Foreword — Editor Comments
Small brands have not received the study they are due considering that, in aggregate, they are a major competitor to category leaders. We therefore benchmarked small brands’ main buying measures (penetration and purchase frequency) in 36 packed goods categories. We make 3 big discoveries:
- The commonly sought “niche” performance, i.e. relatively higher loyalty, is neither common (one in ten), nor a driver of growth.
- A surprisingly large number, over half, of all small-share brands show a persistent loyalty deficit and yet they still survive.
- The evidence shows that when small share brands grow, a Double Jeopardy model always predicts their performance metrics. Therefore for every small-share brand, buyer attraction must be the managerial priority.
This evidence strongly suggests that managers of small brands who want them to grow need to be wary of inadvertently or intentionally niching their brand.
— Byron Sharp
Small brands suffer twice
Small brands are at a disadvantage to bigger competitors – they have fewer resources, lower scale economies, are less well known, have fewer customers and often struggle to maintain distribution even for shorter product lines. More attention is paid to the rivalry between the biggest brands than to their more numerous but smaller challengers, a fact that we now address in this report.
Competition in Consumer Packaged Goods (CPG) is not restricted to the few large brands that usually account for over half of a category’s sales. On many occasions the many customers of these leading brands select a familiar but far smaller share brand instead, and these purchases add up to substantial market share.
The expected performance of competing brands is described by the Law of Double Jeopardy (see Report 26) which says that small brands are small because; first, they have fewer buyers than bigger rivals, and second, because their buyers buy at a slightly lower average rate (they are a little less loyal). A still surprising implication of this regularity is that the niche brand outcomes, which are often planned by brand managers (targeting relatively few customers who buy more often), are an exception. That said, it’s plausible that small brands in particular might be niched:
- lack of resources may motivate brand managers to focus efforts on a segment, rather than all category buyers
- attractive new brands may show high loyalty amongst the small customer base they have recruited – and this excess loyalty might be an indicator that they are destined to be a leading brand
- highly differentiated brands are more likely to be small because they appeal to some buyers but not to others, this would give them the niche pattern of a small customer base with higher than expected loyalty for their market share
So we expected a higher proportion of niche brands among our sample of small brands than normally seen in samples of larger brands. Surprisingly we found the opposite.
Plotting Double Jeopardy for small share brands
The repeat-purchase loyalty of any brand can be predicted just from its size. A number of models will calculate this but one of the most versatile is the NBD-Dirichlet (Ehrenberg, Uncles & Goodhardt, 2004). Figure 1 shows an example.
Figure 1: Double Jeopardy in UK deodorant & body spray buying (14 brands, annual metrics)
Figure 1 shows a theoretical Double Jeopardy line to show the expected relationship between annual penetration and average purchase frequency in the UK deodorant and body spray category. We divided the brands (excluding private labels) into leaders (the top five by share) with the small share competitors to the left of the vertical division. Anything beyond a 10% deviation in purchase frequency is termed “change of pace” if the loyalty is below expectation, or “niche” if there is excess loyalty relative to size (Khan, Kalwani & Morrison, 1988; Sharp, 2010). In Figure 1, the Double Jeopardy slope is typically shallow because rival brands differ greatly in their buyer numbers but relatively little in their loyalty. Brand purchase frequencies are ranged quite closely around the average of 2.2 annually, and largely within the 10% margins shown, except for a cluster of small share brands highlighted in red. For the small share brands we see just one example of niching (in green) – the model suggests a tendency small brands to under, rather than over, perform on loyalty.
To discover how common this tendency is, we examined 130 small brands in 36 categories in one year, and then 70 small share brands in 15 categories over five years. We considered manufacturer brands only, removing all Private Labels before the analysis.
How many smaller brands are niching?
Traditionally, marketing textbooks have encouraged marketers to differentiate their brand, and to target segments. Particularly in old established markets it’s often felt that the only way of gaining any share is to follow this strategy.
This is one reason we’d expect that many, if not most, small brands would be niched. Especially amongst small stable brands. Yet surprisingly this is not what we see. Table 1 shows the performance of 264 brands in 36 categories in a year. Separating brand leaders in the same way as before, we see that about one in ten small share brands performed as niche (excess loyalty) over the course of a year.
Table 1. Loyalty to small share CPG brands in one year (36 categories, excluding private labels)
When we examined the niching brands we found they were, as expected, nearly all functionally quite different from their competitive set – for example soya-based ice cream or children’s shampoo. This differentiation is a reason why their market penetration is limited and their loyalty (within their segment) is higher than expected. That is, because many category buyers probably would not or could not consider buying that brand because of its difference. These brands will probably stay small because of their differentiation.
One third of small share brands complied with the Double Jeopardy Law, with average purchase frequencies in line with their relative penetration. However, the main pattern was a systematic exception to the law. Small brands were much more likely to have deficit loyalty/excess penetration.
To identify the differences in the buying patterns between large brands, small niche brands, and small deficit loyalty brands, we compared the observed (O) and expected (T for theoretical) distribution of their most and least loyal buyers. Looking at the expected values (T) in Table 2 we see that bigger brands (average penetration 12%) are expected to have a fifth (18%) of their buyers purchasing five or more times in a year. For smaller brands (suffering twice) the heaviest buyers should account for 12% of their lower penetration (< 5%). Leading and small share brands are both expected to have many one-time buyers (about half the customer base: T = 49%), but smaller brands proportionately more (56%).
