What Does a Promotion Really Earn? Trade Return and Trade Efficiency

Trade return is incremental sales minus incremental cost, where incremental cost is the trade money spent plus the cost of goods on the extra units. Trade efficiency is the return per dollar of trade invested; above $1.00 the event paid for itself. In Nielsen's 2015 Trade Promotion Landscape Analysis, most events did not.

Lift counts volume. Return counts money. The gap between them is where the buyer's second question ("what did it cost you?") lives, and it is the layer most recap decks never reach.

The framework, step by step

Nielsen published the calculation it used to benchmark 76 million event weeks, and it is the cleanest public statement of the return math [1]:

1. Isolate events: identify promotion weeks by depth of discount against everyday price.
2. Incremental sales = total sales minus baseline.
3. Incremental cost = direct trade expense + incremental cost of goods.
4. Trade return = incremental sales minus incremental cost.
5. Trade efficiency = trade return per dollar invested.

NielsenIQ's dictionary gives the reading scale for that last number: above $1 the investment breaks even, and the higher above $1 the more efficient the promotion; between $0 and $1 the investment drove lift but did not break even; a negative value means the promotion drove no lift at all [2]. Because that scale puts break-even at $1.00, the practical arithmetic for efficiency is incremental margin (incremental sales less incremental cost of goods) divided by trade dollars spent. That is the convention this guide uses, and it is the one that makes the dictionary's bands and Nielsen's benchmark charts, which draw break-even at $1.00, agree.

Steps 1 and 2 are Chapter 4. Step 3 needs the cost file from Chapter 3. The framework is short. The discipline is in feeding it real inputs.

The retail-dollar trap

The syndicated file reports incremental sales in retail dollars, at the shelf price the shopper paid. The brand did not earn those dollars. It earned its own price to the retailer, less cost of goods, on the incremental units. Comparing retail incremental dollars to trade cost overstates return by the retailer's margin and then some. Convert to manufacturer terms before any comparison: incremental units times your net price, less incremental units times your unit cost. This is a practitioner rule rather than a provider definition, but every step of Nielsen's framework assumes it, which is why incremental cost of goods appears in step 3 [1].

The worked example, continued

Chapter 4 left us with an illustrative feature-and-display event at one retailer: 3,000 units scanned over two weeks at a $3.00 promoted price against a $4.00 regular price, of which 1,000 units were incremental and 2,000 were subsidized. The lift was 50%.

Now add the brand's side, with round numbers invented for the example. The brand's net price to the retailer is $2.40 a unit and its cost of goods is $1.20 a unit. The deal was a scan-back paying $0.85 on every unit scanned at the promoted price, so the brand funded 85% of the $1.00 discount, inside the 80% to 90% range Nielsen used as its industry assumption [1]. The retailer also billed a $500 display fee.

Written out:

• Incremental sales, manufacturer terms: 1,000 units at $2.40 = $2,400.
• Incremental cost of goods: 1,000 units at $1.20 = $1,200.
• Incremental margin: $1,200.
• Direct trade expense: 3,000 units at $0.85 = $2,550, plus the $500 display fee = $3,050.
• Trade return: $1,200 minus $3,050 = a loss of $1,850.
• Trade efficiency: $1,200 / $3,050 = $0.39 per dollar invested. The event drove lift and did not break even.

The lift was real. Every one of the 3,000 scanned units drew the $0.85 payment, and 2,000 of them were units the brand would have sold at full price anyway. The subsidized volume from Chapter 4 did the damage, and it is the number a lift-only recap never shows.

Change one input and the story flips. If the same event had run as a shallower deal at $3.50 with the brand funding $0.40 a unit, and lift held at even 35% (700 incremental units on 2,700 scanned), trade expense falls to $1,580 and incremental margin of $840 gives an efficiency of $0.53; still short, but the direction is the point. Depth and funding rate move return faster than lift moves it. Chapter 7 takes that up.

The break-even record, dated

Nielsen's 2015 analysis, covering 331 categories and 76 million event weeks across the US, Canada, the UK, Spain, France, Germany, and Italy, found that 59% of promotion events did not break even. By country the US was the worst at 71%; France and the UK were at 58%; Germany, Canada, and Italy were in the low 40s [1]. Across US food, drug, mass, convenience, and dollar channels, 63% of sales, spread over more than 200 categories, did not break even on trade, and the analysis found no correlation between promotion frequency and trade ROI [1].

The department view, from Nielsen's Q3 2014 benchmark database, ranged from dairy at 75% of weeks failing to break even and produce at 74% down to beauty care and personal care at 50% [1]. In the five large European markets, the best-in-class top 10% of events returned 7 times what the bottom 10% returned [1]. Nielsen's global manufacturer trade efficiency line sat just under $1.00 per dollar across 2012 to 2014 while time on promotion rose [1].

Two cautions travel with every one of those figures. They are eleven years old, and the analysis relied on the 80% to 90% funding assumption rather than actual cost data. Use them as the shape of the problem, never as your number.

Why lift looks good while ROI looks bad

When the two layers disagree, the cause is almost always one of four things, and each has a place in the framework.

Symptom	Where it hides	What it does to return
Deep discount the brand funded	Direct trade expense	Every scanned unit, incremental or not, draws the payment
High subsidized share	Incremental versus promoted volume	You paid the discount on volume you already had
Low pass-through	The gap between what you funded and the shelf price the data shows	The shopper saw less discount than you paid for, so lift was smaller per dollar
Cost of goods on the extra units	Incremental cost	Volume that costs $1.20 to make earns less than its retail dollars suggest

The pass-through study adds a useful split. Among manufacturer deals that lost money under an offer-it-to-everyone strategy, the authors attributed 63% to inelastic demand and cost conditions and 37% to insufficient pass-through by the retailer [4]. Both are diagnosable from your own data: elasticity from the lift ladder, pass-through from the gap between funded and observed discount.

And keep the negative case in view. NielsenIQ's dictionary notes that incremental dollar lift can be negative when the extra units do not cover the price reduction [3]. A recap that clips at zero hides the events that lost the most.

The cost input decides whether a return exists

A return number is only as good as the cost behind it. The practitioner standard is simple to state: when the trade cost for an event is not in the data, report lift and say the return cannot be computed. Do not substitute a funding-rate assumption and present the result as a return. If you must estimate, label the estimate, name the assumed rate, and show the range.

Sous holds that standard by default. Its stop-and-ask behavior on missing measures applies to promo return the same way it applies to a renamed column on a data drop: if the cost input is absent, the workbook reports the lift table and states that return is not computable from the loaded data, rather than inventing a discount-funding rate. When the cost file arrives, the return gets computed. Until then the workbook says so.

Common questions

How do you calculate trade promotion ROI? Incremental sales minus baseline gives incremental sales; add direct trade expense to incremental cost of goods for incremental cost; trade return is the difference; trade efficiency is incremental margin per trade dollar. Convert to manufacturer dollars first.

What is trade efficiency? Return per dollar of trade invested. Per NielsenIQ, above $1 breaks even, $0 to $1 drove lift without breaking even, and negative means no lift.

What percentage of promotions break even? In Nielsen's 2015 analysis, 41% globally and 29% in the US. No newer public benchmark of that scale exists.

Why is my lift 50% but my ROI negative? Because you paid the discount on every scanned unit, including the ones you would have sold anyway, and the incremental units carry cost of goods. See the worked example.

Can I use retail dollars for ROI? No. Convert incremental units to your net price and subtract your unit cost. Retail dollars include the retailer's margin.

What if I do not know the trade cost? Report lift and state that return cannot be computed. If you estimate, label the funding-rate assumption and show a range.