What incremental sales are
Incremental sales are total sales minus baseline, the units a promotion actually added on top of what you'd have sold anyway. It is the real payoff of a promo, and it is almost always smaller than the number a brand manager celebrates. When I evaluated promotions at a natural-products brand, the gap between "we sold 50,000 units on deal" and "the deal added 18,000 units" was where most of the quarter's bad decisions got made.
Take a refrigerated-pasta SKU at Kroger that normally sells 10,000 units a week. During a four-week endcap at a feature price, weekly sales jump to 25,000. The total looks enormous. But 10,000 of every promoted week would have sold at full price regardless, so the incremental piece is the 15,000 above baseline, not the 25,000 on the tag. If you searched "promotional lift" trying to defend or kill a deal, this distinction is the entire argument.
How lift is computed
Lift is incremental units expressed against the baseline, so percent lift equals incremental divided by baseline. The arithmetic is trivial. The hard part is the baseline, which is why baseline sales is the page you read first. Every error in the baseline flows straight into the lift number, magnified.
Run a clean example on a four-week TPR for a cold-brew coffee SKU at 200 Sprouts stores. Baseline is 9 units/store/week. Promoted weeks run 24 units/store/week.
| Line | Value |
|---|---|
| Baseline units (9 × 200 × 4) | 7,200 |
| Total promoted units (24 × 200 × 4) | 19,200 |
| Incremental units | 12,000 |
| Percent lift (12,000 / 7,200) | 167% |
So the deal drove 12,000 incremental units and a 167% lift. Check it: total 19,200 minus baseline 7,200 equals 12,000, and 12,000 / 7,200 = 1.67. That 167% is the gross lift, and it is the number most decks stop at. The trouble is that gross lift overstates the win, because two leaks sit between gross incremental and the units that genuinely grew the business.
Netting out the leaks: cannibalization and forward-buy
The first leak is cannibalization. When you promote the cold-brew SKU, some shoppers who would have bought your regular-price ready-to-drink coffee switch to the deal, so part of that 12,000 "incremental" is stolen from your own shelf. If 2,000 of those units are switchers from a sister SKU, your true portfolio incremental is 10,000, not 12,000. Read cannibalization at the brand level, not the SKU level, or you'll keep crediting one product for sales it pickpocketed from another.
The second leak is forward-buy. Promotions pull demand forward: shoppers stock up at the deal price and don't return at full price for a few weeks, so baseline dips after the promo ends. If 1,500 of the 12,000 are just demand borrowed from the next month, the post-promo trough will claw it back, and counting it as net-new growth double-books the same unit. Net both and the worked example moves from a 12,000-unit headline to roughly 8,500 units of durable, true incremental.
| Adjustment | Units |
|---|---|
| Gross incremental | 12,000 |
| Less cannibalization (switchers) | -2,000 |
| Less forward-buy (pull-ahead) | -1,500 |
| Net true incremental | 8,500 |
That is a 30% haircut on the celebrated number, and it is the difference between a promo you repeat and one you quietly retire.
Why incremental is the only number that pays the bill
Incremental sales are what justify trade spend, full stop. A promotion funds a markdown on every promoted unit, including the baseline units that would have sold at full price, so the contribution from the incremental units has to cover the markdown on all of them. Gross lift makes that math look easy. Net incremental makes it honest. Pair net incremental with base velocity and you can tell whether a SKU needs more distribution, more promotion, or neither.
The full walk-through of separating base from incremental in SPINS data, including the post-promo trough, is in post-promo lift and baseline in SPINS.
Where Scout fits
Computing net incremental, after cannibalization and forward-buy, across every promo and retailer is exactly the reconciliation that eats an analyst's week. Scout connects your SPINS or retailer data, models the baseline, and surfaces gross and net incremental so the lift you report already has the leaks subtracted. It measures and analyzes the lift. It does not execute trade promotions or fund the markdown. Those decisions stay yours.
The short version
- Incremental sales are total sales minus baseline, the units a promotion genuinely added. Percent lift is incremental over baseline.
- Gross lift overstates the win. Net out cannibalization (sales stolen from your own SKUs) and forward-buy (demand borrowed from later weeks).
- Only net true incremental justifies the markdown a promo funds on every promoted unit, baseline included.