Return on Ad Spend (ROAS): What Counts as a Good Number

1. What Return on Ad Spend Actually Measures

Quick Answer: Return on ad spend is revenue attributed to your ads divided by what you spent on those ads. Spend RM 10,000, get RM 40,000 back, and your ROAS is 4x or 400%. It measures revenue efficiency only — not profit, not cash, and not whether the sale needed the ad at all.

The formula is one line: revenue ÷ ad spend. Google expresses it as a percentage, so a 500% target ROAS means RM 5 of revenue for every RM 1 spent, as set out in Google’s Target ROAS documentation. Meta and TikTok express it as a multiple. Same maths, different notation.

What trips people up is what it leaves out. It counts revenue, not margin — ignoring cost of goods, staff, rent and agency fees. A 6x ROAS on a 10%-margin product loses money. A 2x on software at 85% margin prints it. Three metrics get confused with it constantly:

• ROAS is a revenue ratio. Revenue from ads ÷ ad spend. Ignores every other cost.
• ROI is a profit ratio. (Profit − total cost) ÷ total cost. Always smaller, and more honest.
• CAC is a cost per buyer. Acquisition spend ÷ new customers won — see what a buyer actually costs you.

Return on ad spend fits when each sale carries a value you can measure as it happens — e-commerce, bookings, quotable jobs. When the sale closes offline weeks later, cost per lead does a better job.

2. Break-Even ROAS: The Number Your Margin Sets

Quick Answer: Break-even ROAS equals 1 divided by your gross margin. At 50% margin you break even at 2x; at 25% margin you break even at 4x. Anything below that line loses money on every sale, no matter how healthy the ratio looks in the dashboard.

This calculation should happen before anyone opens Google Ads, and almost never does. At 40% gross margin, every RM 100 of revenue leaves RM 40 to pay for the ad — so RM 40 of spend producing RM 100 is break-even, a ROAS of 2.5x. Below that, you’re paying customers to take your product. Across common margin profiles, the famous benchmark stops looking like a target:

Two things jump out. 4:1 is break-even at 25% margin — the benchmark everyone chases is where a general retailer makes exactly nothing. And then the right-hand column, which almost no ROAS guide covers and where Malaysian sellers get caught.

If you sell on Shopee or Lazada, or ship cash-on-delivery, your margin isn’t your margin. Marketplace commission, payment processing, free-shipping subsidies and unpaid COD parcels all sit between the revenue in your dashboard and the money in your account. Strip twelve points out and a 35%-margin fashion seller needs 4.3x to break even — nowhere near the 4:1 they were told was a win.

3. What Counts as a Good ROAS in Malaysia

Quick Answer: A good return on ad spend is any figure comfortably above your own break-even. Across ZenWeb’s Malaysian accounts, typical reported ROAS runs 3x to 7x depending on sector — but several sectors sit at or below their break-even line while the dashboard still shows a healthy-looking multiple.

Sector medians are useful for one thing only: telling you whether your account is normal for your category. They are not targets. Here’s how reported figures compare with each sector’s break-even, drawn from campaigns under ZenWeb management.

Electronics is the cautionary tale. A 6.1x return on ad spend gets celebrated in most meetings, yet on a 15% margin it loses money on every order. The supplement seller at 7.2x against a 2.2x break-even runs one of the healthiest accounts in the sample — and nobody mentions it.

Benchmarks like our Malaysian CPC, CTR and CPL benchmarks tell you if your inputs are competitive. Only your margin tells you if the output is profitable.

4. Why Your Reported ROAS Beats Your Real One

Quick Answer: Platforms report the return on ad spend they can claim credit for, not the revenue you kept. Brand search, retargeting and view-through conversions all book sales that would have happened anyway. Across ZenWeb accounts the gap between reported and verified ROAS runs from 16% on non-brand search to 76% on retargeting.

Every ad platform is both player and referee. Google counts a conversion when its click was involved; Meta when its ad was seen. Neither asks whether the sale needed the ad. Add returns, cancelled COD parcels and double-claimed orders, and the reported figure drifts far from your bank balance.

