The promotion worked, did the business make money?

A promotion that lifts volume has not proved anything yet, because a good promoted week does not tell you whether the business made or lost money once everything is accounted for. Most of what a promotion costs sits outside the promoted week and outside the team that ran it, which is why Trade (Revenue Management) and Marketing so often finish the same promotion with two different answers.

If you are responsible for marketing effectiveness, this is worth sorting before your next planning round, because promotion is often one of the biggest commercial levers in the business and it is rarely measured alongside the media that supports it.

Promotion is often a bigger budget than media, and it is judged somewhere else entirely

Promotion frequently costs more than media, yet it sits on a different P&L line, goes through a different planning process and is run by a different team, so it ends up evaluated in a different spreadsheet against a different KPI to the rest of marketing. Marketing and Trade rarely align on what a promotion actually delivered, so both can claim the same sales and the credit given to promotion and media may add up to more than the business sold.

Marketing also often doesn’t see the promo calendar early enough to put the right media behind the right offer.

The real cost of a promotion lands in the weeks you were not looking at

Most promotion measurement is strongest in the promoted week, where uplift rests on a baseline that is an estimate, or on tags such as “sold on promotion” that ignore all other factors that may have driven the sale. The costs that matter sit elsewhere:

Sales you would have made anyway, now at a lower margin Sales pulled forward from the weeks after Shoppers trained to wait for the next offer Sales taken from your own range (cannibalisation)

• Sales you would have made anyway, now at a lower margin
• Sales pulled forward from the weeks after
• Shoppers trained to wait for the next offer
• Sales taken from your own range (cannibalisation)

That waiting effect is already visible in the weeks before Black Friday, where sales are getting weaker as people hold off to see what will be on promotion.

Volume is not revenue and revenue is not profit, so once you account for the lower price, the uplift is smaller than it looks. If you lower the price by 50%, you need to sell twice as many products just to hold revenue, before margin is even considered, so the depth of the discount and the media plan need to be decided together and not one after the other.

Credit turns political when media and promotion are measured apart

Fewer consumers would have known about the offer without the marketing, and the marketing may not have converted without the offer, so both are doing work, but in most businesses the loudest team wins the argument over where the credit sits. It gets harder to settle because media looks better or worse depending on the offer beside it, and if every promoted sale is tagged “sold on promotion”, the credit follows the tag, which is digital media’s last click problem by another name.

The model that decides whether the promotion made money is the one that decides whether the media worked

The usual approaches each see only part of the picture. Shopper and panel data shows stockpiling and buying behaviour but rarely connects to media, and a media-only model advises on a small fraction of what drives sales while leaving promotion in the baseline, where it gets misread as a media effect.

The answer is one marketing mix model, with price and promotion as key drivers alongside media, seasonality, distribution and external factors, and with the shopper and panel data shaping the hypotheses the model evaluates.

Once price, promotion and media sit in the same model you can see the cannibalisation of your own range, the dip in the weeks after and the evidence that shoppers have learned to wait, and each lever is read against the others and against everything else moving sales, so the credit debate stops being political and becomes an answer you can defend.

Get promotion right first, because controlling it is the prerequisite for a clean media read.

Make sure you know which mechanic was used so you can compare mechanics in the model, because a buy one get one free offer does not necessarily have the same impact as a 50% reduction, even though the percentage discount looks the same.

From there, run any change in promotion, price or media through the full P&L, so a promotion that lifts volume and loses margin becomes a decision taken with eyes open, and you can answer cross-lever questions with a number, such as whether a pound of promotion is worth a pound of media, or what happens if budget moves from one to the other.

Expect resistance, because you will be showing people numbers they are not used to seeing

Stakeholder management matters as much as the modelling here, since Trade, Marketing and Finance will each be looking at figures that differ from the ones they know. Bring Trade and Finance in from the start, build trust in the numbers before you ask anyone to act on them, and make recommendations that cover price, promotion, mechanics and marketing together, so no single team feels the finding is aimed at them.

Then make the output something each team can act on next week, because insight only creates value when it becomes action.

To summarise

A promotion has not made money until profit says so, and promotion and media can only be judged properly in one model, so there are three things to do:

Ask the team that ran your last big promotion, and the team that ran the media beside it, how many sales in that period each of them credits to their own activity, then add the two together and compare the total with what the business actually sold. If the total is higher, you have found double counting, and until the credit is agreed nobody in the business can say what the promotion or the media really delivered. Check whether price and promotion are drivers in your current model or sitting inside the baseline, because if they are in the baseline, media is taking credit or blame for what the discount did and every decision built on that read is skewed. Before your next planning round, put the promo calendar next to the media plan and agree with Trade and Finance how each promotion will be judged before it runs, so profit is the test and not something argued about afterwards.

• Ask the team that ran your last big promotion, and the team that ran the media beside it, how many sales in that period each of them credits to their own activity, then add the two together and compare the total with what the business actually sold. If the total is higher, you have found double counting, and until the credit is agreed nobody in the business can say what the promotion or the media really delivered.
• Check whether price and promotion are drivers in your current model or sitting inside the baseline, because if they are in the baseline, media is taking credit or blame for what the discount did and every decision built on that read is skewed.
• Before your next planning round, put the promo calendar next to the media plan and agree with Trade and Finance how each promotion will be judged before it runs, so profit is the test and not something argued about afterwards.

What would the two sales numbers add up to in your business? If you want to compare notes on getting promotion and media into one model, get in touch.