1. INTRODUCTION

Brand differentiation is essential for brand success – or so says conventional wisdom and much of the branding literature (e.g., Trout, 1999; Aaker, 2001). This belief affects marketing strategies, advertising briefs, tracking studies and more. Marketers and marketing strategies are judged on their success in differentiating their brand.

However, very little is known about the level of buyer perceived differentiation. Do buyers always need to believe the brand is different from competitors in order to buy it?

Background

In economic theory, brand differentiation simply means that buyers show preferences. The greater the preference the more secure customer base. The famous downward sloping demand curves illustrate the theory, they slope down to show that when a brand increases its price a bit it loses some, but not all, customers (supposedly those who do not value its point of difference so greatly).

In marketing theory, identifying and communicating a point of difference is considered essential. This point of difference, be it functional or emotional, must be ‘meaningful’ in that it provides buyers with a reason to purchase the brand (Aaker 2001). For example Shape yogurt is meant to be bought because it is lower in fat than many yogurts, Body Shop supposedly for its perceived values and ethics.

Established brands are said to need to maintain a point of difference in order to stay desirable to their customers. This influences the copy testing approaches used to select hopefully more effective advertisements. Advertising that does not give buyers a unique reason to buy is thought to be less effective. Some creative concepts/executions are therefore rejected in favour of advertising that conveys a point of difference, even though likeability and ‘cut through’ may be sacrificed.

Breakthroughs in perceived differentiation through either product features (e.g. Apple’s iMac) or clever advertising (e.g. Marlboro man) are seen as the pathway to growth. Likewise differentiation is considered essential for a new entry into the market. Undifferentiated new entrants supposedly fail because no customers should be motivated to buy them over existing brands.

Conversely a brand’s decline in share is often attributed post-hoc to it “losing its point of difference” in the marketplace – though empirical evidence is seldom cited.

Differentiation: A Feature of Brands or of Markets

Buyers do show loyalty and preferences, our extensive R&D shows that restricted repertoires appear to be near universal (see Ehrenberg-Bass Institute Report 1 for Corporate Members). So differentiation in the economic sense exists. Brands do not compete as commodities where lowest price always wins. But does loyalty exist because buyers see the brand as meaningfully different? Are brands really differentiated so that they appeal to distinct sets of customers or needs? Two of the regular patterns we have documented across markets suggests that this should be questioned:

1. the Duplication of Purchase Law and
2. the lack of distinctive user bases for competing brands.

The Fit of the Duplication of Purchase Law

One reason to question the conventional view of brand differentiation is the widespread fit of the Duplication of Purchase law (DoP). The DoP law states that brands share customers in line with popularity, not with image positioning or marketing strategy. Any brand shares more customers with larger competitors, and fewer customers with smaller competitors. So any brand shares customers with all other brands – it competes with all the other brands in the category. That there is “sharing” at all is something of a blow to the idea of differentiated brands. But that this sharing is regular and predictable (i.e., the same pattern for all competitors) is something that would not be predicted if brands were highly differentiated.

Brand differentiation should produce unequal sharing (referred to as ‘partitioning’). Brands that are differentiated in a similar way (e.g. both trendy) should share customers more than they do with other brands. Brands that are differentiated in opposing ways (e.g. one stylish, one simple and tough) would be expected to share few customers.

The widespread fit of the Duplication of Purchase law shows that partitions are weak, for example while there is a diet partition in the Soft Drink market double the number of Diet Coke drinkers also buy normal Coke as buy Diet Pepsi (or any other diet softdrink for that matter – Ehrenberg-Bass Institute Report 12 for Corporate Members), and BMW wins (and loses) more customers each year from Ford as it does from Mercedes (see Ehrenberg-Bass Institute Report 6 for Corporate Members). Stronger partitions do exist but are typically linked to subcategories (i.e. leaded versus unleaded fuel; foreign versus domestic automobiles) rather than specific brands (see Ehrenberg-Bass Institute Report 1 for Corporate Members).

