How Brands Grow

by Byron Sharp

  Price promotions are popular with managers because the results are easy to see. More stock is sold. If a brand is sold through retailer outlets, sales tracking immediately shows an impressive spike. The immediacy is attractive. This may be important for a company that needs to transform inventory into money. There are other related reasons – such as managers being under pressure to make unit sales budgets. Price promotions deliver these results, but they also lead to next year’s volume targets that require more price discounting. Price promotions can become a way of life for organisations because winding them back would result in failing to hit sales targets. Therefore, such promotions are highly addictive for marketers.

  Another reason for the popularity of these discounts is that distribution channels (such as retailers) demand them, because a retailer can then offer discounts to consumers and build an image of good value. Discounts at the retail level also allow a retailer to maintain competitive pressure on its rivals. Therefore, price promotions on the part of manufacturers can be one of the costs of maintaining good relationships with retail distributors. Unfortunately, in-store promotions empower the distributor because the market share of the brand becomes dependent on the distributor's decisions. Price promotions can also 'buy time' for a brand – boost sales in the short term or save it from being delisted. However, something must be done while the clock is ticking, and there is evidence that more emphasis on price promotions is accompanied by lower emphasis on innovation (Pauwels et al., 2004). Finally, it may be that price promotions are so popular because it is hard to think of more creative ideas for brand strategy.

  Price promotions don't win new customers

  Some marketers believe that price cuts attract new customers to buy a brand: customers who might later come back and make more purchases at full price. However, there is little evidence this occurs. Ehrenberg, Hammond and Goodhardt (1994) found that almost everyone who bought a brand during a price promotion had bought the brand previously. Some academic researchers have even found that some sales in a price promotion have been brought forward from future sales, so the spike in volume at the time of the sale is cancelled out to some degree because of lower sales post promotion. Fortunately, the evidence is that the main effect is brand substitution, not forward buying.

  If price promotions can't attract new buyers, can they at least bring back infrequent buyers? Perhaps the acts of purchase and use could reinforce infrequent buyers' propensity to buy again. This possibility has been investigated as the 'purchase reinforcement effect'. What has been found is that a price promotion pulls in a large proportion of infrequent buyers (i.e. buyers who have a low propensity to buy a brand). They buy during the promotion and then afterwards resume their very low buying propensity. In other words, after the promotional purchase it is as if nothing happened. Why is this the case? Let's think about consumers' purchases over a long period. They may have purchased that category dozens or hundreds of times before. Buying this brand on deal is nothing special to consumers; they have bought that brand before, just as they have bought competitor brands before, either on special or at normal price. Now let us consider what might occur before these infrequent buyers purchase from the category again. Ehrenberg's work over decades shows that the incidence of category purchasing follows a negative binomial distribution (NBD): lots of infrequent buyers and a few frequent buyers (Ehrenberg 1959; Goodhardt, Ehrenberg & Chatfield 1984). So for the majority of buyers, a lot of time may elapse before they buy from that particular category again. While the elapsed time between purchases may only be a week or two for some buyers in some categories (e.g. regular milk or bread buyers), it could be several months for instant coffee or toothpaste and six months to a year for less regularly purchased goods. When the time does come for customers to re-buy from the category in question, they may have purchased dozens of brands in a range of other categories. Some of those purchases will have been on special, some not. This means that the act of buying any one particular brand on promotion in any one particular category will easily be forgotten before the next time. Buying brands that are temporarily promoted is a routine, very frequent behaviour for consumers. There is little scope for a permanent or semi-permanent behavioural change. In summary, what price promotions do (for established brands) is to jolt the short-term buying propensities of mainly infrequent buyers who take the opportunity to buy the brand cheaply and then resume their normal purchase behaviour afterwards (i.e. to buy it sometimes as part of a wider repertoire). As Pauwels, Hanssens and Siddarth (2002, p. 437) state, “Price promotions produce only temporary benefits for established brands”.

  Negative after-effects

  Many marketers are justifiably worried about the possible negative after-effects of price promotions. Such after-effects include consumer resistance to paying normal price after purchasing at a discounted price; this is a manifestation of the 'reference price effect'. A number of studies have found reference price effects (e.g. Hardie, Johnson & Fader, 1993; Lattin & Bucklin 1989) in grocery markets. However, another stream of research on price knowledge has found that many consumers struggle to recall the prices they paid for goods (Vanhuele & Dreze 2002). How can there be a reference price effect if consumers have a low level of price recall and buying brands on special is routine? Let's tackle this paradox by reviewing the reference price research and the studies on price recall, and then reconcile the results from both research streams.

  Reference price

  The academic research on pricing emphasises the notion of a consumer reference price – especially the consumer's internal reference price, which is a memory or expectation of what prices should be (note: the concept of an external reference price refers to information on the prices of alternatives). This expectation of price is thought to be generated by exposure to prices in the past, either by purchasing or observing communications such as ads. The idea of a reference price is that 'past prices matter' and if consumers encounter a price above their reference price, this dampens their propensity to buy.

