Sign in

Blog · Territory, capacity and quota planning · FMCG and CPG brands

Net lift on one promotion at one account: the whole arithmetic on one page

The complete promotional net lift calculation on one SKU at one retail account over sixteen weeks, small enough to check by hand: the eight pre-promotion weeks and their median as the baseline, the two promotion weeks and the gross lift, the four post-promotion weeks and the dip, net lift, trade spend, cost per net unit, the same mechanic at a second account for comparison, and the assertion that the account's weekly sell-out sums to the retailer's total, so a reader can reproduce every figure and then run it on their own sell-out files.

The short answerOne SKU at one account. Eight pre-promotion weeks sell a median of 4,100 units, the baseline. Two promotion weeks sell 9,100 and 9,000: gross lift 9,900 over baseline. Four post-promotion weeks sell 3,600, 3,700, 3,900 and 4,000: a dip of 1,200 below baseline. Net lift is 8,700 units for $21,000 of trade spend, $2.41 per net unit. The same mechanic at a second account has a gross lift of 14,100 and a dip of 12,900: net 1,200 at $28 per unit, volume moved forward. Weekly sell-out per account sums to the retailer's total for the SKU. Every number can be reproduced by hand.

Net lift is a baseline, a gross lift and a dip, and on sixteen weeks of one SKU at one account every step can be checked. This page works the baseline, the lift, the dip, the net, the cost per unit, the comparison account and the assertion.

Account R-1, sixteen weeks

Week Sell-out Window
1 4,200 Pre
2 3,900 Pre
3 4,100 Pre
4 4,400 Pre
5 4,000 Pre
6 4,100 Pre
7 3,800 Pre
8 4,300 Pre
9 9,100 Promotion
10 9,000 Promotion
11 3,600 Post
12 3,700 Post
13 3,900 Post
14 4,000 Post
15 4,100 Back at baseline
16 4,200

The baseline

Baseline = median of weeks 1 to 8 = median(3,800, 3,900, 4,000, 4,100, 4,100, 4,200, 4,300, 4,400) = 4,100

Gross lift, dip, net lift

Gross lift = Σ (promo week − baseline) = (9,100 − 4,100) + (9,000 − 4,100) = 5,000 + 4,900 = 9,900 Dip = Σ (baseline − post week), floored at zero = 500 + 400 + 200 + 100 = 1,200 Net lift = 9,900 − 1,200 = 8,700 units

Cost per net unit

Trade spend on this promotion at R-1: $21,000.

Cost per net unit = 21,000 ÷ 8,700 = $2.41

The comparison account, R-2, same mechanic

Element R-2
Baseline 6,300
Promo weeks 13,400 and 13,300: gross lift 14,100
Post weeks 2,900, 3,100, 3,600, 4,100: dip 12,900
Net lift 1,200
Trade spend $34,000
Cost per net unit $28.33

R-2's shoppers bought a month's supply in two weeks and stayed home. On shipments, R-2 was the better promotion.

The assertion

Σ accounts' weekly sell-out for the SKU = retailer's total per week

Week 9: R-1 9,100 + R-2 13,400 + others = the retailer's reported total for the SKU. Holds each week, or a missing account file is listed and that account's promotion is marked unmeasured.

Where it goes wrong, even on sixteen weeks

Measured on shipments. R-2 shipped the most; the sell-in spike is the retailer stocking.

No post-window. R-2's net lift reads 14,100 and it gets next year's spend.

Baseline including a prior promotion. If week 3 had been a promotion at 8,000, the median rises to 4,150 and the lift understates.

Mean baseline. One odd week moves it; the median does not.

From one promotion to a season

The same baseline, windows and arithmetic per account per promotion, from the weekly sell-out files and the calendar. Covirage runs it after every promotion. The promo lift guide covers the measure, and the weeks of cover guide covers what the promotion weeks look like in channel inventory.

Questions people ask

Why the median of eight weeks?

Because one high or low week in the pre-window would set a mean baseline the account does not normally sell at. The median is the account's typical week, and the pre-window excludes any earlier promotion.

How long is the post-window?

Until sell-out returns to baseline, capped at a stated number of weeks, here four. The dip is the shortfall below baseline in that window. An account still below baseline at the cap is flagged: the promotion may have pulled forward more than four weeks of demand.

Why compare two accounts?

Because the same mechanic produced a real lift at one and a pull-forward at the other, and trade spend next year goes where the net lift is. The second account's shipments during the promotion looked like the better result.