How Brands Grow

by Byron Sharp

  This marketing orthodoxy means that it is derogatory to suggest to a brand manager that the loyalty his or her brand enjoys is due to habit, availability, and/or lack of caring on behalf of buyers. Academics have long derided such buying behaviour with pernicious terms such as 'spurious loyalty'.46 Marketing consultants and professors urge brand managers to build deep relationships with their buyers. While market researchers offer tools that guarantee to find (some) passionately loyal buyers.

  Consumers can apparently even have deep attachments to cheap everyday consumer goods. In 1997 a highly respected academic journal included an article based on long (2-3.5 hours), paid interviews with eight coffee drinkers (see text box at the start of this section for a flavour of the article). The authors of the article proposed that consumers have deep and complex relationships with (coffee) brands – and somehow their research findings turned out perfectly in line with their prior beliefs.

  If you think that is all just a bit silly you’ll be dismayed by the lead article in the March 2012 volume of our discipline’s top cited journal (the Journal of Marketing) titled “Brand Love”. Qualitative interviews were followed by a one-off cheap online survey of 268 US undergraduate college students at a single university. The students were asked to pick a consumer electronic brand they loved, and another they felt “mostly neutral, not love, about”, and were then asked 59 largely attitudinal questions about each. No surprise that the iPod was the most chosen brand, the authors didn’t mention the popularity of digital music with college students but instead attributed this love to the fact that it allowed personalising of music selection, and that “the iPod’s industral design (the simplicity of its click-wheel mechanism and menu navigation) is likely to promote a sense of natural and intuitive fit”. What this psycho-babble has to do with brand love is beyond me, and obviously beyond Apple too, as they have dropped the click wheel.

  The article’s statistical analysis shows every question in the survey is highly correlated with every other question which apparently somehow gives “richness and diagnostic insight”. To disguise their lack of findings the authors use many vacuous phrases that are meant to impress like:

  •“leads to a much more comprehensive and integrated understanding of how consumers actually experience brand love than the prior academic study of its individual components”,

  •and “we believe that our higher order prototype model adds value”

  The trivial nature of this research is well illustrated by the vague recommendations for marketing managers to:

  •“facilitate passion-driven behaviors...e.g. Scion cars encourage owners to choose...wheels, sports mufflers and so on”

  •“build brands that facilitate self-brand integration”

  •“create positive emotional connections”.

  All this in a journal that prides itself on producing findings that are valuable to practitioners!

  Almost buried on the very last page is a single line within a paragraph: “However, the brands consumers really love are likely to be a small minority of the total brands they purchase”. To put this another way, any brand manager will find that only a tiny number of their buyers have anything that approaches love for the brand.

  Another such academic article (titled “Emotional Branding Pays Off”) that desperately tried to hype up the importance of brand love actually found that a mere 4% of the average Australian beer brand’s male buyers ticked the survey box “I would say that I feel deep affection for this brand, like ‘love’, and would be really upset if I couldn’t have it”. If an Australian male can’t get emotional about beer brands then what brands can anyone get emotional about? And these few brand lovers reported only buying the brand half the time (i.e. they mostly bought other brands they did not love).

  Sadly I expect a raft of misguided PhD theses on brand love over the next decade, as academia falls in and then out of love with its latest fad.

  While it may be new to academia, the idea that consumers have 'relationships' with brands is an old idea, which is routinely repackaged by marketing consultants. Kevin Roberts' Lovemarks is a classic, and quite humorous, example. To quote, “Lovemarks are brands that create an intimate emotional connection that you simply can't do without. Ever.” (Roberts 2004).47

  A more down-to-earth view is that if buyers don't care deeply about a brand then they could be easily lured away to buy another brand. Yet even this supposedly commonsense statement is actually an untested empirical assumption. It's just as plausible to argue that a lack of caring is the cornerstone of ongoing loyalty.

