Why Being Interesting Might Be More Important Than Being Different

Part 1: The Evidence


Aisles1

 

“The lead car is unique, except for the one behind it which is identical”

Murray Walker

 

 

 

Inherited wisdom
It is, perhaps, more than a little ironic that while our creative possibilities (both in their form, content and delivery) are exploding and evolving like ever before, how we actually think our efforts actually work in the real world has barely evolved. If at all. 

Community, fans, engagement, propagation, diffusion, interaction, participation, transmedia… We shouldn’t be fooled by the oversupply of new and fashionable language for it disguises a deeper truth. That is that for all the undoubted executional innovation and (less satisfyingly) all the rhetoric our industry puts out, the old notions of ‘messaging’, persuasion, positioning and brand image remain much as they were over forty years ago. Unchanged. and for the most part, unquestioned.

One of the central tenets of marketing is that in the battle for the hearts, minds, and wallets of customers, differentiation matters. A lot. Indeed it’s hard to find a marketer or marketing text that doesn’t argue in its favour. 

Rosser Reeves (1961) famously suggested that all brands and advertisements must offer a ‘Unique Selling Proposition’  – one that the competition either cannot, or does not, offer and that will convert people to your brand.

Ries and Trout (1972) did much to insitutionalize the concept of positioning, arguing that “to succeed in our cover-communicated society, a company must create a ‘position’ in the prospect’s mind. A position that takes into consideration not only its strength and weaknesses, but those of its competitors as well.” 

Of course, we might not be so quick to embrace their thinking if we remembered that in their original series of articles on positioning for Advertising Age, Ries and Trout also argued that “… advertising is entering a new era. An era where creativity is no longer key to success… strategy is king.”

And here, several decades later, is marketing guru David Aaker (1997): “Brand position… demonstrates an advantage over competing brands…. develop a point of advantage that resonates with the customer… provide a point of difference with respect to competitive offerings”

One of the key mandates of effective brand management is therefore to develop and sustain a unique point of difference. This, we are told should manifest itself in three consumer valuable dynamics:

Perceptions: Within a given category, people should have a very different set of mental associations for brand. 

Attitudes: People should believe that brands are different from one another.

Behaviours: All this should be evidenced in their people’s brand loyalties. Seeing and believing brands as being quite different should result in people treating brands as being poor subsitutes for each other. 

Not surprisingly then, ad- and marketingland spend a lot of time (and money) identifying a brand’s unique point of difference. And a lot of time (and money) monitoring those differences.

 

Challenging wisdom
What complicates things however, is that there exists sufficient evidence – stretching back some thirty or so years – to challenge this unwavering belief in the importance of positioning and differentiation in building successful brands.

Perceptions:
When the effect of usage and prototypicality is factored out, brands don’t have exclusive image attributes but share them with other brands.

Knowledge about a brand  – the root of any positioning – is not held uniformly by all buyers. A minority of a brand’s users account for half of brand knowledge, with the remainder spread thinly across the majority of its user base. In other words, while some people know a lot about a brand, most of a brand’s users  will know little.

Far from having fixed perceptions of brands, people’s view of brands are in fact very far from stable and vary across time. Brands are not ‘things’ in people’s brains, etched in, permanent and enduring. They must be constructed and reconstructed each time from all the myriad of associations in the mind. Whether these associations are evoked and assembled is – like any memory – a matter of cues, circumstances, and indeed luck

Attitudes:
When directly asked, the vast majority of people don’t think brands are unique or different

Behavior:
People don’t regard evaluating brands as very important, and don’t spend much cognitive time and energy making purchase decision.

Brands don’t attract different kinds of users.

Brands don’t have exclusive customers – instead, people are quite happy to buy from a repertoire of brands.

Could it be that people don’t actually need brands to be differentiated to make purchase ‘decisions’?

Given the incredulity that invariably greets the suggestion that brands really aren’t that differentiated, in Part One, I’ll go through some of the key research exhibits. When so many marketing texts are devoid of empirical data, it is worth wading through just a bit of it. Even though it may feel like some kind of Alice in Wonderland experience.