Notice the pattern: the channels reporting the best return on ad spend inflate it most. Brand search and retargeting both advertise to people who already decided. They look spectacular precisely because they take credit for demand you created elsewhere.

Timing distorts things too. Google advises excluding the most recent conversion delay period when evaluating ROAS, because conversions keep landing after the click. Meta’s default windows credit both clicks and views, which is why its numbers routinely exceed analytics — see why Meta claims more sales than GA4, and choosing an attribution model for the Google side.

Three fixes shrink the gap, none requiring enterprise tooling:

1. Split brand from non-brand. Blending hides whether your ads create demand or just intercept it.
2. Feed back returns and cancellations. Import net revenue so your conversion values reflect money you kept.
3. Reconcile monthly against your books. Reported revenue above what your accounts recorded is double-counting.

Tightening the underlying data is the real fix: server-side tracking recovers signal browsers block, first-party data gives platforms something durable to match against, and cookieless measurement is how this gets read once third-party cookies go.

5. Average ROAS vs Marginal ROAS

Quick Answer: Average ROAS tells you how the whole account performed. Marginal ROAS tells you what the next ringgit returns — and it’s always lower. Accounts routinely show a comfortable 3.3x average while the last slice of budget returns 1.4x and quietly loses money.

This is the distinction that decides budget, and the one most owners never see. Account ROAS averages every ringgit you’ve spent, including the cheap early ones that were always going to convert. But the decision you face — should I spend more? — depends on the next slice. Returns diminish as budget grows: the first RM 5,000 reaches people already searching for you, the tenth reaches people who’ve barely heard of you.

At RM 50k a month this account still reports 2.9x — respectable, nothing obviously wrong. But the last RM 15k returned 1.4x, burning roughly RM 9,000 of margin a month while the dashboard looks fine, because strong early spend props up the average.

The usual advice is to run geo lift tests or build a media mix model. On RM 20k a month, that isn’t happening. The practical version: read marginal ROAS out of your own budget history. Every budget change you’ve made was an experiment — compare revenue and spend either side of it, divide the differences, and you have that step’s marginal ROAS. Knowing when to scale comes down to exactly this.

6. How to Lift ROAS Without Fooling Yourself

Quick Answer: Cutting spend raises ROAS and shrinks profit — the easiest way to win the metric and lose the business. Real gains come from the revenue side: higher order value, better conversion rates, accurate conversion values. Most reported “improvements” are measurement changes, not business changes.

ROAS is a ratio, so there are two levers: shrink the bottom or grow the top. Most guides reach for the bottom, because pausing campaigns improves the chart instantly — and produces less profit. A 6x return on ad spend on RM 5k earns less than a 3x on RM 30k. So grow the top. Five ways, in rough order of payoff:

• Raise average order value. Bundles, volume breaks and free-shipping thresholds lift revenue per conversion without touching spend — the fastest honest ROAS gain available.
• Fix the landing page before the bids. Doubling conversion rate doubles ROAS at identical spend. Our CRO basics for Malaysian sites covers the usual leaks.
• Send real values, not flat ones. If every order reports RM 100 regardless of size, Smart Bidding chases the wrong customers. Accurate values let PPC bidding pursue profit over volume.
• Bid warmer, not wider. Remarketing lists for search ads bid up people who already visited — just report them separately so their inflated returns don’t flatter the account.
• Let budget follow performance. Campaign budget optimisation on Meta and automated rules on Google shift spend toward what works without daily babysitting.

On targets: Google recommends setting a target ROAS at or below your historical performance, warns that too high a target limits traffic, and requires at least 15 conversions in the past 30 days for Search and Shopping. An aspirational target on a thin account strangles campaigns that were fine.