  Connection to a brand

  Consumers are busy people. They have hundreds of thousands of brands vying for their attention. Thanks to the competitive modern economy, and government regulation, most of these options are sound. Therefore, the choice of brand is trivial compared to the decision of whether or not to purchase from the category. And aside from cars, houses and a minority of other categories, even the decision to buy from the category is a minor one – “Shall I have a snack or wait until lunch?”

  Brands are a necessary evil: they add a layer of complexity to the buying decision, but they also allow for routines (“Ah, there's my brand” or, “Oh yes, I've heard of that one”); such habits make buying easier – automatic even. From cars to canned soup, routine results in passionless48 brand loyalty. Also, we're usually so unaware of our habits that we barely notice that we eat the same meals, shop from the same shops, etc. And, as we've seen, loyalty is rarely exclusive; it's nothing like loyalty to a sports team or a country. Buyers typically have a number of brands that they routinely buy, so 100% exclusive loyalty is much rarer than marketers expect.

  Marketing texts sometimes dismiss these buying patterns as simply a case of the messy real world deflecting customers from their true intentions. But this interpretation doesn't fit the facts when attitudes, perceptions and intentions are measured. Consumers' thoughts and plans also show this prosaic uncommitted loyalty.

  Brand knowledge, attitudes and intentions

  Long ago marketing wisdom was that Avis buyers rented from Avis and Hertz buyers rented from Hertz. So if Hertz had 23% market share in a particular country, it was assumed that 23% of car renters used Hertz. As soon as marketing scientists starting looking at ongoing panel records of real-world repeat buying they realised the error of their assumption. Buyers are polygamous; brands share consumers. In the long term, 23% market share means that most category consumers use Hertz, but on average, they only use Hertz not much more than 23% of the time.

  The story is similar with attitudes, perceptions and intentions (i.e. people's thoughts about brands). If surveys show that 30% of people say Hertz rents attractive cars then the common assumption is that 30% of people believe this and 70% don't. While the reality is that a lot of people believe this (or remember they believe this), but just not all of the time.

  Let me explain. We can see the fickle (actually probabilistic) nature of people's beliefs if we ask them the same question more than once. This is seldom done in market and social research. Surveys generally ask different people each time, which is why even experienced market researchers are generally unaware of this phenomenon.49 If, on the first survey, 30% of people agreed with the statement, “Hertz rents attractive cars”, then on any subsequent survey the figure is also usually very close to 30%. This (misleadingly) suggests that the individuals largely always give the same response, either agreeing or not agreeing with the statement. But, if we analyse the answers of each individual we see something startling. Typically, only about half the people who on the first survey agreed with the statement do so again on the second survey, and an equal number of people who did not agree with the statement the first time now agree with it. So the overall level of agreement remains at 30% but the repeat or stability rate (people saying yes both times) is only half (i.e. 50%). Scale-type questions show the same instability.

  Table 7.3: Image belief stability survey to survey, Australian banks, small &
medium size business respondents

  “friendly and helpful”

  RL1 RL2 RR

  Brands % who agreed on 1st survey (Response Level 1) % who agreed on 2nd survey (Response Level 2) % of those who initially agreed who repeated that agreement on the 2nd interview

  (Repeat Rate)

  CBA 30 35 57

  Westpac 19 20 48

  NAB 19 20 45

  ANZ 16 17 44

  Average repeat rate= 49%

  correlation between the response level (R1) and the repeat rate, r=0.98

  Source: Survey data from Ehrenberg-Bass Institute, University of South Australia.

  Only half (or less) of the respondents who agreed with the statement (“Bank X is friendly and helpful”) agreed when interviewed again. However, this does not mean that people are becoming increasingly cynical about how banks serve them. Overall, the brands get very similar scores survey to survey (RL2 is very similar to the previous RL1), because every person who no longer votes for the brand is matched by someone who now votes for it (i.e. some of the people who did not agree with this statement on the first interview do so on the second). This instability in individuals’ response does not mean that those who agreed on the first interview but not the second have fundamentally changed their minds about the brand. It's just that they sometimes think this, so on some interviews you catch them saying it, and some you don't.

