The Long-Run Promotion Effect

What it is

Two papers carry it. The first, led by Carl Mela in the Journal of Marketing Research in 1997, used eight and a quarter years of panel data on a frequently purchased packaged good. It asked whether consumers' responses to marketing variables such as price change over time, and whether the changes follow shifts in advertising and in retailers' promotion. They did. Consumers grew more price and promotion sensitive over time, and the authors linked that to reduced advertising and increased promotions.

The second, led by Kamel Jedidi in Marketing Science in 1999, built a varying-parameter model that lets consumers' short-term response change as a brand's marketing changes, and that allows for competitors' reactions. It was estimated on more than eight years of panel data from a consumer packaged-goods category. Manufacturers had been moving money from advertising to promotion, and the question was what that did to profit once the long run was counted.

In effect

The 1999 paper reports several results at once. Advertising had a positive long-term effect on brand equity, while promotions had a negative one. Price promotion elasticities were larger than regular price elasticities in the short term and smaller once long-term effects were included. Most of a price cut's short-term effect appears in brand choice, a pattern that no longer held when long-term effects were counted. Overall the long-term effect of promotions on sales was negative, about two-fifths the size of their positive short-term effect.

The profit simulations, run on reasonable cost and margin assumptions, add the part a manager would act on. Regular price decreases generally reduced long-term brand profits, advertising was profitable for two of the brands, and increases in price promotions were unprofitable for all of the brands modelled.

The article Why Luxury Brands Hold Their Prices uses this as the second part of the house's case: customers learn the calendar. It is an argument by analogy. Nothing in either paper is about a handbag.

What it does not say

It does not say luxury goods behave like groceries. Both studies concern things bought often, in categories where the shopper has many repeat purchases to learn from. A luxury purchase is rarer, and the house's customers may learn from other cues.

It does not say a brand should never promote. The model estimates an average over the brands in one category, and its own advertising result was profitable for two of the brands, not all.

It does not say the effect is large for any one firm. The two-fifths figure is a ratio inside one model, not a forecast.

Both papers were read here at the level of their published abstracts and summaries, not in full.

How the two papers fit together

The two papers answer different questions, and they do so at different distances from the shopper. The 1997 study asks whether consumers' sensitivity to price and promotion changes over time, and finds that it does. It uses a two-stage approach to separate medium-term from long-term effects. The 1999 study asks what that drift does to profit. Its model lets short-term response change as a brand's marketing changes, and lets competitors react, which is why it can speak to the long run at all.

Read together, they give a chain of claims. Brands moved money from advertising to promotion. Consumers grew more sensitive to price and promotion. The brand's equity weakened and the long-run effect on sales turned negative. The chain is plausible, but each link was estimated on one category of packaged goods, and the page records the chain without extending it.

Why the two-fifths figure needs care

The figure is a ratio inside one model: the long-run effect of promotions on sales was negative and about two-fifths the size of the short-run gain. A reader can mistake it for a rule of thumb that two-fifths of any promotion's gain is lost later. It is not that. It is the model's estimate for the brands in the category it was fitted to.

The profit results are easier to carry across. Regular price decreases generally reduced long-term brand profits, and increases in price promotions were unprofitable for all the brands modelled. Those are conclusions about the direction of effects within the sample, and the authors ran them on reasonable assumptions about costs and margins.

Why a luxury house cites it

The attraction for a luxury house is that the finding comes from real purchase records over years, which is rare. Most arguments about markdowns and prestige rest on theory or on laboratory ratings. Here is a result in which promotion looked good in the short run and cost something later.

The weakness is the one the article names. The category was packaged goods, where shoppers buy often and learn the calendar of promotions from many repeat purchases. A luxury buyer purchases rarely and may take cues from elsewhere. The concept therefore supports the house's case as an analogy and nothing stronger, and the article says so.

---