In Part Two, I’ll consider the implications. While this might feel for some like the marketing equivalent of taking the red pill, this isn’t mere academic navel gazing or planner-y theory.  It has a very direct impact on the work. On how we expect marketing communications to work. On the creative content of what we produce. On our inputs into getting to that work. And how we measure and evaluate its success.

 

The data
It is worth pointing out that there is absolutely nothing original in what follows. All of these patterns in people’s behaviours and perceptions have been long-documented. 

The debt must be acknowledged. All I have done is to follow the paper trail of references and sources provided primarily in the works of Byron Sharp (particularly How Brands Grow), John Scriven, and the incomparable Andrew Ehrenberg.  The only way I could truly wrap my head around some of the implications (and perhaps let go of some my own inherited assumptions) was to get my own hands on as much of the data as possible. 

Dr. Byron Sharp is Professor of Marketing Science at the University of South Australia and Director of the Ehrenberg-Bass Institute for Marketing Science. Professor John Scriven is the Director of the Ehrenberg Centre for Research in Marketing at the London South Bank University. These institutes continue the work of Professor Ehrenberg. 

Both Dr. Sharp and Professor Scriven and have been enormously helpful in helping me better understand the data, were very generous in sharing additional references, papers and studies, and very kind enough to put up with a good number of stupid questions.

So let’s take the red pill and see how deep the rabbit-hole goes.

 

If brand differentiation matters, why do brand perceptions not differ significantly?
If differentiation were successful then we would expect brands to have very different brand image profiles, whether the attributes relate to tangible or intangible nature. 

Brand image tracking is good at giving us the impression that brands are indeed different.

But key factor that distinguishes brands with high brand image scores is not the depth of the positive disposition towards the brand. Big brands simply have more users. And users are more likely to give any kind of response in these surveys. So what distinguishes big brands in these kinds of surveys is not the depth of brand equity, but simply the number of people who have any mental associations with the brand at all.

There is another recurring phenomenon in brand image data. Responses don't just vary according to size of brand. Responses also vary according to the attribute. 

Romanuik and Sharp (2000) term this effect ‘prototypicality’. This describes those attributes which contributing more to category membership than others. That is, they represent those essential qualities that are needed to be part of any category or market.   For example in the beer category ‘refreshing’ is a prototypical attribute in as much as it would be expected that every brand that was a beer would be refreshing.

So usage effect is reflected in the number of responses for each brand and prototypicality is reflected in the number of responses for each attribute.

Here’s an example (Collins, 2011) for the Australian soft drinks market:

ALL TABLES FINAL.001

 

Knowing these two patterns exist in any data set, we can now set about to factor them out, to reveal where the brands really over- or underperform.

To work out what we’d expect a brand and attribute to score we calculate:

Expected: row total x column total/total for all cells

Deviations from what we’d expect happen when there is greater agreement between users and nonusers of the brand:

A positive deviation is typically when non-users also show a strong tendency to associate this attribute with the brand

A negative deviation is typically when users behave more like non-users in that they do not tend to associate this attribute with the brand

And we calculate the deviations from the expected by calculating:

Deviation: actual proportion of responses – expected

Here’s the output:

ALL TABLES FINAL.002

 

If we remove these brands and the attributes that describe them and re-run the analysis for the remaining five brands and attributes, we get this:

ALL TABLES FINAL.003

The deviations are small, averaging +/–2 points. Nobody is really ‘owning’ anything.

Positioning theory argues that brands must develop and maintain  points of differentiation and uniqueness. And who has not been in a strategy meeting which has centered around what our brand can ‘own’? Yet the data repeatedly shows that if a characteristic matters in a category, it is shared by brands. Indeed, brands share characteristics more so than they exhibit marked differences.

 

If brand differentiation matters, why do personality traits not differ between brands?
Ever since Ernest Dichter first suggested that a brand could have a personality, distinguishing between “sombre, utilitarian, thoroughly cleansing character” Ivory and  seductive Camay, marketing and adland have embraced the idea that brand personality is a vital source of differentiation.