The Lack of Brand Segmentation

The concepts of differentiation and segmentation are typically aligned. Textbook convention is that brands are differentiated to appeal to a specific segment of the market (e.g. 18-25 y.o. males). Yet competing brands within a category tend to appeal to the same types of customers. While there are differences between markets (e.g. buyers of nappies tend to be people with babies and not people without babies) brands within the market (e.g., Huggies and Pampers) tend to have buyers with much the same characteristics (see Ehrenberg-Bass Institute Report 7 for Corporate Members).

Therefore the objective of building a brand that faces less direct competition and/or appeals to a certain segment of the market doesn’t match what we observe when we look at buyer behaviour and buyers’ characteristics. In the “real world” brands sell to the same type of customers.

In line with this, Sharp and Dawes (2001) argue the level of differentiation is largely a market characteristic, i.e., something shared by all the brands within the category. While some markets are more differentiated than other markets (e.g., softdrinks are slightly more differentiated than water – see Table 3), it makes less sense to talk of brands being more differentiated than their competitors. Loyalty does not vary much between competing brands and so nor does the degree of differentiation.

Thus there appears to be enough empirical evidence to seriously question the conventional view of differentiation. In this report we go further to empirically examine:

Do Buyers see Their Brands as Differentiated from other Brands in the Market?

Specifically:

1. Do many of a brand’s buyers say it is differentiated from other brands?
2. Does this vary between brands ? E.g., are big brands more (or less) likely to be seen as differentiated ?

2. RESEARCH METHOD & DATA

We draw on data from 13 different markets. Ten product categories were in Young and Rubicam’s BrandAsset Valuator (BAV) UK data collected in 1999¹. The other three were collected in Australia in 2002-4. As part of a larger questionnaire, respondents were asked to indicate whether they perceived particular brands to be different, distinctive, original or unique. These four perceived differentiation attributes were interspersed within other brand attributes. Questioning was ‘free response’: respondents could name as many or as few brands or attributes as they liked.

The markets varied considerably (see Table 3). We deliberately examined categories with obvious functional differences between brands (e.g., computers) as well as categories where buying is argued to be “image driven” (e.g. skin care, spirits).

There were some deliberate differences in the questioning and data collection method between studies, yet the results were consistent.

• The UK data was collected by a sizable self completion booklet. Brands from different product categories were interspersed and respondents were instructed to fill in the booklet one brand at a time. Hence respondents were not asked to compare the brand in question with other brands directly (Coke with Pepsi or Dr Pepper). Instead, they were just asked whether the listed brand, on its own, is different (and so on for each of the 24 attributes contained in the set).
• In contrast the Australian data was collected via telephone with instruction to answer for each attribute, across all brands. So respondents were given a list of brands and asked which of those brands was different etc.

Rather than inferring the degree of differentiation from brand image analysis (the traditional way) our data provides direct measures of perceived differentiation, whether respondents perceive the brand as different, or distinctive, or possibly original, or even unique. This direct approach avoids the problem of having to subjectively infer from the image scores respondents give whether or not they really feel a brand is different.

For example, while more people associate Coca-Cola with “red” than do Pepsi is it right to infer from this that Coca-Cola is seen as differentiated? Any more so than Pepsi is because of its association with “blue”? Whereas if many Coca-Cola users say that Coca-Cola is different, distinctive, original and unique this would appear to be straightforward and solid evidence of a high degree of perceived differentiation.

We were mainly interested in the responses of users, that is, people who had bought the brand. Did people need to see the brand as different in order to buy it (or at very least having bought it did they now appreciate its differentiation)? We calculated the % of users of a brand that associated that brand with each of the “differentiation” measures. We separately calculated these associations for each brand’s non-users.

3. RESULT

We first give results for Soft Drinks and Information Technology. Then we summarise the overall results across all 13 categories.

Soft Drinks

The Soft Drinks in the Y&R BAV data were functionally very different from each other: from colas to energy drinks (e.g. Lucozade) to diet soft drinks (Diet Pespi). Table 1 shows the results for the 5 largest and smallest market share brands. In general, less than ⅓ of users link a brand to any single differentiation attribute.

