A cashback portal might offer several percent at a clothing retailer and a fraction of one percent at an electronics chain. The spread reflects what each merchant can afford to pay.
Cashback is paid out of a marketing budget
The money comes from a commission the merchant pays for a referred sale, and the portal returns part of that commission to the shopper while keeping the rest.
That commission is booked as customer acquisition spending. It sits alongside search advertising and email marketing, and it competes with them for the same budget.
A merchant therefore sets the rate at whatever it can pay for an incremental sale while still profiting from it, which anchors everything to the margin on the goods.
Gross margin sets the ceiling
Apparel, cosmetics and home decor carry wide gross margins, so a mid single-digit commission still leaves the merchant well ahead on a referred order.
Consumer electronics runs on thin margins, often only a few points on a major appliance or television. A comparable commission would erase the profit on the sale entirely.
Groceries and gift cards sit lower still, which is why they frequently appear as excluded categories rather than as low-rate ones. There is no room to pay at all.
Attribution quality changes what a referral is worth
Merchants know that many portal users would have bought anyway, having decided on the purchase and then routed through the portal for the rebate at the last moment.
Where that behavior is common, the merchant is paying for sales it already had. Rates fall in categories with strong direct demand and rise where the referral genuinely originates the purchase.
This is why rates on a merchant's own branded searches are often reduced or disallowed, and why rates climb during periods when the merchant needs new customers.
Rates move with the merchandising calendar
Commission rates are not fixed. Merchants raise them when inventory needs to clear and cut them during peak periods when demand arrives without incentive.
A rate that doubles for a weekend usually signals an inventory position rather than generosity, and a rate that collapses in late November signals a merchant that expects traffic regardless.
Portals negotiate these changes continuously, which is why a rate seen one week is not a reliable guide to what will apply the next.
The published rate is a maximum, not a promise
Most programs state their headline figure as an up-to rate, with lower tiers applying to specific departments, sale merchandise or third-party sellers within the merchant's site.
Exclusions cluster where margin is thin, so the departments a shopper is most likely to be buying from are often the ones carrying the reduced rate.
Reading the category table beneath the headline is the only way to know which rate a given basket will actually earn.