Guide
Why three-quarters of your promotions lose money
The scan report comes back and volume is up 40 percent. The promo gets declared a win and booked again next quarter. But most of that volume was your existing buyers stocking up at a discount, and the weeks after the promo dip below normal while their pantries empty. Nielsen found roughly three-quarters of CPG promotions do not break even. The scan report will never tell you which side of that line you are on.
Pantry-loading: the lift that is not lift
A discount pulls demand forward. Loyal buyers who would have paid full price next month buy double now at 25 percent off. The promo week looks spectacular, the following month sags, and the net effect is selling the same units for less money plus a slotting fee.
McKinsey puts money-losing US trade promotions at 59 to 72 percent. For a small brand, trade spend often consumes around 20 percent of revenue, which makes this the biggest quietly-losing line on the P&L.
Why the loss is invisible
The only number most brands see is scanned volume during the promo window. Incrementality, the units that would not have sold anyway, requires comparing against baseline velocity and watching the post-promo trough, and nobody has time to build that analysis per promo per retailer.
So the decision defaults to the retailer's narrative: volume was up, let's rebook. The brand pays for the same shelf theater again next quarter.
Measure true lift, not gross volume
The fix is a promo-ROI dashboard built from the scan and shipment exports you already receive: baseline velocity before, volume during, the dip after, and the discount cost, netted into one number per promo: this event made or lost X dollars.
With that number, the conversation changes. Rebook the two promos that genuinely recruit new buyers, renegotiate or kill the ones that only subsidize existing ones, and put the freed budget where it grows the brand.
The promos that do work
Promotions that recruit new buyers, win display placement that persists after the event, or defend against a competitor's launch can absolutely pay. The point is not that promos are bad. It is that without incrementality math, the winners and losers look identical in the scan report.
A brand that measures true lift runs fewer, better promos, and usually spends less to sell more.
Key takeaways
- Roughly three-quarters of CPG promotions never break even.
- Pantry-loading makes losing promos look like winners in scan data.
- True lift means incremental units net of the post-promo dip and discount cost.
- A per-promo profit number turns rebooking from a habit into a decision.
Know what every promo really netted
Tablebond builds promo-ROI tracking for packaged-goods brands: true incremental lift per event from your scan exports, with a made-or-lost number before you rebook. One-time setup, $1,200.
Measure my promo ROI →Questions, answered
Existing buyers stocking up during a discount on units they would have bought later at full price. It inflates promo-week volume while borrowing sales from future weeks, which is why gross volume overstates promo performance.
The scan or depletion exports you already get from retailers and distributors, plus your shipment records and the promo's discount terms. Baseline velocity, promo volume, and the post-promo trough come straight from those.
Sustained post-promo velocity above the old baseline is the fingerprint of new buyers. A deep trough that gives back the gains is the fingerprint of pantry-loading.
No. Retail relationships often require promo participation. The goal is knowing which events pay, negotiating harder on the ones that do not, and never rebooking a loser out of habit.