Your ROAS is flattering you: how to work out what you really keep

A dashboard that says 4.0x can sit on top of a month that lost money. ROAS is revenue divided by ad spend, and the revenue half of that fraction is where the flattery lives: it's what the ad platforms claim, not what reached your bank, and it's gross of everything the product, the delivery and the failed parcels cost you.

The receipt your dashboard doesn't print has six lines, and the numbers for it sit in four places: your ad accounts, your store, your courier's reports and your books. Take a made-up month in which the dashboard shows 4.0x and the store loses AED 5,150.

The receipt, in six steps

1. 01Take the revenue your ad platforms reported for the month and add them together. Meta, Google and TikTok can each claim the same order, so the total starts out suspect. Divide it by ad spend and you have the dashboard ROAS, the number you're about to correct.
2. 02Match that revenue against the store. Export the month's orders with their source, and keep only the ones you can tie to a click. If you can't match orders, compare totals instead: when the platforms together claim more than the store took in from all channels, the excess is double-counting.
3. 03Take off orders that never turned into money: refused, unreachable, cancelled after dispatch and returned. Use the courier's report and your store's statuses, and wait until every parcel in the month has a final status. If last month's parcels haven't settled, use the month before.
4. 04Take off what the products and their delivery cost on the orders that did arrive: landed product cost, packaging, outbound delivery and payment fees. If you don't have it by product, use one blended percentage for now and refine your top sellers later.
5. 05Take off the courier cost of the parcels that came back, trip out and trip back. It's the line most receipts leave out, because the courier bills it separately. Ask for a monthly breakdown of failed deliveries if your invoice doesn't show it.
6. 06Take off the ad spend, plus whatever you pay someone to run it. What's left is what you kept. If it's negative, don't cut the budget first: find which of steps 2, 3 and 5 is the biggest leak.

A made-up month, worked through

Picture a made-up store that spent AED 40,000 on ads in a month, with its platforms reporting AED 160,000 of revenue. Suppose its orders average AED 200, 25% of orders never turn into money, products and delivery cost 60% of the money that comes in, and each failed parcel costs AED 35 for the trip out and back.

In the made-up month the dashboard said 4.0x, the money that came in said 2.55x, and the store lost AED 5,150. The 25% failure rate is a made-up input, though it falls within the 20% to 35% range Quiqup, a UAE delivery company, quoted for many UAE merchants' return-to-origin rates in May 2026.

What is a good ROAS for a UAE e-commerce store?

The only good ROAS is one above your break-even, and break-even is 1 divided by the margin you keep before ads. In the made-up month the store keeps AED 34,850 of the AED 102,000 that came in, which is 34%, so its break-even ROAS is 2.9x. Its real ROAS of 2.55x sits below that, which is why it lost money.

The margin has to be measured on the same basis as the ROAS. In the made-up case, taking only the 40% left after products and delivery would put break-even at 2.5x, and a real ROAS of 2.55x would look fine. Counting the failed parcels moves it to 2.9x. Break-even ROAS is only useful once every leak is inside the margin.

Recompute break-even whenever the price changes. Imagine a product that keeps 40% of its price before ads and goes on sale at 20% off. The discount comes straight out of the margin, which falls to 25% of the new price, so the break-even ROAS would be 4.0x instead of 2.5x. Sale periods call for a higher ROAS, not a lower one, which is worth knowing before White Friday.

Any ROAS target you set in an ad platform has to be in the platform's own units, because the platform can't see your leaks. In the made-up month the dashboard needed to show 4.6x for the store to break even, which is the 160,000 reported divided by the 34,850 kept before ads, and it showed 4.0x. Recompute that ratio every month, because it only holds while your leaks stay the same.

Give the numbers time. Google suggests at least 30 conversions before you judge a Smart Bidding strategy, and 50 for Target ROAS, so don't reset a target every few days. And hold above break-even for four weeks in a row before you raise the budget.

Fix the biggest leak before you cut the budget

• Double-counting is the biggest gap: fix the tracking first, and compare the platforms' combined total with the store every week. Server-side tracking helps each sale count once.
• Failed orders are the biggest gap: confirm orders on WhatsApp before dispatch and optimise the ads to confirmed orders, as set out in Meta ads for UAE e-commerce. Return to origin (RTO) shows how to measure it.
• Margin is the biggest gap: raise the price, bundle, or stop advertising the products that can't carry their delivery.
• Every campaign looks fine but the business doesn't: check MER, total revenue divided by total marketing spend. Channel ROAS can rise while MER stays flat, which means the channels are claiming demand that already existed.

Two blind spots work in opposite directions. Cash on delivery orders are usually counted when the order is placed, so they inflate the dashboard. Sales closed in a WhatsApp chat may never reach it, so they deflate it. Add those by hand in step 1 until your tracking sees them.

Put the cost of running the ads on the ad-spend line. That includes us: our e-commerce marketing service is priced on scope from AED 12,000 a month, excluding VAT, and the fee belongs in this sum. If a store keeps AED 5,000 after ad spend and pays AED 12,000 a month to have its ads run, it has lost AED 7,000.

Run last month through the six steps tonight. If the answer surprises you, put it in front of whoever runs your ads before the next budget meeting.