  This phenomenon applies to all brands, and to all image beliefs, attitudes, purchase intentions and most awareness measures. Table 7.4 gives examples across different attitudinal or image statements and across different categories. We can see that the repeat rate varies from brand to brand; it also varies across different beliefs. That's because the repeat rate largely depends on the overall response (agreement) level for that belief and brand. Logically, if on each survey 100% of respondents agreed with a statement (e.g. “How Brands Grow is a superb book”) then the repeat rate must also be 100% – each survey can only record a score of 100%. However, for beliefs that score less than 100% their repeat rate could be 100% or it could be lower (i.e. a score of 30% in each survey could be because 30% of people agree on every interview, or everyone agrees but only 30% of the time, or something in between). In the real world we find the repeat agreement level is almost always far lower than 100%. In fact we can predict the repeat rate based on the level of initial agreement (Dall'Olmo Riley 1997) – this is the whole story, the repeat rate has surprisingly little to do with the type of belief or other factors.50

  Table 7.4: Image belief stability survey to survey, multiple categories, different types of image belief

  Category Brand image belief Initial agreement (%) Repeated agreement (%)

  Fast food outlets Good for a snack 29 56

  Banks Progressive 22 47

  Banks Expert in the areas it deals in 21 48

  Insurance Provides fast service 17 42

  Supermarket Sells low-quality fruit 14 36

  Liquor/spirits Unique 10 36

  Data source: Ehrenberg-Bass Institute, University of South Australia.

  Note: Repeat rates are typically low. Usually under 50% of those people who agreed with the statement on one survey do so on another interview.

  This doesn't mean that our attitudes and beliefs are random, in the sense of being without direction, mindless. Rather it means that our individual brand memories, like our brand buying, are probabilistic. We each have a steady, ongoing propensity to think something, and for most of our beliefs that propensity is not 100%. Not surprisingly, whether or not we recall a belief is highly dependent on the situation, and is very much affected by the cues that are used to elicit the belief, and can change depending on what other things were going through our head at the time. 51

  Our brand buying behaviour is like this: I might choose Avis 60% of the time, but that doesn't mean that sometimes I won't choose Hertz (indeed it suggests that I definitely will sometimes choose Hertz) even two or three times in a row. This behaviour appears to be random, even though each time I buy Avis or Hertz there are particular reasons for this. But those reasons are so variable that anyone watching my choices from a distance might think I was merely tossing a coin to make my choice (a coin that just happened to be weighted 60% in favour of Avis).

  So perhaps our attitudes merely reflect our buying. Our behaviour has a strong effect on our attitudes. But also, the fickle nature of our intentions (or their recall) is one of the many causes of our probabilistic buying patterns. Presumably attitudes play a role, albeit a small one, in the weighting of that mental coin.

  The key discovery is that most of what we think about brands is not absolute. It is natural to tick the survey box that says we are completely satisfied with the brand and one hour later tick the 'only somewhat satisfied' box. Indeed, most of what we think about brands is so trivial, so barely thought through, that we will happily change our mind in a second 52. Actually, it's misleading for me to talk about changing of minds. It's more apt to say that we haven't completely made up our minds yet. Or rather that we have made up our minds to sometimes think/like/buy the brand.

  Therefore, attitudinal commitment to brands (or to buying them) is much weaker than marketing mythology makes out. But it can exist; for example, the Apple computer on which this book is being typed is rather dear to me personally. Which leads us to question whether it is possible for a brand to build exceptional emotional or symbolically based loyalty. How important to a brand are these special, committed customers? Have some brands excelled at building passionate user bases?