Leading brands perform just about as expected (average values within about 10%) but smaller brands do not. They generally fall below the DJ line, and deficit loyalty brands have a dramatically skewed customer base; while they attract over a third more customers than expected (35%), seven in ten make just one brand purchase in a year. Deficit loyalty brands have less than half their expected heavy buyers (6% vs 14%) and so the sales required to maintain share are being made to additional, lighter buyers.
One possible explanation for this excess penetration and deficit loyalty is that these small brands are somehow faulty (e.g. don’t taste that nice) and so many of their buyers do not repeat-purchase. This seems implausible because these brand do not fall out of the market.
Table 2: The nature of Deficit Loyalty
We now answer five questions about the performance of brands ranking below the top five – i.e. small share brands.
1. How common is deficit loyalty?
Brands outside the top five tend to have deficit loyalty: we found that over half had repeat purchase rates at more than 10% below expected levels. By contrast, very few (about 10%) were niching.
2. In the long-run, does deficit loyalty lead to decline?
No. Small share brands with deficit loyalty are able to hold their market share near stationary against much larger rivals and at least over five years (we also found that niche brands don’t grow either).
3. What causes deficit loyalty?
The customer base of a small share brand in deficit loyalty is dramatically skewed towards one-time buying. Around 70% of customers do not make a repeat purchase within the year (against a normal rate of about 50%). This means that the brand’s survival depends upon its managers being able to attract many more light buyers than predicted by brand size alone, and continuing to do so.
4. Are deficit loyalty brands somehow different?
No. We observed that niching brands are by and large functionally different from category competitors. However, no one simple explanation seemed to hold for small share brands in deficit loyalty. The Change of Pace idea is that repeat purchase is restricted by the brand appeal to variety seeking (the usual example is 7-Up the “uncola”) or by seasonality (Easter eggs). However our analysis showed that few deficit loyalty brands clearly fitted this description. Instead it seemed to be quite a normal function of brand size.
5. Does Deficit Loyalty predict brand decline? Does Niching predict growth?
Brand share and loyalty metrics fluctuate a little in time-series (Graham, 2009), but it was also possible that our cross sectional analysis had captured a moment in a trend (Pare and Dawes, 2012). In order to find out we compared the dynamic behaviour of brand shares and the persistence of loyalty patterns for 70 small share brands in 15 categories, over five consecutive years.
Using opening and closing brand shares (i.e. Year 1 vs Year 5) to identify dynamics we found that about 80% (56 brands) remained stationary (plus or minus one share point). And just five (7%) showed substantial increase in market share. We categorised these and the 14 brands that grew each as having Deficit, Normal or Excess Loyalty on the basis of performance in at least three of the five years (70% were persistent in four or all five, and importantly no brand moved from deficit to niche).
The continuous data reflected the annual picture – with over half (53%) of brands being persistently in deficit loyalty but it was also clear (Table 3) that every one of the five substantially growing brands had normal DJ performance metrics – share increases were primarily driven by annual penetration growth.
Table 3. Loyalty and brand share change
Summary & implications
In this report we compared the performance of small share brands against predicted Double Jeopardy outcomes and found that niche-marketing outcomes are rare, and that over half of all small share brands show the opposite pattern, i.e a persistent loyalty deficit characterised by an unexpectedly high volume of one-time buyers.
When discussing small CPG brands we have heard a number of managers referring to “small brand syndrome”, a condition in which smaller brands in the portfolio show a lacklustre repeat purchase rate, do not grow, and yet do not fail. Our analysis certainly proves the existence of such brands, indeed they seem to be the norm.
In an exciting development, our research showed that neither niching nor deficit loyalty brands showed significant share changes over five years, while about one in four small share brands with normal loyalty did increase their share over the same time. This suggests that Double Jeopardy benchmarking can reveal whether a brand has a hope of growing. More research is needed to see if this applies across more categories and countries.
Note: Ehrenberg-Bass Sponsors who wish to examine their brands to see if they are normal, niche, or deficit loyalty can access our Laws of Growth service.
References
Ehrenberg, A. & Goodhardt, G., (2002), “Double jeopardy revisited”, Report 26 for Corporate Sponsors, Adelaide, Ehrenberg-Bass Institute for Marketing Science.
Ehrenberg, A. S. C., M. D. Uncles, and G.J. Goodhardt (2004). “Understanding Brand Performance Measures: Using Dirichlet Benchmarks.” Journal of Business Research 57: 1307-1325.
Graham, C. D. A. (2009). What’s the point of Marketing anyway? The prevalence, temporal extent and implications of long-term market share equilibrium. Journal of Marketing Management, 25(9-10), 867-874.
Graham, C, Sharp, B., Trinh, G and Dawes, J (2017) The unbearable lightness of buying. Report 73 for Corporate Sponsors, Adelaide, Ehrenberg-Bass Institute for Marketing Science
Hammond, K., Ehrenberg, A and Long, S (2000) The Case Against Price-Related Promotions. Report 8 for Corporate Sponsors, Adelaide, Ehrenberg-Bass Institute for Marketing Science.
Kahn, B., Kalwani, M., and Morrison, D. (1988), Niching Versus Change-of-Pace Brands: Using Purchase Frequencies and Penetration Rates to Infer Brand Positionings, Journal of Marketing Research, 25 (November), 384-90.
Nenycz-Thiel, M., Romaniuk, J. and Sharp, B (2017) Physical Availability – What makes a brand easy to find & buy? Report 76 for Corporate Sponsors, Adelaide, Ehrenberg-Bass Institute for Marketing Science.
Pare, V., & Dawes, J. (2012). The persistence of excess brand loyalty over multiple years. Marketing Letters, 23(1), 163-175.
Sharp, B., (2010) How Brands Grow. Melbourne: Oxford University Press.
