Ad-reported revenue vs real revenue: why they differ
Why ad managers report more revenue than you made: attribution windows, double counting, GMV and cancellations. How to reconcile, and true ROAS.
By The DataxLand team· · 6 min read
Add up the revenue that Facebook Ads, Google Ads and your marketplace ads report for the month, and many shops get a figure larger than the shop's total revenue. No platform is lying, but each one counts by its own rules, and those rules are designed to answer "are my ads working?", not "is your shop profitable?".
This post explains six reasons ad-reported revenue differs from real revenue, gives an example on sample data, and shows how to reconcile the two figures into a ROAS you can actually use for budget decisions.
Numbers come from the DataxLand sample data (a simulated scented-candle and essential-oil shop selling on Shopee, TikTok Shop and its own website, 1 Aug – 29 Oct 2026). They are not a real customer's numbers.
Two numbers, two questions
- Ad-reported revenue (conversion value): the total value of the orders an ad platform credits to itself, under the platform's attribution rules.
- Real revenue (net revenue): the sales money the shop actually keeps after discounts, vouchers and refunds, counting each order once.
In the sample data, for the website channel alone: ad platforms report a conversion value of 222,400,000 ₫, while the website's net revenue for the whole period is only 129,572,000 ₫. The gap is 92,828,000 ₫, meaning the platforms report about 72% more than real revenue.
Six reasons the two numbers differ
| Reason | Makes the reported figure larger or smaller | Example |
|---|---|---|
| 1. Attribution window | Usually larger | A buyer clicks an ad, then 10 days later finds the website on their own and buys: still counted |
| 2. Several platforms claim the same order | Larger when you add platforms together | One order is counted as a conversion by both Facebook and Google |
| 3. Value before discounts, vouchers, shipping | Larger | The pixel sends the cart value, not revenue after vouchers |
| 4. Orders cancelled or returned after attribution | Larger | An order cancelled the next day still sits in the ad report |
| 5. Attribution by interaction date | Shifts over time | The conversion is assigned to the day of the ad click, not the order date |
| 6. Modelled conversions, missing events | Can be larger or smaller | Modelling fills in what could not be measured; a blocked pixel loses events |
On reason 1: each platform has its own conversion window. For example, Google Ads help defines the conversion window as the period after an ad interaction during which a conversion is still recorded, 30 days by default and settable from 1 to 90 days depending on the conversion source (Google Ads Help: About conversion windows, read on 16 Nov 2026). Facebook and TikTok also have their own attribution settings in their ad managers; check the setting used by each campaign before comparing numbers.
On reason 2: each platform only sees its own part of the journey, so each one can claim 100% of the same order. Adding up three platforms' reports counts one order three times.
Platform ROAS and true ROAS
Because reported revenue differs from real revenue, ROAS has two versions too:
Platform ROAS = Conversion value reported by the platform ÷ Ad spend
True ROAS = Net revenue of the channel ÷ Ad spend for the channelIn the sample data, website channel:
| Metric | Value | Note |
|---|---|---|
| Ad spend | 41,011,000 ₫ | Ads driving buyers to the website |
| Conversion value reported by platforms | 222,400,000 ₫ | Under the platforms' attribution rules |
| Website net revenue | 129,572,000 ₫ | After discounts, vouchers, refunds |
| Platform ROAS | 5.42 | Looks great |
| True ROAS | 3.16 | Use this for decisions |
| Contribution margin / net revenue | 41.8% | Before ads |
| Break-even ROAS | 2.39 | 1 ÷ 41.8% |
| Channel profit after fees | 13,175,700 ₫ | Contribution margin − ads |
Here the website ads are still profitable (true ROAS 3.16 is above the 2.39 break-even), but the profit is only about a quarter of the contribution margin, far from the "ROAS 5.4" impression the ad manager gives.
The opposite case is more dangerous: in the sample data, campaign TTA-201 on TikTok has a platform ROAS of 3.04 in September and
2.14 in October (to 29 Oct). The TikTok Shop channel's contribution margin is only about 18.5% of net revenue, so break-even ROAS on
net revenue is about 5.4. Even the platform-reported number is already below break-even, and it is falling. The detailed break-even
calculation is in What ROAS is profitable?.
Reconciling ad-reported revenue with real revenue
- Use the same period and the same time zone for the ad reports and the order data. A time-zone mismatch shifts daily numbers.
- Group ad spend by destination sales channel, not by ad platform: Facebook ads that send buyers to the website belong to the website; marketplace ads belong to that marketplace.
- Take the channel's net revenue from order data (after vouchers, net of returns), not from the ad manager.
- Compute the gap = reported conversion value − net revenue, weekly. Watch the trend: a sudden jump in the gap is often a pixel event firing twice or an attribution setting that just changed.
- Make budget decisions with true ROAS against the channel's break-even ROAS. Use platform ROAS to compare campaigns within the same platform, where the counting rules are the same.
When to trust platform numbers
Platform numbers are not useless. They are good for comparing creatives, audiences and campaigns within the same account, because their errors are roughly the same. They are not good for "are my ads profitable overall?" or "should I move budget from platform A to platform B?", because each platform counts by different rules.
How DataxLand does this
DataxLand puts the numbers side by side with fixed definitions: platform-reported ROAS (platform conversion value divided by spend), true ROAS (channel net revenue divided by the channel's ad spend) and the reported-vs-actual ad revenue gap (conversion value minus net revenue). Ad spend is assigned to the destination sales channel, then allocated to orders by net revenue to give profit after fees by channel and product. Each metric's definition and version is in Metrics and the semantic layer.
You combine your order source with your ad source, for example Shopee with Facebook Ads, then open the retail marketing dashboard. When you ask Si "true ROAS for the website last month compared with the ROAS Facebook reports", the answer states which metrics it used. See also the solution page Measure true ROAS.
Frequently asked questions
Why is the total revenue reported by ad platforms bigger than the shop's revenue?
Because each platform can claim the same order as its own conversion, on top of long attribution windows, pre-voucher values and orders that are later cancelled or returned. Adding the reports together double counts.
Should I shorten the attribution window to get closer numbers?
A shorter window makes the reported figure smaller but does not turn it into real revenue; it also changes how the platform optimizes delivery. Keep the setting that suits optimization, and use the channel's net revenue for profit decisions.
Should ROAS be calculated on GMV or net revenue?
On net revenue, and compared with a break-even ROAS also calculated on net revenue. ROAS on GMV always looks better because vouchers and returns have not been deducted.
Do shops that sell only on marketplaces see this gap?
Yes, but less: in-marketplace ads can see in-marketplace orders, but they still attribute by the marketplace's rules and usually on values before cancellations and returns. Reconciling with net revenue from settlements is still necessary.
Read next: What ROAS is profitable? · How to calculate true profit · Profit by SKU: finding best sellers that lose money