ROI vs ROAS: Why a 4× Campaign Can Still Lose Money

Why the two numbers disagree

ROAS is the number the ad platform shows you, because it is the only one the ad platform can compute. It knows what you spent with it and what revenue it can attribute. It does not know your cost of goods, your shipping, your payment processing fees, your salaries or the agency retainer that produced the creative.

ROI is the number your accountant recognises, because it counts what the business actually kept. And the gap between them is not a rounding error — on a thin-margin business it is the difference between a campaign to scale and a campaign to stop.

The reason this persists is that both numbers are honest within their own frame. Nobody is lying. ROAS is simply answering a narrower question than the one being asked of it.

ROAS asks whether the ads paid for themselves. ROI asks whether the business made money. Those are not the same question and they frequently have different answers.

The definitions, precisely

ROAS — return on ad spend

Revenue ÷ ad spend, reported as a multiple. Spend ₹50,000, generate ₹200,000 in attributed revenue, that is 4×.

It counts only media cost, and only revenue. It is the right measure for comparing two ad sets against each other inside the same account, where the other costs are identical and cancel out.

ROI — return on investment

(Gross profit − total cost) ÷ total cost, reported as a percentage. Same campaign at 25% margin: ₹200,000 revenue is ₹50,000 gross profit. Minus ₹50,000 of ad spend, that is zero. 0% ROI on a 4× ROAS.

And that is before the agency fee, the creative production, the tools, or anybody's time. Include those and the 4× campaign is losing money.

The two supporting numbers

Cost per lead = spend ÷ leads. Easy to compute, and the least useful number in the set on its own. Cost per acquisition = spend ÷ customers, which is where the truth lives — and which gets its own post, because the gap between the two explains most bad channel decisions.

Your break-even ROAS

This is the number worth writing on a wall. It is set entirely by your gross margin:

Break-even ROAS = 1 ÷ gross margin

Read the table twice. A 3× ROAS is a triumph for a software business and a slow bankruptcy for a retailer. This is why the endless online argument about what constitutes a good ROAS never resolves — everyone is right about their own margin and wrong about everyone else's.

One important adjustment: for repeat-purchase and subscription businesses, break-even on first order is often perfectly rational, because the second and third purchases carry no acquisition cost at all. If that is your model, the input that matters is lifetime value, not order value, and using the latter will understate your returns substantially — frequently by several multiples.

The costs that get left out

When ROI is calculated at all, it is usually calculated on media spend alone, which quietly makes it ROAS wearing a percentage sign. The costs most often omitted:

• Agency or freelancer fees. On a small budget this can exceed the media spend itself.
• Creative production. Photography, video, design, copy. Amortise it across the campaigns that used it.
• Tools and platforms. The landing page builder, the email platform, the analytics stack.
• Internal time. The most commonly ignored and often the largest. Somebody spent four days on this.
• Discounts and offers. A 20% launch discount comes directly out of margin and is rarely reflected in the margin figure people enter.
• Returns and refunds. Attributed revenue in an ad platform generally does not net these off.
• Payment processing and shipping. Small individually; not small across a quarter.

You do not have to include all of them. You do have to be consistent about which ones you include, or two months are not comparable — and inconsistent inputs are how a campaign appears to improve for reasons that turn out to be an accounting change.

Which number to use when

Use ROAS when comparing inside one channel

Two ad sets in the same account, same product, same margin. All the other costs are identical, so they cancel, and ROAS is a clean comparison. This is the job it was built for and it does it well.

Use ROI when comparing across channels

Meta Ads against SEO against email. The cost structures are completely different — SEO is mostly time, email is mostly a platform fee, ads are mostly media — and ROAS cannot see any of that. Only ROI puts them on the same scale.

Use ROI when deciding whether to continue at all

“Should we keep spending on this?” is a business question, not a media question. It needs margin, total cost and the honest revenue figure.

Use both when reporting to somebody who pays for it

Show the ROAS because it is the number they will see in the platform anyway, and show the margin-based ROI beside it. Presenting only the flattering one is how trust in marketing reporting gets destroyed — not through a single lie, but through a number that was technically true and consistently optimistic.

Why a calculator that shows its working

Most ROI calculators return a number and expect you to trust it. That is exactly backwards for a figure you are about to use in a budget conversation, where the first question will be “where did that come from?”

The ROI calculator here shows the formula behind every figure it produces — cost per lead, cost per acquisition, conversion rate, revenue, ROI and ROAS. You can check the arithmetic, rebuild it in a spreadsheet, or hand it to a finance team without being asked to explain a black box.

It also reports margin-based ROI alongside revenue-based ROI, and says plainly that leaving the margin field empty flatters the result. That is the number most calculators quietly omit, and it is the one that changes decisions.

And it is arithmetic in your browser. Nothing is uploaded, logged or stored — which is deliberate, because spend, margin and customer value are among the most commercially sensitive numbers a business has, and there is no reason to post them to somebody's server to perform a division.

One thing it cannot fix: bad inputs. If your conversions are being counted twice, or not at all, every figure downstream is wrong no matter how sound the formula. Clean campaign tagging and working conversion tracking come first — the calculator is the last step, not the first.

FAQ