Three habits do the opposite — raising the number while the bank balance sits still:

• Summing platform reports. Google and Meta both claim the same order. Adding their revenue together invents sales that don’t exist.
• Ignoring agency fees and tools. Media spend isn’t total spend. Include management fees and software before calling anything profitable.
• Forgetting COD. Cash-on-delivery is still common here, and the platform books the conversion when the order is placed, not when the customer accepts the parcel. Every refused delivery is revenue in your ROAS and a loss in your warehouse — no amount of bid tuning fixes that.

For the account-level versions of these, see 10 Google Ads mistakes that waste money.

7. How to Know Your ROAS Is Real

Quick Answer: A trustworthy return on ad spend reconciles with your accounts, holds up when you pause a channel, and separates brand from non-brand. If revenue doesn’t move when spend does, the ROAS was reporting demand you already had.

Five checks, none requiring specialist software:

1. Reconcile against your books. Compare platform-reported revenue with what your accounting system recorded that month. A gap over 10% means double-counting or COD leakage.
2. Run a pause test. Turn one channel off for two weeks and watch total revenue, not the channel’s own figure. If overall sales barely move, that ROAS was mostly credit-taking.
3. Compare against break-even. Sit your figure next to the number from your margin — the only comparison that says profitable or not.
4. Check the trend, not the day. Read ROAS over 30 days minimum, excluding the last few days of unconverted clicks.
5. Split brand and non-brand. If the account only looks good with brand included, it isn’t good.

The pause test is the most uncomfortable and the most useful — the closest an SME gets to a real incrementality experiment. For a lighter routine, see whether Google Ads are actually profitable.

With 98.0% of Malaysians online as of late 2025, per DataReportal, nearly every customer you’d win offline is reachable through paid search anyway — which is exactly why brand traffic gets mistaken for incremental return. SEO in Malaysia and search engine marketing lower how much brand demand you rent back.

8. Conclusion

Quick Answer: A good return on ad spend is the one that clears your break-even with room to spare, holds up when you verify it, and still clears break-even on the next ringgit you spend. Three tests — margin, honesty, and margin at the edge.

“What’s a good ROAS” has no benchmark answer — and the benchmark answers are the dangerous part. Work out break-even from margin after fees and returns, verify the reported figure is revenue you kept, then judge new budget on marginal return rather than the flattering average.

Do those three and the metric becomes what it should be: an honest read on whether your advertising buys growth or rents demand you already owned. The complete Google Ads guide for Malaysian SMEs is the wider playbook, and ZenWeb runs these accounts for 500+ Malaysian businesses as a Google Partner.

9. Frequently Asked Questions

1. What is a good return on ad spend in Malaysia?

There’s no single good number. A good return on ad spend is any figure above your break-even, which is 1 divided by your gross margin after marketplace fees and returns. A 50%-margin brand breaks even at 2x, so 4x is strong. A 15%-margin electronics seller breaks even near 6.7x, so the same 4x loses money.

2. Is 4:1 ROAS good?

Only above roughly 25% margin. At exactly 25% gross margin, 4:1 is your break-even — you make precisely nothing. The rule circulates because it suits typical American e-commerce margins, and ignores the marketplace commission, payment fees and COD returns Malaysian sellers carry.

3. What’s the difference between ROAS and ROI?

ROAS is revenue divided by ad spend and ignores every other cost. ROI is profit divided by total cost, including goods, staff and overheads. ROAS says whether the ads generated revenue efficiently; ROI says whether the business made money. A campaign can post a strong ROAS and a negative ROI at once.

4. Why is my Google Ads ROAS different from my actual sales?

Three usual causes: conversions arrive after the click so recent data reads low, brand campaigns count buyers who were already coming, and returns or cancelled COD orders stay counted as revenue. Google advises excluding the recent conversion delay period when evaluating ROAS. Reconcile platform revenue against your books monthly to size the gap.

5. Should I use Target ROAS bidding?

Only with enough conversion data. Google requires at least 15 conversions in the past 30 days for Search and Shopping campaigns, and recommends setting the target at or below your historical ROAS. An aspirational target on a low-volume account limits traffic and can stall a campaign that was performing acceptably.