A founder showed me a report last month. Meta said 6.2x ROAS. Google said 8.1x. By those numbers, every ₹1 spent was returning ₹6 to ₹8. The founder should have been thrilled. Instead, he was confused — because his actual bank account wasn't growing anything like those numbers suggested it should.
He wasn't being scammed by a dishonest agency. He was being misled by the platforms themselves — and his agency was simply passing along the numbers the platforms reported without questioning them. That's the more common version of this problem, and it's worth understanding exactly how it works, because it's costing Indian businesses enormous amounts of misallocated budget.
1. The 6x That's Actually 1.8x
Here's the founder's actual situation, simplified. He was spending ₹5,00,000 per month on ads. The platforms claimed credit for ₹34,00,000 in attributed revenue — a blended 6.8x. But his total business revenue, across every channel, was ₹20,00,000. His total ad spend was ₹5,00,000. So his actual marketing efficiency was 4x at the business level — not 6.8x.
And it got worse. When we ran an incrementality test — turning off his "best performing" retargeting campaign for two weeks — his total revenue barely moved. That campaign was reporting 12x ROAS. Its real incremental contribution was close to zero, because it was almost entirely reaching people who were already going to buy.
The platforms reported ₹34 lakh of revenue against ₹20 lakh that actually existed. They were each claiming credit for the same sales — and some sales that would have happened with no ads at all. That ₹14 lakh gap between reported and real isn't fraud. It's how the attribution system is designed to work, and it works in the platforms' favour every time.
2. What ROAS Actually Measures (And What It Pretends To)
ROAS — Return on Ad Spend — is supposed to answer a simple question: for every rupee I spend on ads, how much revenue do I get back? The honest version of that question has one more word in it: how much additional revenue do I get back — revenue that wouldn't have existed without the ad?
That word, "additional" (or "incremental," in the technical term), is the entire ballgame. Platform-reported ROAS systematically ignores it. When Meta reports that a campaign drove a sale, it's not claiming the ad caused the sale — it's claiming the ad touched the customer somewhere along the way. Those are completely different things, and the difference is where your money disappears.
3. Inflation Mechanism 1 — Attribution Window Games
Meta's default attribution is 7-day click. This means if someone clicks your ad on Monday and buys anything on your site by Sunday — for any reason, through any path — Meta claims credit for that sale. The customer might have clicked your ad, forgotten about it, then come back six days later through a Google search and bought. Meta still counts it.
The longer the attribution window, the more sales the platform can claim. A 1-day click window shows lower ROAS than a 7-day window, which shows lower than a 28-day window — for the exact same campaign, with the exact same real performance. Nothing changed except how generously the platform credited itself. Agencies that want to show impressive numbers simply select the longest attribution window available. The campaign didn't get better; the accounting got more flattering.
4. Inflation Mechanism 2 — Branded Search Counted as Paid
One of the most common ROAS-inflation tactics: running Google Ads on your own brand name. When someone searches "[Your Brand] India" — they already know you, they already want you, they were always going to find you. But if you're running a branded search ad, that click goes through the ad, and Google claims the sale as paid-driven. It reports a spectacular ROAS because the "customer acquisition" already happened organically.
Branded search campaigns routinely show 15-30x ROAS, which agencies love to put in reports. But these sales would have happened anyway through the organic result directly below the ad. The ad didn't create the customer — it just charged you to intercept a customer you already had. There are legitimate defensive reasons to run branded search (competitors bidding on your name), but counting it as acquisition performance is one of the most misleading things in a typical ROAS report.
5. Inflation Mechanism 3 — View-Through Attribution
View-through attribution credits the platform for a sale when a user merely saw an ad — didn't click it, may not have even consciously noticed it — and then bought later through some other path. Meta can show your ad in someone's feed, they scroll past it without engaging, they buy from you a day later because they were already a customer, and Meta counts that as an ad-driven conversion.
View-through attribution is the single most generous gift the platforms give themselves. At sufficient ad volume, your ads are being "seen" by a huge portion of your potential customers — so the platform can claim view-through credit for a large share of all your sales, including the ones that had nothing to do with the ads. Always check whether the ROAS you're being shown includes view-through conversions. If it does, the real number is meaningfully lower.
"If an ad can take credit for a sale the customer made without ever clicking it, then 'ROAS' has stopped measuring advertising performance and started measuring how much of your existing demand the platform can stand in front of."— Saksham Mehra, Founder & CEO, ENZO Digital
6. Inflation Mechanism 4 — The Platform Grades Its Own Homework
This is the structural problem underneath all the others. The companies reporting your ROAS are the same companies selling you the ad space. They have a direct financial incentive to make their own ads look as effective as possible — because the better the ads look, the more you spend. No platform with this conflict of interest will ever choose conservative attribution over generous attribution by default.
This doesn't make the platforms evil — it makes them self-interested, which is exactly what you'd expect. But it means their reported numbers should be treated as a sales pitch, not as objective truth. You would never let a vendor both deliver a service and be the sole judge of whether the service worked. Yet that's precisely the arrangement every advertiser accepts when they take platform-reported ROAS at face value.
7. The Number That Actually Matters — MER
Here's the good news: there's one number that cannot be inflated by any of these mechanisms, because it doesn't rely on attribution at all. It's called MER — Marketing Efficiency Ratio, sometimes called blended ROAS.
MER is brutally simple: total revenue divided by total ad spend. It doesn't care which platform claims which sale. It doesn't care about attribution windows or view-through or branded search. It just measures the one thing that actually matters — for every rupee you put into advertising across everything, how many rupees came into the business across everything.
The most useful thing about MER is what it reveals over time. When you increase ad spend and your platform-reported ROAS stays high but your MER drops, that's the tell: the platforms are claiming credit for more sales, but those sales aren't actually new. The ads are getting more "efficient" on paper while the business gets less efficient in reality. No amount of attribution cleverness can hide a declining MER.
8. How to Read Your Real Numbers
You don't need a data science team to stop being misled. Here's the practical approach:
| Instead of trusting... | Do this |
|---|---|
| Platform-reported ROAS as truth | Treat it as a directional signal only — useful for comparing campaigns within a platform, useless as an absolute measure of business impact |
| A single blended ROAS number | Track MER (total revenue ÷ total ad spend) every month as your north-star efficiency metric |
| "Our best campaign does 12x" | Run an incrementality test — pause it for 2 weeks, measure the actual revenue drop. If revenue barely moves, the ROAS was fiction |
| Branded search ROAS in the headline | Separate branded from non-branded reporting — judge acquisition performance only on non-branded campaigns |
| View-through conversions in the total | Ask for click-only attribution numbers and compare — the gap tells you how inflated the headline was |
| Reported ROAS rising as success | Check if MER rose too. If reported ROAS is up but MER is flat, the platforms are just reshuffling credit |
The single most clarifying exercise: every month, write down your total revenue and your total ad spend, divide one by the other, and track that one number over time. When your agency tells you performance improved, ask one question — "did our MER go up?" If they don't know what MER is, or can't answer, that tells you something important about whether they're measuring reality or just forwarding the platforms' marketing.
None of this means paid advertising doesn't work — it absolutely does, and most businesses should be running it. It means the numbers you use to judge whether it's working need to come from your business reality, not from the sales pitch of the company selling you the ads. An agency worth its retainer reports MER and incrementality alongside platform ROAS — and tells you the difference. We've written separately about how to evaluate whether your agency is actually being straight with you.
Want to Know Your Real ROAS — Not the Platform's Version?
ENZO Digital reports MER and incrementality alongside platform numbers, so you know what your ads are actually doing for your business.
Get an Honest Audit →