. The economics of the business are different. A single qualified lead can be worth thousands of dollars in commission, which means an advertiser can still run a profitable account at a much lower ROAS. Sports and apparel businesses often work with very different product prices and margins, so the return they need from ad spend can look very different too.

Local Services campaigns make the same point even more clearly. Their median ROAS in the report was just 5.30%, but the figure is heavily affected by how conversion value is tracked. Many Local Services conversions do not pass a dollar value back to Google, which makes the reported ROAS artificially low. Before comparing your account with any benchmark, make sure both numbers are measuring value in roughly the same way. Otherwise, the comparison can be misleading.

ROAS is a lever. It need not be a goal.

Fred Vallaeys makes the underlying point directly in his book, The AI-Amplified Marketer:

“ROAS is not a business goal, it’s just a lever to help you achieve the actual business goals of profitability or higher revenue.”

You can dial that lever up or down. What matters is whether the number you land on still makes you money.

This is what he calls the Genie Problem. Smart Bidding does exactly what you ask, not what you mean. Tell it to maximize conversions and it might hand you a pile of leads, 95% of them junk, because you asked for quantity when you meant quality. The same failure shows up when you copy a ROAS target from a benchmark report. You’ll likely hit the ratio, but you might not hit the profit.

This is where value-based bidding becomes more useful than simply adjusting the target. By assigning different values to different conversions, you give Smart Bidding a better signal about which outcomes matter most to the business. Fred recommends reviewing those conversion values every quarter, because the bidding system can only optimize based on the values it receives.

Let your business economics set the ROAS target

The best place to start is with your own margins rather than an industry median. A business with a 30% gross margin, for example, needs a higher ROAS to break even than a similar business operating at a 60% margin. That difference alone makes it hard to define one ROAS benchmark that works for everyone.

Your own historical performance is also a more useful reference point. Look at how ROAS has changed over the past few quarters and whether those changes have come alongside healthier revenue and profitability. An account that is maintaining or improving its return over time may be in a much better position than one that simply happens to sit above an industry median.

Benchmarks can still give you useful context when the comparison is relevant. Optmyzr’s PPC Vertical Benchmarks lets you compare your account with advertisers in your own industry rather than relying on the overall Google Ads median. If the values you’re sending to Google don’t accurately reflect what different customers or conversions are worth, value-based bidding is a better place to start than simply changing your ROAS target.

Google’s August 17 bidding change also makes the target itself more consequential for budget-limited campaigns, where the ROAS target you choose is now more closely tied to the result the bidding system aims to deliver.

So I wouldn’t take 557.17% from the Q2 report and use it as the definition of a good ROAS. Instead, pull your own ROAS for the past four quarters, compare the trend with your margins and business results, and then use the market benchmark as context for what you’re seeing.