After Black Friday, your store and your ad platforms will report different results for the season, because each source counts something different at a different moment. There is no single "true" number for every purpose. Revenue and margin come from the store. The platforms are for comparing campaigns against each other, not for counting how much you earned.
The gap does not need fixing. It needs writing down. The rule for which number you use for which decision is best set before 27 November, because nobody has time for that argument mid-season.
This article gives you five reasons the numbers disagree, a "decision → number" table, and four rules a leadership team can approve in one afternoon.
Cyber Monday, 6 p.m.: three reports, three answers
An illustrative scenario: a typical mid-sized online store, not a description of a specific company.
A D2C store, a few dozen people, a strong season. On Monday evening the Head of Growth opens three reports.
The store admin shows sales for the whole weekend. The ad manager of one platform shows attributed sales that look excellent. The second platform does too. When someone adds the two together, the attributed sales come out clearly higher than what the store sold in total. In this scenario, say one and a half times higher, because gaps like that happen.
The conversation starts. Marketing says the campaigns worked. Finance says there is less in the bank than the reports suggest. Someone proposes trusting the store, someone else the platforms, "because they delivered the traffic". The budget decision for the final days of the season, and for January, gets made on whichever number someone happens to remember better.
Nobody lied or made a mistake. Each report counted what it was built to count. Nobody had agreed in advance which one decides the budget.
Where the gap comes from: five reasons
1. Each platform counts the sale for itself. If a customer saw an ad in one place, clicked in another and bought, both platforms may claim that transaction. The store counts it once. This is the most common reason the total is larger than actual sales.
2. Attribution windows differ. A platform credits a sale to an ad if the purchase happened within a set time after a click or a view. The length of that window depends on the platform and the account settings. The same ad "earns" more with a longer window and less with a shorter one, though customers bought exactly the same.
3. The click date is not the order date. A platform often assigns the sale to the day of the click, while the store counts it on the day of the order. In a season, when people click on Thursday and buy on Monday, daily reports diverge sharply, even if over the whole period the difference is smaller.
4. The store counts after returns, the platform often before. An ad report usually shows the basket value at the moment of purchase. Returns, cancellations, discounts and shipping costs show up later, in store or accounting data. After Black Friday there are more returns than in a normal week, so this reason matters more in season.
5. Every source has blind spots. A customer who did not consent to tracking does not exist in the platform report, but does exist in the store. And some sales no platform will ever see: referrals, direct returns to the site, purchases several days after the first contact. So platforms overstate (reasons 1 and 2) and understate (reason 5) at the same time. The net result is neither "too high" nor "too low". It is simply different.
Which number to use for which decision
"Which number do I believe?" is the wrong question. The better one is "which decision do I need this number for?"
| Decision | The number you use | What you do not do |
|---|---|---|
| How much the company earned and whether the season was profitable | Revenue and margin from the store and accounting, after returns | Do not add up platform results |
| Which campaign or channel performs better | Results from one platform, compared with other campaigns on the same platform, in the same window | Do not compare two platforms directly or add them together |
| How much budget to shift between channels | The change in store results against the change in ad spend, plus the direction shown by the platforms | Do not move budget based on a single day |
| Whether platforms overstate sales | Total sales attributed by platforms divided by store revenue, tracked week to week | Do not judge the value itself, only the trend |
The fourth row is the least obvious. That ratio will not tell you how much the advertising really earned, but it will tell you whether the gap is growing or shrinking. If it is growing, someone changed settings, more platforms are overlapping, or something stopped being measured.
Test your own gap on an ordinary week, before the season starts
You do not have to wait for Black Friday to see how big the gap is. Take the last ordinary week and do the exercise on a sheet of paper or in one spreadsheet.
Write down three numbers: store revenue after returns, sales attributed by the first platform, and sales attributed by the second. Divide the platform total by store revenue. That ratio is your baseline. In season you check it again and see whether the gap has grown.
Then go through the five reasons from the previous section and answer each one yes, no or I don't know. Can the same sale be counted by two platforms? Do you know the attribution windows of both? Do you know how the store report treats returns and discounts? Has anyone checked what share of traffic is covered by tracking consent? Are you adding new traffic sources this season that have had no report until now?
If most answers are "I don't know", you have information worth more than the result of the exercise. It means that in season the leadership team will receive numbers whose origin nobody at the table can explain. That is not marketing's fault or finance's fault. It is what a company looks like when it has grown faster than its way of working with data: every department has its own report and its own definition of a sale, and nobody is responsible for making those definitions meet.
If most answers are "yes" and the gap is still large, that is news too. Then you look for the cause somewhere you have not thought of yet, for example in how orders from different sources end up in one summary.
Either way you know more than before the exercise. It does not replace a diagnosis, but it helps you decide whether you need one now.
Four rules to set before 27 November
This is a leadership decision, not an analyst's. An analyst can do the calculation, but the rules of the game must be approved by someone who answers for the result.
- One revenue number. You write down where it comes from (the store admin, the accounting system) and who owns it. It goes into every season summary.
- One comparison window for the platforms. You pick one and keep it for the whole season, so that campaign comparisons mean something.
- A decision threshold set in advance. Before the season starts, you define what has to happen for you to move budget. For example: a channel shows a clearly weaker result in the same window for at least two days in a row. Without this, decisions get made under the pressure of one bad evening.
- One person settles disputes. When reports disagree, it is clear who decides and by what time. In season, an hour of argument is an hour in which the budget keeps working to the old decision.
The gap between your store and your platforms is not a bug to fix. It is a difference in definition that has to be written down before anyone spends the budget.
What it costs when you don't have these rules
The first cost is time: the hours in which marketing and finance agree on which number applies. In season those hours are the most expensive.
The second cost is decisions made on whichever number happens to fit. I wrote about this in more detail in why a board that gets three different numbers picks the one it likes. In season it translates into budget burned in a channel that only looks good, or a channel cut that was actually working.
Add to that team fatigue. When the same argument about whose number is the real one returns every season, people stop trusting reports, even the correct ones. They start counting their own way, in their own spreadsheets, and the gap you wanted to close grows by another version of the truth.
The third cost only shows up in January. Without an agreed definition, nobody can calculate whether the season paid off, so the assessment depends on who is presenting. If you want to see how a single reporting error can drag a budget down, the reporting error that burned a marketing budget shows it on a typical scenario. It is one of the places where EBITDA leaks out before anyone notices.
A nice dashboard will not solve this either. A dashboard is a snapshot of the past, not control over your data. It will show the gap just as clearly, but it will not tell you where it comes from.
When setting rules is not enough
The rules in this article work when the gap can be explained by the five reasons. If in your company the difference is large, changes from day to day with no clear cause, or nobody can say which orders are counted twice, then this is not a question of definitions. It is Data Debt: data that accumulated without one owner and one logic, where every further decision pays interest on it.
That is where a diagnosis begins. It is not about building something right away. It is about seeing exactly where the numbers diverge, what it costs, and what to fix first, so that the leadership team looks at the same numbers.
There are only a few days until the season. If you want to check whether your gap fits the five reasons in this article or needs a different look, book a 15-min call.