Trade Promotion ROI: Real Uplift Versus Borrowed Volume

Most trade promotion reviews in India end the same way. Someone puts up a chart of sell-out during the promo window, points at the spike, and calls it a win. The spend gets renewed. Nobody asks the only question that matters: would some of that volume have happened anyway.

Trade promotion is usually the second or third largest line in a consumer brand’s P and L, behind cost of goods and often ahead of media. It is also the least audited. That combination is expensive.

The four things people call a trade promotion

Lumping them together is the first mistake, because they behave differently and they need different measurement.

Price off. A temporary reduction in the shelf price, funded by you, executed by the retailer. Fastest response, weakest memory. Shoppers who buy on price off are the most likely to be buying ahead rather than buying more.

Bundled or multipack. Two for one, a larger pack at the single pack price, a combo. This is a volume commitment device. It fills the consumer’s pantry, which is exactly what makes it dangerous. A pantry that is full in week two does not get refilled in week four.

Consumer offer. Cashback, a gift with purchase, a contest, a scratch code. Slower to move volume, but the response is usually more honest, because the shopper is buying for a reason other than stockpiling a discount.

Retailer incentive. Slotting money, a display fee, a slab-based margin kicker, a target bonus to the distributor. Here the money never reaches the shopper at all. It buys attention, space or push. Whether it buys volume is a separate question, and one that has to be tested separately.

The baseline is the whole argument

Uplift means nothing without a baseline. Baseline is your best estimate of what would have sold in those stores, in that window, if you had done nothing.

The common lazy baseline is the four weeks before the promo. That is wrong more often than it is right. It ignores seasonality, it ignores the festive calendar, it ignores whether a competitor was on promo before you, and it ignores that you may have pushed extra stock in to prepare for the promo, which lifts the pre-period.

A defensible baseline needs at least two of these. First, a longer history, so the shape of a normal year is visible. Second, control stores or control cities that did not run the promotion, matched on format, throughput and geography. Third, an explicit adjustment for known events, the obvious ones in India being Diwali, Holi, the school reopening cycle, and monsoon onset in the categories that swing with weather.

If you cannot build a control group, say so and lower your confidence. Do not pretend a pre-period is a control.

The post-promotion dip is the tell

Run any price off and plot the eight weeks after. If sales fall below baseline for a period and then recover, you did not create demand. You moved it. The shopper who was going to buy in week six bought in week two, and week six is now empty.

The size and length of the dip varies a lot by category. Stockable, low-perishability goods with long consumption cycles show the deepest dips. Detergent, hair oil, staples and packaged snacks in large packs are classic. Fresh, short-shelf-life and impulse categories show a much shallower dip because nobody stockpiles them. Do not carry a benchmark dip percentage from one category into another. Measure yours.

The practical rule: your measurement window must extend past the dip, not stop at the end of the promo. If the window closes when the price goes back up, you have guaranteed yourself a flattering answer.

Forward buying by the trade

There is a second, quieter leak. When you announce a promotional price to distributors or modern trade buyers, a rational buyer loads up at the promo rate and sells through at the normal rate afterwards. That is forward buying, and it is invisible if you measure at primary sales.

Symptoms are easy to spot once you look. Primary sales spike hard, secondary and sell-out spike much less. Order sizes in the week the scheme is announced are far above trend. Reorders go quiet for longer than the consumer dip would explain. Stock in trade rises and stays risen.

Forward buying does not just borrow volume, it transfers margin. You funded a discount on units that would have sold at full price. Measure at sell-out, or at the very least at secondary, and track days of stock in trade before and after every scheme.

The arithmetic

Incremental volume, not total volume, is the numerator. Work it in this order.

One. Total volume in the promo window across participating stores. Two. Subtract baseline volume for the same stores and window. That gives gross uplift. Three. Subtract the shortfall against baseline in the post-promotion period. That converts gross uplift into net incremental volume. Four. Subtract any forward-bought volume that has not yet cleared to the consumer.

Then the cost side. Add the discount funded per unit across every unit sold at promo price, including the units that would have sold anyway. Add display fees, slotting, promoter cost if a promoter was deployed, and the cost of the extra service and freight. That is total promotion cost.

Divide net incremental contribution, meaning incremental units times contribution margin per unit at normal price, by total promotion cost. Below one, the promotion destroyed value. The uncomfortable part is that a large share of trade promotions in stockable categories land below one once the dip and forward buying are netted out. That is not a reason to stop promoting. It is a reason to know which ones work.

What to do with the answer

Rank promotions by incremental contribution per rupee, not by uplift percentage. Kill the bottom of the list before you fund anything new. Shift budget from depth to frequency where the dip is deep, and from frequency to depth where it is shallow. Cap scheme quantities and tighten scheme windows to blunt forward buying. And insist that every promotion is booked with a stated baseline and a stated post-period before it runs, not after.

The discipline is unglamorous. It is also the cheapest margin you will find, because you are not buying more volume, you are stopping payment on volume you already had.