  Price knowledge

  We have to reconcile the notion of a reference price effect with consumers' less-than-perfect knowledge about prices and the routine nature of shopping described above. Many studies of consumers' ability to recall prices have shown their recall to be poor. In the US, a large-scale study (Dickson & Sawyer 1990) was conducted in which shoppers were intercepted shortly after selecting a product; of these shoppers, only about half knew the price of the item they selected to within 5%. In France, the figure was about a third of shoppers (Vanhuele & Dreze 2002). It seems that either the memory trace for the price of the item was very momentary, or that little attention was paid to the price of the item in the first place. In the Dickson and Sawyer study, many shoppers claimed they did not check the price of the item or check the prices of alternatives.

  The initial findings about consumers' price knowledge, such as the Dickson and Sawyer (1990) study, were based on price recall. More recently, researchers have argued that two other measures are more appropriate: recognition and deal spotting. Recognition studies show shoppers three prices for an item that are (for example) the normal price, 10% above the normal price and 10% below the normal price. Vanheule and Dreze (2002) applied this approach to consumers, and found that only 13% of shoppers identified the correct price. This suggests that price recognition is also rather poor. In terms of deal spotting, approximately 50% of the time shoppers appear to be able to spot a 'good deal' for a particular brand based on being shown a (discounted) price for it.

  The conclusion is that while consumers' knowledge about prices is hazy, an appreciable proportion of consumers has a fair idea about price within a zone of the actual price. Also, consumers do seem to be aware of relativities between brands (i.e. X is usually more expensive than Y), rather than absolute price levels, and that comparison between available brands is more important.

  So, will a temporary price promotion have negative after-effects for a brand? The answer is no, but repetitive
ly doing it may have negative after-effects. This is because if consumers commonly encounter low prices and 'deal' signals for a brand, their reference price for it will be lowered. Also, customers will get used to seeing price-related information for the brand, which will raise the salience of price (and/or on-deal stickers) and possibly lower the salience of other important brand attributes. US studies (Mela, Gupta & Lehmann 1997; Mela, Jedidi & Bowman 1998) found that more frequent promotions result in heightened sensitivity to price among consumers, as well as slightly longer interpurchase times and slightly higher purchase quantities. That is, consumers learn to not only buy on deal, but to also buy a little more during the price promotion, which results in less frequent purchases.

  In terms of long-term adverse effects, competitive retaliation is another factor to consider. Managers tend to over-react to each other's price reductions (Brodie, Bonfrer & Cutler 1996). More broadly, competitor response to price has been shown to be less than rational, with some firms placing more emphasis on 'beating’ the competition with low prices than on pursuing profit. Companies that do this tend to depress their profits (Armstrong & Collopy 1996; Armstrong & Green 2007). An interesting example of the destructive potential of a 'beat the competition' mentality is the price war between Greyhound and Peter Pan Trailways (Heil & Helsen 2001 p. 86). The price of a bus fare went from $25 to $9.95 to $7 to $6.95, and then Greyhound offered a $4 fare, which was lower than the price it had charged 40 years previously. This rapid, ruinous price descent took place over a short period of three weeks. In extreme cases, price wars can bankrupt multiple players. Therefore, a decision to engage in discounting, or to escalate it, must consider that competitors are likely to respond in kind, which will result in lower profits, or even severe hardship, for all players.

  How much extra volume does a price cut generate?

  Price promotions boost sales volume in the short term; but by how much? The term used to discuss this is price 'elasticity' – the percentage change in volume from a 1% change in price. For example, if a price is cut by 10% and sales volume increases by 20%, the elasticity is -2 (20/10 = 2, the minus indicates that price and volume move in opposite directions).

  Studies have found a reasonable degree of consistency in the average elasticity for temporary promotions. Danaher and Brodie (2000, p. 923) found an average price elasticity of -2.3 across 26 categories, with almost all the price changes examined being for temporary promotions. Scriven and Ehrenberg (2004) found an average elasticity of-2.6 in an exhaustive series of experiments that included both price increases and decreases. Steenkamp and colleagues (2005) examined 442 product categories and reported a mean elasticity of -4.0. Summaries of in-market studies have found an average price elasticity of-2.5 (Bijmolt, van Heerde & Pieters 2005; Tellis 1988).

  These results suggest that on average we can expect an increase in sales volume of approximately 25% from a 10% price cut. Most of the authors cited above point out that individual elasticity can vary greatly from 0 to -20 or bigger, although the majority, 70% or so, fall in the range just smaller than -1 to -4. Scriven and Ehrenberg shed light on this variation, showing that elasticity is not a fixed property of a brand, but varies systematically with the context in which a choice decision is made. They found that five factors consistently led to bigger price elasticities, which are now explained.

  Factors leading to bigger price elasticities

  1) If the brand's price moves past a local 'reference price'

  People are more likely to switch from brand A to brand B when B becomes cheaper than A (where previously, B was more expensive). Changing relative price positions is far more effective than just narrowing the price gap. Similarly, widening a price gap that already exists has less effect than changing relative positions. Also, the larger competitor brand B is, the bigger the effect. Therefore, setting the price of a brand slightly below that of the biggest brands in the category may be a better tactic than either deeply undercutting the leaders' prices, or setting the price close to the leader's price, but not undercutting it.

