What is ROAS?

ROAS is total revenue generated from ads / total ad spend. Spend $10,000 on a campaign that returns $40,000 in attributed revenue and the ROAS is 4x, or 400%. Spend $10,000 to return $15,000 and the ROAS is 1.5x. The number is almost always expressed as a multiplier in conversation (“we're running at 3.2x ROAS”) and as a percentage in finance decks (320%), but the arithmetic is identical.

ROAS is reported automatically by every major ad platform — Meta Ads Manager, TikTok Ads Manager, Google Ads, LinkedIn Campaign Manager — against whatever conversion event the campaign is optimizing for. Different platforms attribute differently (Meta defaults to 7-day click + 1-day view, TikTok to 7-day click), so two platforms measuring the same campaign on the same store will report meaningfully different ROAS numbers. The metric is universal; the attribution behind it is not.

It is the most-tracked number in performance marketing because it combines a revenue input and a cost input into one clean ratio that fits on a dashboard tile. That virtue is also its hazard. ROAS collapses everything happening between the impression and the purchase into a single number, and any flaw in attribution, margin, or order economics gets hidden behind it.

The three metrics sit on the same axis but measure different things, and serious operators track all three. ROASis gross revenue divided by ad spend — revenue side only, no costs subtracted. ROI(return on investment) is net profit divided by ad spend — revenue minus cost of goods sold, fulfillment, payment processing, returns, and overhead, then divided by spend. ROI is what actually pays the rent; ROAS is a proxy for it.

MER(Marketing Efficiency Ratio, sometimes called blended ROAS) is total business revenue divided by total marketing spend, including organic traffic and channels you didn't pay for. MER is the cleanest read on whether the marketing function as a whole is generating leverage, because it's immune to the attribution games that distort platform-reported ROAS. A 4x platform ROAS with a 1.8x MER usually means the platforms are claiming credit for sales that organic would have closed anyway.

Operators with discipline track all three at once: platform ROAS to steer individual campaigns, ROI to gate channel-level decisions, and MER as the integrity check on the other two. When platform ROAS rises and MER doesn't, the channel is cannibalizing organic. When MER rises and ROI doesn't, you have an order economics problem, not a marketing problem.

Heavily category-dependent — there is no universal answer, and benchmarks pulled from generic blog posts are almost always wrong for any specific business. The right ROAS is the one that clears your gross margin floor plus overhead with room left for profit. That said, the durable patterns by vertical:

• —DTC ecommerce.2–4x ROAS on cold prospecting audiences, 4–8x on retargeting, 6–12x on email plus remarketing combined. Brands at the low end of these ranges usually have 60%+ gross margin; brands at the high end usually have thinner margins and need the multiple to cover overhead.
• —SaaS.ROAS is less useful here because the value of a customer is paid out over months or years, not at the first purchase. LTV/CAC is the dominant metric. That said, a first-purchase ROAS of 1.5–3x on a free trial or low-tier subscription is typical for healthy SaaS funnels.
• —Affiliate and creator campaigns.ROAS is often inverted — the “revenue” in the numerator is commission earned, not gross sales, so the multiples look smaller. A 1.4x ROAS on an affiliate campaign can be excellent if the commission rate and AOV combine to leave margin after the spend.
• —Lead generation.ROAS is calculated against pipeline-weighted lead value, not realized revenue. A 5–15x ROAS on raw lead value is typical, which after close-rate adjustment usually pencils to 2–3x on closed-won revenue.

Cold-prospecting ROAS varies by platform driven mostly by audience intent quality and attribution depth. The ranges below are 2026 medians for direct-response campaigns serving North American delivery, synthesized from The Ad Bench's calibration set and publicly reported benchmark data.

• —TikTok cold. 1.8–3.2x median. Lower than Meta because TikTok's in-house attribution under-reports and many buyers compensate by treating it as a top-of-funnel discovery channel.
• —Instagram Reels cold. 2–3.5x median. Meta's attribution stack credits Reels more aggressively than TikTok credits itself, which lifts reported ROAS even when underlying conversion is comparable.
• —YouTube Shorts cold. 1.5–3x median. Shorts is younger inventory with thinner direct-response signal than Reels or TikTok, but the price is cheaper and the variance is wider.
• —Facebook Feed cold. 2.5–4x median. The highest cold ROAS of the major consumer placements because Facebook's aging audience has the deepest history of in-platform purchases, which trains the algorithm to find buyers more efficiently.
• —LinkedIn cold. 1.2–2.5x on first-touch attribution. B2B sales cycles stretch ROAS measurement over months, and most LinkedIn campaigns under-report ROAS at the platform level while over-delivering pipeline downstream.

