Did the promotion make money? How retailers evaluate a promotion after it runs
A promotion made money only if the margin it added is more than the margin it gave away, once pull-forward, cannibalisation and the discount on sales that would have happened anyway are counted.
Start from a baseline of what would have sold without the promotion, keep only the extra units it really added, and turn every part into margin. Judged on promoted units alone, almost every promotion looks like a success.
Steps / 01
Step by step
Set the baseline
Estimate what the line would have sold in the promotion week at its normal price, usually from recent weeks with no promotion running, adjusted for season. Every later step is measured against it. In the illustrative pasta-sauce example below, the baseline is 1,000 units a week.
Count the promo-week units
Take the units sold during the promotion and subtract the baseline. The sauce sold 3,000 at half price, 2,000 more than its baseline. That lift is where most promotion reports stop, and it is only the starting point.
Subtract pull-forward
Look at the weeks after the promotion. A dip below the baseline usually means shoppers stocked up and bought early what they would have bought anyway. The sauce sold 600 the week after, not 1,000, so about 400 of the extra units were pulled forward.
Subtract units switched from other lines
Check the lines a shopper could have bought instead. A fall in their sales during the promotion is cannibalisation: units that moved from one line to another without adding to the category. A rival sauce the retailer stocks fell by 500 units, which leaves 1,100 incremental units out of the 2,000 extra.
Count halo only if it can be measured
Halo is extra sales of other lines bought alongside the promoted one. Count it only where you can measure those lines against their own baseline for the same weeks; an assumed halo can turn a loss into a profit on paper. The example leaves halo out.
Turn the units into margin
Price each part at the margin it actually earned or lost. Every unit sold in the promotion week earns the promoted margin, including the baseline units that would have sold at full price anyway, so the margin given away on them counts against the promotion. Pulled-forward units lose the full-price margin they would have earned later, and switched units lose the margin the other line would have made. In the example, the promotion sold 2,000 extra units and still cost $3,950 of margin against doing nothing.
Add supplier funding, if any
If a supplier paid towards the promotion, add what it paid before judging the result. Funding that covers the margin given away can change the answer for the line. Still check the whole category, because a funded line can take sales from own label and other lines. The example has no supplier funding.
Decide whether to repeat, change or drop the promotion
Repeat a promotion that added margin. Change one that lost it for a reason you can fix: a shallower discount gives away less on units that would have sold anyway, the same trade-off as a markdown (try different cuts in the markdown calculator), and a gross-profit floor catches a promoted price whose margin falls too far. Drop one that only moved demand between weeks and lines.
Examples / 02
Illustrative examples
The business and figures are illustrative. None of it is customer data.
A half-price pasta sauce, from units to margin
| Pulled forward from next week | 400 |
|---|---|
| Switched from a rival sauce | 500 |
| Incremental | 1,100 |
Just over half the extra volume is new, and all of it sold at half price.
Read the example as text
A pasta sauce normally sells 1,000 units a week. During a half-price week it sells 3,000. The following week it sells 600 as shoppers use up their stock, and a rival sauce the retailer stocks drops by 500 units during the promotion. Of the 2,000 extra units, about 400 were pulled forward and 500 switched from the rival, leaving 1,100 incremental units, all sold at half price.
Did the promotion make money?
Add an illustrative price and cost: the sauce sells for $5.00 and costs $2.00, so it makes $3.00 a unit at full price and $0.50 at half price ($2.50). The rival sauce makes $2.50 a unit.
| Against a normal week | Margin |
|---|---|
| Promo week: 3,000 × $0.50 = $1,500, not 1,000 × $3.00 = $3,000 | −$1,500 |
| Week after: 600 sold, not 1,000, so 400 × $3.00 lost to pull-forward | −$1,200 |
| Rival sauce: 500 units lost to cannibalisation × $2.50 | −$1,250 |
| Net margin against doing nothing | −$3,950 |
Counted the other way: the 1,100 incremental units earn 1,100 × $0.50 = +$550; the 1,000 baseline units give away $2.50 each, −$2,500; the 400 pulled-forward units earn $0.50, not $3.00 next week, −$1,000; the 500 switched units earn $0.50, not the rival's $2.50, −$1,000. Same −$3,950: 2,000 extra units sold, and margin lost.
Halo, extra sales of other lines bought alongside the sauce, counts only where it can be measured; this example leaves it out.
A shallower discount, which gives away less on units that would have sold anyway (the markdown trade-off), or supplier funding that covers the margin given away would change the answer; at half price the sauce makes 20% gross margin, the kind of margin a gross-profit floor is there to catch.
In twelfth / 03
Where twelfth fits
twelfth is a commercial workspace for retail and category teams that joins pricing, range, promotion, inventory and supplier signals into ranked, owned decisions. For merch analysts it reads price, promo, range and space off one line, with promotion results back before the next deal sheet is committed.