The results are similar for each brand. Despite a 7-fold difference in the size of the user bases. For example, market leader Coca-Cola has 32% saying it is original, while Canada Dry (the smallest brand here) has a very similar 33% of its users saying it is original.

More generally:

• Only 1 in 10 or so of each brand’s users associated it with any one of the four differentiation attributes.
• About half did not associate it with any of the four attributes.
• These scores do not vary much between the brands.

The majority of brand users therefore do not consider the brands they use to be original, distinctive, different nor unique. This held true irrespective of brand size.

Some brands were slightly higher than the norm. These brands (e.g., Tizer) are different in formula and taste from the 4 biggest brands in the market (which are all colas). Unsurprisingly, users give these brands slightly more responses for being differentiated. However, even for these brands, still only about ½ of any brand’s customer base explicitly state that a specific brand is differentiated (in any of the 4 measures) from the other brands it competes with. Having a functional difference does not guarantee buyers will think of the brand as different (see Tango).

Information Technology

Our second example is deliberately chosen as a very different market from soft drinks. Computers and computing equipment:

• Are less frequently bought
• Have greater value per purchase
• Are much more complex products etc

This market also encompasses brands that operate in several different but related sub-markets (e.g. some hardware, some software, some both). Despite these differences, the results are similar to Soft Drinks.

• Typically only 1 in 10 or so of a brand’s users associated it with each particular attribute.
• Only about ⅓ of each brand’s customer base perceive the brand they use to be either original, distinctive, different or unique (i.e., associated the brand with any one of these attributes).
• And these results do not vary substantively between brands.

Apple Users – Thinking Different?

Microsoft scores similar to Apple on Original and Distinctive, but is of course very large (and a near monopoly in some markets). Otherwise the only notable deviation is Apple as 25% of its customer base thought it was unique, twice that of any other brand and 5 times more that of some of its most direct competitors, like Dell. But 25% is still not high (75% then did not claim it as unique). And still 65% of Apple’s customer base (2 out of 3 customers) claim it is neither original, distinctive, different nor unique. This figure is similar as that for other brands.

This may seem amazing given that Apple computers tend to look distinctive and have a UNIX based (not Windows) operating system. However, on reflection, an Apple Macintosh is a personal computer (PC) that runs software (e.g., Microsoft Office), sends email, stores files, and prints. Most Apple customers did not, it seems from Table 2, buy it to be different. Presumably (like buyers of other brands) they bought it because they thought it was a good computer.

Results Across all 13 markets

Table 3 shows the average results across all 13 markets. The results for these other 11 markets mirror those shown in Tables 1 and 2, i.e. low response levels across all 4 attributes (and little variation across brands).

Deviations – Some More Differentiated Brands

Buyers did perceive some particular brands to differ a bit more than their peers. These brands tended to be higher priced and generally of small market share. For example, in Table 4 Häagen-Dazs often scores several times more than Walls. Other examples were Bang and Olufsen (in Consumer Electronics) and New Convent Garden Soup (Soups) and Shisiedo (Skincare). In total we found 12 brands (out of over 115) that exhibited a deviation on at least one attribute. That there were many small brands that were not seen as any more differentiated suggests that pricing (not size) is probably the key factor. That they tended to be small brands is consistent with the idea that differentiation (e.g. higher prices and quality) can restrict market share (appealing to a niche and/or occasional demand).

Some perhaps unexpected similarities also occurred, e.g. Table 4 shows that while Walls has 3 times the customers of Nestle, both perform virtually identically across the four attributes.

4. SUMMARY

Buyers are not saying ‘I buy this brand because it is different’.

The majority of a brand’s users do not explicitly perceive it to be differentiated from other brands. Generally only 1 in 10 of a brand’s users are likely to associate it with any perceived differentiation attribute, and the majority will consider brands as neither original, distinctive, different nor unique. Therefore the degree of direct influence brand differentiation can have on buyer behaviour looks at best low.

Could this result simply be a measurement effect? Were respondents lazy? We think not. The simple fact that other attributes were able to gain response levels of greater 50% strongly suggests that this result was not simply a measurement effect (see appendix).