  Brand fanatics

  It's long been fashionable in marketing to uphold brand fanatics as ideal customers. While most buyers who are 100% loyal to a brand are simply light category buyers, there are a (very) few customers who both regularly buy from the category and yet are very loyal to one brand. There are also a few buyers who claim to be passionate about a brand. How important are these people? Who are they? How does a brand attract more passionate consumers?

  Don Gorske is a perfect example. This Wisconsin resident has consumed more than 23,000 Big Macs; he claims that the Big Mac constitutes more than 90% of his solid food intake. We know that he's eaten more than 23,000 Big Macs because he records the ongoing total in a notebook. This apparently is just one example of his obsessive behaviour. Don Gorske, to put it mildly, is a little odd.53 Every brand has a few very loyal and heavy buyers, a few passionate advocates.

  The Lovemarks website lists thousands of brands that have been nominated as Lovemarks by visitors to the website. These are brands that some people feel passionate about (or at least feel so some of the time). The Lovemarks list is moderated and nominees have to make a case for a brand's inclusion. Yet when I looked the list still includes prosaic brands such as 7-Eleven. Also, Air New Zealand, Continental Airlines, Delta Airlines, Air Jamaica, KLM, Korean Air, Qantas and Swiss Air all got nominated, as did many other airlines. Among car manufacturers, General Motors, Ford, Honda, Mazda, Mitsubishi, Mercedes, Nissan, Toyota and Saturn have all been nominated54. This shows that any brand can have a few fans, a few Don Gorskes; but, it does not show that some brands are special or that these fans are of any financial or strategic consequence to marketers. These make for an entertaining story and that's all. Advertising agencies, who in general know very little about buyer behaviour, seem to love these stories.

  Brand loyalty: harley-davidson and apple

  “Most people merely buy a computer. Apple Macintosh customers fall in love” (New York Times, 2 April 1998).

  “There is no product on the planet that enjoys the devotion of a Macintosh computer. Famously dedicated to their machines, many Mac fans eat, sleep and breathe Macintosh” (‘The Cult of the Mac’ by Leander Kahney, 2004).

  ‘Baba Shiv, a professor of marketing at the Stanford Graduate School of Business, compares Apple's fan base to Harley-Davidson motorcycle riders who pass over arguably higher-quality Japanese bikes” (‘Apple fans loyal despite iPod, iphone 3G woes', CNN, Fox News, 20
08)55.

  Brands like Harley-Davidson and Apple are the poster children for emotion-based brand loyalty. They are regularly cited as having passionate, highly loyal customer bases – though few writers provide evidence to back up such claims. Also, no one questions why this loyalty failed to protect Harley-Davidson from losing market share to Japanese brands, or Apple to PC clones.

  I'm sure it is true that tattoo engravers (if this is the right word) receive more requests for Harley-Davidson tattoos than they do for Kelloggs Cornflakes tattoos (this slightly ludicrous example provided by Parker and Stuart 1997) but this says more about the respective product categories than it does about loyalty. And the fact that Harley-Davidson riders are more likely to have tattoos of any sort than the average breakfast cereal buyer (i.e. a statistical selection effect). And that first purchases of motorbikes and tattoos tend to coincide, whereas consumers tend to be first introduced to Cornflakes at an age when their mum won't allow them to get a tattoo, even if they wanted one!

  What are the buying facts about Apple and Harley-Davidson loyalty? First, just like other brands, their customers are polygamously loyal to a number of brands. The Harley-Davidson's share of category requirements (SCR) metric is reported at about 33%. In other words, Harley-Davidson buyers purchase other bikes twice as often as they buy Harleys. This is a very normal sort of loyalty figure for a brand. Similarly, Apple's repeat-buying level is much lower than 100% and is not very different from its competitors; see Table 7.5.

  Table 7.5: Apple owners show only moderately higher than expected loyalty

  Brand Repeat (%)

  Dell 71

  Apple 55

  HP/Compaq 52

  Gateway 52

  Source: MetaFacts Inc., 2002-03 www.technologyuserprofile.com