Meta is showing you 5x. Your accountant is showing you basically nothing. Both are looking at the same month. Here's why they disagree, and why the dashboard is the one you can't trust.
The ROAS on your dashboard is a credit metric, not a profit metric
Start with what the number actually measures. Meta takes credit for any sale that happens inside its attribution window. Default is 7 days after a click, 1 day after a view, and that view-through attribution is where a lot of the inflation lives. If someone sees your ad in passing and buys a day later, that's logged as a conversion. Multiply that by thousands of impressions and the reported ROAS balloons far past what your ads actually caused.
The word that matters there is "credit." Meta isn't asking whether the ad caused the sale. It's asking whether the ad was in the room when the sale happened. Those are very different questions, and only one of them matters to your bank balance.
Say 30,000 people saw your ad over a week. On Meta's logic, any of them who bought within a day of seeing it counts. The guy who was already comparing your product on three tabs before the ad showed up? Counted. The woman who clicked an ad from another brand and yours hung around in her feed on the way? Counted. It's not lying exactly. It's just answering a different question than "did my spend produce this."
Returning customers inflate it further
If a loyal customer comes back and buys, and any of your ads touched them in the window, Meta books the full sale as ad revenue. It didn't acquire that person. They were already yours.
Think about how this compounds on a real account. If you've been running for two years, a decent slice of your monthly revenue is repeat buyers. Say 25% of your orders come from customers who bought before. If your ads touched even some of them in the attribution window, Meta absorbed those orders into its reported revenue. Half the "ad revenue" it reports can be customers you already had, which makes the ROAS look strong while the acquisition economics are underwater.
And here's the brutal part. Those repeat customers were going to buy without any ad. Some of them you could have reached with an email that cost nothing. As your brand grows, this inflates reported ROAS more and more while the actual cost of winning a new customer keeps rising. The dashboard tells you the business is getting more efficient. Your books know it isn't.
The margin problem nobody checks
Say Meta shows 5x. Sounds amazing. Now run the real math.
A Rs.1,500 order. Product cost Rs.500. Shipping and packaging Rs.120. Payment and platform fees Rs.60. Returns reserve at 15% Rs.225. That's Rs.905 gone before a single rupee of profit, leaving Rs.595 gross profit. Your break-even ROAS is 1,500 divided by 595, which is 2.5x.
So 5x is profit. But understand what kind. At 5x, every Rs.1,000 of ad spend produces Rs.5,000 of revenue and Rs.2,975 of gross profit. Fine. Now stress it with the things founders forget. If your actual freight cost jumped 20% this quarter, or product cost rose with your supplier, that Rs.595 of profit sits on a wobbly base. Drop it to Rs.430 and your break-even moves to 3.5x. Your returns run hotter than 15% some months, fashion and electronics especially. At 25% returns, break-even is 2.7x.
That's how a 5x gets eaten. Not by one big change, by the quiet ones, and the dashboard never subtracts any of it. It just shows 5x and lets you feel good.
The objection I always get
"If it's all inflated, why does the account keep making money?" Because sometimes it does. High ROAS with healthy real profit happens, mainly on high-margin products sold to a genuinely cold market. The 5x on a 60% margin product at a 1.67x break-even really is money.
The failure mode is assuming that's your situation without checking. Check the margin, check the returns, check how much revenue is repeat. Ten minutes with your books answers it. Most founders who skip it find out three months in, when scale turns a thin margin into a loud loss.
What to do
Run this once a month, and it changes how every decision looks:
- Pull your real gross margin from your books, with a returns reserve honest enough to sting.
- Calculate break-even ROAS from it.
- Calculate acquisition MER, total revenue divided by total ad spend, both from outside the platforms.
- Compare your reported ROAS to break-even, and your MER to it too.
If your MER sits around 1.6x while the platform says 5x, the MER is closer to the truth, and you plan spend against that. The dashboard stays useful for choosing which creative to kill. The books stay the voice on whether the business works.
The biggest objection: "my margins are fine, so why bother?"
Because "fine" is doing a lot of quiet work in that sentence, and this article's whole point is that fine margins break at scale. If your 5x account really runs above break-even with room to spare, you're in the healthy minority, and the discipline still earns its keep in two ways. It tells you when the room disappears, because the dashboard will stay at 5x for weeks after your freight contract, supplier, or return rate shifts under you. And it tells you which margin lever pays most, because once you know that returns are 22% rather than 15%, fixing returns might beat tweaking ads entirely. A founder who can recite their break-even spends their quarter on the right problem. A founder who can't spends it wherever the dashboard pointed last.
I'm Hirdesh Matai, founder of HDM GRO-ads, the science of e-commerce advertising.