  2) If the price change is explicitly signalled

  Signalling the price change has a big effect. The signalling effect links to the notion that consumers have imperfect awareness of prices and their relativities. Therefore, if we draw attention to a price cut (e.g. 'was $54 now $30'), more people will be aware of the price and act on it. Also, of course, consumers like a bargain. Many studies of the effect of in-store price signals such as signage and displays can result in massive, short-term gains in sales. If a price cut is accompanied by in-store advertising, sales can increase by as much as 400% (Woodside & Waddle, 1975). In the UK, Hamilton, East and Kalafatis (1997) reported that price cuts coupled with additional promotional support can increase sales by around 200-500%. Totten and Block (1994, p. 70) reported that a deep price cut of 45%, supported by feature or display advertising, can increase sales by approximately 280% and possibly as much as 400%.

  3) If the brand's share is low

  Big brands have smaller price elasticities; small brands have bigger elasticities. This is an arithmetic effect, a result of elasticity being calculated using percentages. It is easy to imagine a brand moving from a 2% to a 6% market share (e.g. with a 50% price cut, this would require an elasticity of -4). The same elasticity and price cut on a brand with a 30% market share could enlarge the brand's share to 90%, which is almost impossible to achieve. This difference in elasticity means small brands can enjoy bigger uplifts from promotions, but suffer more than big brands when they increase prices. Conversely, bigger brands find it harder to get really big proportional sales rises from promotions, but suffer a little less than small brands when they put prices up.

  4) If the price is increased from its normal level

  It is often assumed that elasticity is the same for price rises as cuts, but there is no reason why this should be so. Empirical results from a series of experiments (Scriven & Ehrenberg 2004) show that price increases have a bigger effect on volume than cuts do – if we consider the sole effect of a price change, divorced from extra effects of signalling or in-store support.

  Larger price elasticities for price increases are consistent with a well-known finding in consumer psychology called 'loss aversion'. Experimental studies show support for loss aversion; for example, many consumers are reluctant to accept fair bets because the potential loss, of say $20, outweighs the attractiveness of possibly winning $20 (Kahneman & Tversky 1979). If consumers have an approximate reference price for a brand, an increase above that reference level is perceived as a 'loss' with a corresponding drop in the likelihood of purchase.

  This differential effect of price increases appears to be even more marked for private label brands: their downward price elasticity tends to be quite low (because they are already often cheaper), but if their price is put up, their sales loss is quite pronounced.

  5) If the brand's normal price is close to the average of the other brands

  If the price of a brand is close to the average prices of competitors, its elasticity will be greater. There are two reasons for this. First, if prices for all the brands are very dispersed, then a price change by one brand bridges the price gap between it and competitors by a smaller amount.

  Second, if the brand's price is close to the average of competitors, a price change – either upward or downward – will take the brand past the price of one or more of those competitors (i.e. go from cheaper to dearer, or vice versa). Pricing experiments showed average price elasticities in double figures when all the brands start at the same price – because a change results in setting one brand's price below all the others, resulting in a huge swing in consumer choice. A good example of this phenomenon is petrol in urban locations where it is easy to switch from one station to another.

  By contrast, elasticities are smaller when a brand's price is further away from the centre or average of the prices in its category. This dampening effect on price elasticity is especially marked if most of the other compet
itor brands' prices are close together.

  There are three other category characteristics that lead to larger price elasticities. These are (1) categories with fewer brands, (2) goods that consumers buy very regularly and (3) categories that can be stockpiled (Narasimhan, Neslin & Sen 1996).

  Can price cuts be profitable?

  An important point to make is that even very large increases in volume may still not make any extra contribution to profit. This is because, when a price is cut, the contribution margin for each item is cut even more. For example, if the normal contribution margin on a product is 50%, a 10% price cut shaves 20% off the contribution margin. In the Totten and Block (1994) example (previously discussed), a price cut as reported of 45% would eliminate virtually all the contribution margin for a normal packaged goods brand; therefore, the brand is being sold at close to manufacturing cost. So while the brand could sell 400% more volume, it would not make any more profit, because there is almost no contribution margin on each item sold.

  The short-term profitability of a price promotion depends on these three factors:

  •the contribution margin of the brand at normal price

  •the depth of the price cut

  •the price elasticity of the brand.

  Using these three factors we can easily work out 'break-even' scenarios that show how much extra volume needs to be sold to break even from a price reduction. If the contribution margin is low to begin with, a massive increase in sales is needed to break even. If the contribution margin is high, a price cut can earn more profit, even from a modest increase in sales. This is because every extra unit still carries enough contribution to pay for the price cut. Three example scenarios are shown in Table 10.2. The first example shows that a brand selling at the normal price and earning a 30% contribution margin and employing a 10% price cut must sell 50% more just to make the same amount of money. If the promotion creates an increase in sales of anything less than 50%, the brand will lose money!