ROAS

What ROAS measures — and what it doesn't

Return on ad spend compares the revenue attributed to advertising against what the advertising cost:

ROAS = Attributed revenue ÷ Ad spend

A ROAS of 4.0 means the measurement system credited four units of revenue for every unit spent. Note the verb: credited. ROAS is not a fact about money in the bank — it is an output of an attribution system, and it inherits every assumption that system makes about windows, touchpoints, and credit rules. Two platforms reporting ROAS for the same campaign routinely disagree, and both can be internally correct.

ROAS is a campaign efficiency measure. It answers "how hard is this spend working?" It does not answer "is this profitable?" — that requires margin — and it does not answer "should I scale this?" — that requires knowing how ROAS decays as spend increases.

How to calculate it properly

The formula is trivial; the inputs are not.

Revenue side. Decide what counts: gross revenue or revenue net of refunds and discounts? First purchase only, or repeat purchases within the window? Platform-attributed revenue, or your own analytics? Each choice produces a different ROAS from identical campaigns. Write the definition down and keep it constant across comparisons.

Spend side. Include the full cost of the channel where possible — agency fees and creative production change the real return even though platforms exclude them. Many teams track platform ROAS for optimization and a fully-loaded ROAS for budgeting.

Worked example. A campaign spends $5,000 and the platform attributes $18,500 in purchase revenue within a 7-day click window: ROAS = 18,500 ÷ 5,000 = 3.7. If your contribution margin is 40%, the margin-adjusted return is 3.7 × 0.4 = 1.48 — meaning $1.48 of contribution per $1 spent, before fixed costs. The ROAS calculator handles these variants.

Break-even ROAS is the floor below which spend loses money:

Break-even ROAS = 1 ÷ Contribution margin

At a 40% margin, break-even is 2.5. A 3.0 ROAS that looks healthy in a dashboard is thin in reality once margin enters the picture.

Benchmarks and targets

Public "average ROAS" figures are mostly noise: they blend attribution settings, industries, margins, and funnel stages that aren't comparable. The usable reference points are internal:

• Break-even ROAS from your margin — the only universal threshold, and it differs per product line.
• Your own history per channel and campaign type — retargeting always reports higher ROAS than prospecting because it claims credit for demand that already existed; comparing the two is a category error.
• Marginal ROAS as you scale — the return on the next dollar, which falls as targeting saturates. Average ROAS can look fine while marginal ROAS has dropped below break-even.

Using ROAS day to day

1. Fix the measurement before judging the number. Confirm conversion tracking fires once per purchase with correct values. A ROAS built on duplicate or missing events is fiction.
2. Set targets per campaign role. Prospecting, retargeting, and brand defense have different jobs; one blended target pushes budget toward whatever claims credit most easily.
3. Mind conversion lag. Recent campaigns under-report because revenue hasn't arrived yet. Evaluate periods after the attribution window closes, or use lag-adjusted reporting.
4. Cross-check platforms against a neutral source. Google Analytics 4 for web, an MMP such as AppsFlyer for apps. The platforms will sum to more revenue than you earned — overlapping credit is normal; the neutral source keeps proportions honest.
5. Pair ROAS with LTV when repeat purchase matters. Optimizing first-purchase ROAS systematically undervalues channels that recruit loyal customers.

Common mistakes

• Treating revenue as profit. A 5.0 ROAS on a 15% margin product loses money. Margin-adjust before celebrating.
• Comparing incompatible attribution windows. A 28-day click+view ROAS versus a 7-day click ROAS is not a comparison; it's two different metrics sharing a name.
• Ignoring delayed and repeat revenue. Subscription and considered-purchase businesses see most value land after the window closes.
• Scaling on average instead of marginal return. The averages stay green while the incremental dollars quietly go red.
• Letting ROAS allocate all budget. ROAS measures attributed efficiency, not incrementality — it cannot see conversions that would have happened anyway. Branded search famously posts spectacular ROAS for exactly that reason.

FAQ

What's a good ROAS? Whatever clears your break-even (1 ÷ contribution margin) with room for overhead and growth — typically that means meaningfully above break-even, not at it. A good ROAS for a 70%-margin SaaS is unprofitable for a 12%-margin retailer, which is why universal benchmarks mislead.

Why do Meta, Google, and my analytics all report different ROAS? Different attribution windows, different credit rules, and each platform claiming conversions it touched. Their numbers answer "what did spend on my platform influence?" — summing them double-counts. Use one consistent source for cross-channel decisions; the broader workflow is covered in the paid acquisition path.

ROAS or CPA — which should I optimize? ROAS when order values vary widely (it weighs a $30 and a $300 order correctly); CPA when conversions are roughly uniform in value, such as leads or installs. Many teams bid to ROAS for e-commerce and CPA for lead generation.

Does ROAS include ad platform fees? Platform spend yes; everything around it (agency fees, creative costs, tools) no, unless you add it. For budget decisions, compute a fully-loaded version — the channels with heavy production needs look noticeably different.

Why did my ROAS drop when I increased budget? Diminishing returns: platforms spend your first dollars on the likeliest converters and each increment reaches colder audiences. Falling marginal ROAS with rising spend is expected; the question is whether the marginal return is still above break-even, not whether the average dipped.