Perceived Differentiation is Consistently Low Across Brands

Our second major finding is that this low figure typically did not vary across brands. This is despite considerable differences in marketing activities, market share and product formulation.

Deviations tended to be for smaller, higher priced brands. Being highly differentiated may allow you to charge a price premium, but at the expense of restricting the size of the brand. Whether this makes for a more profitable brand is a different issue, as there are costs associated with being ‘non-standard’. A simple explanation of this finding is that having higher price/quality may increase your perceived differentiation and lower your demand simultaneously.

Finally, in general, buyers were almost twice as likely to perceive their brand to be original or distinctive than they are to see the brand as different or unique. This was consistent across markets. Therefore we can surmise that while brands sometimes can ‘stand out’, in that they are perceived more often as original or distinctive, it is rare that buyers perceive the brand to be actually different or let alone unique (all brands can occasionally stand out).

5. IMPLICATIONS

For decades the marketing literature has focused on perceptions of brand differences as the main reason why one brand is chosen over another. This emphasis seems misplaced. If buyers of a brand do not think their brand is different (or unique, distinctive or original) then presumably this is not the reason why they buy it. We need to look elsewhere for explanations of why this brand (and not others) is bought. The answer may lie more in the area of physical availability and mental availability (salience).

Marketers are told to make customers perceive differences between brands. Yet even marketers of highly successful brands seem to have failed. This leads us to the conclusion that perceived differentiation plays little part in the success of brands. Or far less than most have stated – it is certainly not a case of “Differentiate or die”. It appears that even without being more differentiated some brands have managed to grow much larger than competitors.

That buyers do not often notice that the brands they use are different from other brands in the market seems odd given that some products within markets were very functionally different. One reason for this is, perhaps, that for most buyers most of the time, brand choice is a fairly trivial task. So not a lot of time is spent comparing between brands. As such differentiation (which is relative to other brands) is not noticed or thought about. We find that buyers seem to recall something about the brands they use, and very little about the ones they don’t. That’s why even fewer responses come from non-users.

The main implication for marketing practice is that marketers need not be concerned with convincing buyers that they are in some way different. This should take a considerable weight off marketers shoulders because our data shows that such a task must be near impossible. Even functionally different brands like Apple, Bang and Olufsen, and Ben & Jerry’s have largely not achieved this. Instead marketers need to focus on the things that seem to make customers buy. Such as brand salience (see Ehrenberg-Bass Institute Report 16 for Corporate Members).

These findings also support our contention in Ehrenberg-Bass Institute Report 13 for Corporate Members, namely that it is useful for brand advertising to have ‘talking points’ rather than ‘selling points’. That is, it is useful to have something that identifies the brand, even something seemingly trivial or unrelated to product benefits (like McDonalds’ golden arches). This has the benefit of making it easier for customers to notice which brand is advertising, thus building/maintaining brand salience. But this sort of ‘distinctiveness’ need not be anything meaningful that provides a reason for buyers to buy the brand. It can be a colour, packaging, a logo or a particular creative device.

Finally, this research also highlights the dangers of relying solely on techniques that focus on identifying differences between brands (e.g. perceptual mapping). Most multivariate statistical techniques are very prone to outliers (often fairly small ones). Our analysis was deliberately simple and transparent and we were able to highlight differences between brands. However, more importantly, we were able to clearly see and highlight the similarities, which most likely would have been hidden in multivariate analysis.

Future Research

The available data does not provide an exhaustive test of the presence or value of differentiation. For example, it may well be that differentiation is more important for new brands, at least at first. However, this report is significant in that it starts to test differentiation theories, which up until now have been largely taken as faith. If differentiation is as vital as most marketing texts claim, then it should stand up to any form of rigorous empirical testing. We think it has failed its first test.

We hope that these results will at least start people questioning how important it is to try to make customers think your brand is different. Marketing managers have spent much effort on trying to find features on which to differentiate their brands. This has led people to rely on ever more obscure aspects of branding (e.g. brand personality), that seem to have little relevance to buyers (Ehrenberg-Bass Institute Report 14 for Corporate Members). This is not beneficial for future brand growth or the effectiveness of marketing in general.

Our R&D on this topic continues.