A 3x ROAS sounds healthy and is, often, unprofitable. The math is simple: 3x ROAS means revenue is 3x ad spend, which means ad spend is 33% of revenue. If COGS plus fulfillment plus payment processing plus returns plus overhead exceed the remaining 67%, the business loses money on every order — and every additional dollar of ad spend deepens the loss.

The standard bake-in for DTC: gross margin of 30–50% after COGS and fulfillment is the floor before ads, and a blended ROAS of roughly 2.5x or higher is needed to clear paid acquisition plus overhead and still leave operating profit. Brands with 65%+ gross margin can survive lower ROAS multiples; brands selling commodity products at 20% gross margin need ROAS well above 5x to make a channel work. The ROAS target is downstream of the unit economics, not the other way around.

The cleanest discipline is to compute a contribution-margin ROAS floorper SKU or category before any campaign launches. That number — the ROAS below which the campaign is destroying value — becomes the target. Everything above is upside; every dollar below is a leak that scaling will amplify.

Four levers, in descending order of how much they actually move the number. The order surprises operators who haven't worked on enough accounts to have seen it play out repeatedly.

• —Better creative.The single biggest lever — creative quality typically drives 30–60% of the variance in ROAS across accounts in the same category. Better hook, better native-feel, sharper CTA, and a clearer demonstration of product value usually move ROAS more than any targeting or bidding change.
• —Better landing page. 15–30% of ROAS variance. The page is where the click becomes revenue or doesn't, and most brands invest 90% of their creative budget in the ad and 10% in the page that receives the traffic. Reversing that ratio usually pays.
• —Better targeting. 10–25% of ROAS variance, and the number keeps shrinking as broad-by-default products like Advantage+ take over. Targeting matters less every year as the algorithms get better at finding buyers inside large audience pools.
• —Better offer. Variable but often huge. A 20% discount, a bundle, free shipping over a threshold, or a money-back guarantee can lift ROAS more than months of creative iteration. Most accounts under-test offer.

The practical implication: “fix the ad” beats “fix the audience” about seven times out of ten on a stalled account. See our companion piece on CPM for how creative quality also routes through delivery cost on the way to ROAS.

The four most common ROAS decay patterns, in roughly the order they appear on a scaling account.

• —Creative fatigue.The #1 reason ROAS decays. Paid creative needs to be refreshed every 7–21 days at meaningful spend levels; longer than that and frequency climbs, novelty drains, and CTR collapses. Our ad fatigue guide breaks down the refresh cadence by platform.
• —Audience saturation. At scale, the addressable buyer pool inside your target audience gets thinner and the next dollar of spend reaches a lower-quality fit. Broadening usually lowers frequency but also lowers average buyer-intent, so ROAS drifts down even as volume holds.
• —Seasonality.Q4 competition pushes CPMs up 40–70% above the trailing average, and ROAS compresses against the higher cost-per-reach. Strong Q4 ROAS usually requires strong Q3 audience-building so the December auctions are fought on retargeting, not cold.
• —Tracking degradation. iOS 14, cookieless browsers, and tightening data-sharing rules under-report conversions by 15–35% depending on platform and store stack. Real ROAS is usually higher than reported ROAS, which is why the MER cross-check matters.

The Ad Bench scores creative quality, which is the largest single lever on ROAS. Higher hook score, stronger native-feel, and a cleaner CTA architecture map to lower CPM (because every major platform rewards quality with delivery discounts) and to higher conversion rate (because a better hook produces a more intentional click and a more receptive post-click audience). Both effects compound into ROAS.

The Deep Dive report's predicted-metrics section forecasts a CPC, CVR, and CPA bracket for each uploaded creative based on the rubric scores and the platform target. Those projections feed directly into ROAS modeling: project AOV against the predicted CPA and you have a forward-looking ROAS estimate before a dollar is spent. For deeper context on the inputs to that model, see our articles on CPM, TikTok ad cost, ad fatigue, and the full ad budgeting guide.