Why your pending commission keeps disappearing

You check your dashboard, see $840 pending, plan around it, and three weeks later it is $610. Nobody emailed you. Nothing is broken.

This is a reversal, it is normal, and the reason it comes as a shock is that most networks do a poor job of explaining it before it happens rather than after.

What actually causes a reversal

The order was returned. By far the most common. You earned commission on a sale; the customer sent the item back; the sale no longer exists.

The order was cancelled before dispatch. Same effect, faster.

A different code was applied at checkout. The customer left, found a voucher on a deal site, came back and used it. Under last-click attribution the deal site now owns the sale and you do not.

The traffic breached program terms. Trademark bidding, incentivized traffic, a placement the advertiser prohibits. This one is different from the others, see below.

A chargeback. Rare, later, and effectively invisible until it lands.

Cause	Reversed by	In your control
Return	Advertiser	Partly
Cancelation before dispatch	Advertiser	No
Voucher code at checkout	Attribution	Partly
Terms breach	Network	Yes
Chargeback	Payment processor	No

Why fashion is worse than everything else

Return rates drive reversal rates, and clothing has the highest return rate in retail. People order three sizes intending to keep one. That is a normal, expected shopping behavior, and it means a meaningful share of your fashion commission was never going to survive validation.

Reversal rates of 15% to 30% are typical in fashion and footwear. That is a property of the category, not a sign that anything has gone wrong, and any network implying otherwise is setting you up to be disappointed.

ASOS shows the widest gap in our directory between clicked revenue and validated revenue for exactly this reason. It is still a good program. You just cannot plan against the pending number.

Travel behaves differently but ends up in a similar place: Booking.com validates on completed stay rather than on booking, so a reservation made in March for a September trip is not really pending, it is provisional for six months. Hotels.com pays on the same basis, and its seven-day cookie window is a separate and unrelated constraint that people routinely confuse with this one.

Days to validation, retail

30 to 60Days to validation, retail

Days to validation, travel

60 to 90Days to validation, travel

Typical fashion reversal range

15% to 30%Typical fashion reversal range

Reversals cost most where the baskets are biggest

Reversal rates are quoted as percentages, which flattens something important: the same percentage costs wildly different amounts depending on the program.

Take the middle of the published fashion range, call it a quarter of sales reversing, and apply it to three programs in the directory. ASOS pays up to 6% on an $88 average order, so about $5.28 a sale, and a quarter of that is $1.32 lost per sale. Revolve pays up to 8% on a $310 order, about $24.80, so the same rate costs $6.20. Farfetch pays up to 7% on $540, about $37.80, and a quarter is $9.45.

Those figures are an illustration built on a published range rather than a measurement, so treat the shape rather than the digits. The shape is that premium and luxury fashion carries seven times the per-sale reversal exposure of high-street fashion at the same reversal rate, and higher return rates on top of it, because expensive clothing is bought on approval more often than cheap clothing is.

This does not argue against premium programs. Their net numbers are still better, which is the whole point of promoting them. It argues against reading a pending balance from a luxury program the same way you read one from a high-street program. The first will shrink by considerably more in absolute terms, and if you have planned around it, that is the month it hurts.

The practical fix

Plan against validated earnings, not pending.

Our dashboard reports validated by default for this reason. Pending is still visible and clearly labelled, but it is not the number we put in front of you first, because a figure that reliably shrinks is a bad basis for a decision.

If you are working out whether a program is worth your time, wait for one full validation cycle, typically 30 to 60 days in retail, and judge it on what actually cleared. Judging a program on week-one pending figures will make almost every program look better than it is.

Putting a number on your own reversal rate

Category averages are a starting point and a poor substitute for your own figure, which will differ from the average because of what you promote and how you describe it.

The calculation is one division and it needs one full quarter of history. For a single program, take the commission credited in a settled period and the commission that eventually validated from it, then divide the second by the first. That ratio is your survival rate on that program. One minus it is your reversal rate.

Two things make the number trustworthy. Do it per program rather than across the account, because a fashion program and a tech program will produce very different ratios and the blended figure describes neither. And only use periods that have fully cleared their hold, because a period that is still validating will make your reversal rate look better than it is, in precisely the way this whole post is about.

Once you have it, forecasting is straightforward. Multiply today's pending by the survival rate and plan against that, not against the headline. A creator who knows their ASOS traffic validates at around three quarters is in a completely different position from one who is surprised every month by the same thing, and the only difference between them is one quarter of patience and one division.

Recalculate it every few months. Reversal rates move with the seasons, and sale periods are the worst of them: heavily discounted stock is returned more often and it is also when your content performs best, which is an unhelpful combination and a predictable one.

When a clawback pattern means something

Ordinary reversals are scattered: a return here, a cancellation there, spread across programs and weeks.

Worth investigating:

• Concentrated in one traffic source. If one platform or one campaign generates most of your reversals, the traffic quality from that source is different in a way worth understanding.
• Concentrated in one piece of content. Usually means the content oversold. If a video promises something the product does not deliver, people buy and return, and the reversal rate is the audience telling you.
• A sudden rate change with no volume change. Often an advertiser tightening terms, sometimes a tracking problem.

We raise this with partners before it becomes a reason to remove access, because on a sub-affiliate network the upstream network sees one publisher account. A pattern that would be a private problem on a direct relationship is a shared problem here.

What reversals are not

They are not a network keeping your money. The commission was never paid to us either, the advertiser reversed it upstream and it disappeared from our balance the same way it disappeared from yours.

They are also not a reason to avoid categories with high return rates. Fashion has excellent economics despite reversals, because the volume more than compensates. You simply have to price the reversal rate into your expectations rather than discovering it at the end of the month.

Full detail on validation timing and hold periods is on the payouts page.