Aside from the fact that interpretations of exactly what it is vary, there’s little evidence to suggest that brand personality attributes are successfully at really differentiating brands.

Romaniuk and Ehrenberg (2003) analyzed the data from Y&R’s Brand Asset Valuator.

The full set of 28 BAV personality attributes are:

 

ALL TABLES FINAL.004

The table below shows some of these personality associations for readymade sauces. What’s striking is that all but two of the response levels were 10% or less. The category average was just 7%.

ALL TABLES FINAL.005

 

What this means is that 9 out of 10 buyers of a particular brand did not nominate it as having a particular personality trait.

The responses for users of automobiles is not much different.  The average response across all personality traits is just 12%:

ALL TABLES FINAL.006

 

Brands do not differ much from each other whatever personality traits are ascribed to them. As we can see from the two tables above, response levels between any two brands seldom differ by more than a few percentage points.

It would appear that consumers don’t perceive substantive brand personalities, and that these do not differ significantly between brands.

The table below shows the findings from a study that asked customers of two banks (their names are disguised) to nominate which of two bank brands possessed a series of personality attributes. They could nominate their own bank possessed it, the other bank, or neither.

The response levels overall are quite high, and this may simply because only two banks were asked about. If the study had asked for more, we could expect the proportions to be much lower.

What is striking is that the personalities of each bank mirror each other. Personality attributes are not uniquely owned by any bank . Redbank customers have a very similar view of their bank’s personality as Bluebank’s customers do of their bank. Indeed, the correlation between the proportion of Redbank customers agreeing it has a personality trait and the proportion of Bluebank customers agreeing it has that same trait is 0.98:

ALL TABLES FINAL.007

Despite the amount of time we put in crafting those personality traits for our brands, this evidence challenges our unwavering conviction that if brands are not differentiated on the basis of product features, then their intangible characteristics do. I don’t believe this dimension of brands is utterly irrelevant and redundant though. I think that what we’re trying to describe is not differentiation, but the phenomenon of brand salience – which I’ll get to in Part 2.

 

If differentiation matters, why are people’s brand perceptions not stable?
We’re told that in a world of product parity brand personality is a key means of differentiating our brand from competitors.

If we were successful in clearly differentiated brands, we’d expect to find that we were building enduring brand perceptions. That is, at an individual level what people think of a brand relative to its competitors would not vary dramatically over time.

But it appears that it does. It’s a pretty well-documented pattern. This is for the simple reason that the kinds of surveys that monitor people’s brand perceptions tend to interview a different group of individuals with each wave of interviewing. And these give the illusion that an individual’s brand perceptions are generally fairly stable.  Indeed according to Winchester and Sharp, at an aggregate level over time, brand image remains stable in stable markets – whether it is 3 weeks or one year between interviews, the results tend to be relatively the same.

But the picture changes dramatically when the same group of people are interviewed – and then re-interviewed at a later date.

Across a number of studies, the finding is replicated. Repeat responses – that is the proportion of people attributing an image attribute to the same brand on subsequent interview occasions ranges between 40% to 60%.  And the 50% who do not give the same answer as last time associate the attribute with some other brand.

Here is data from Byron Sharp showing the image stability for Australian financial services brands. Specifically, it looks at the proportion of respondents who agreed with the image statement ‘Would value me as a whole person not just a transaction’ from one survey to another. Only half of the people who agreed with the statement agreed with it again when interviewed again:

 

ALL TABLES FINAL.008

The analysis undertaken by Castleberry et al (1994) of almost one hundred attribute measures for nine product categories in the US and UK shows that the attitudinal repeat rates are mostly well below 100%.

Significantly, this predisposition to change response isn’t the behaviour of a specifically fickle group of respondents, but is a characteristic across all respondents.

What, we might ask, is going on?

The ‘moveability’ as Ehrenberg puts it, of these brand associations “reflects the perceived similarities of the brands. Consumers would be unlikely to respond in this variable way if they saw the brands as generally different.”

Why this happens is not entirely understood, but Dawes (2011) has suggested that this instability may simply indicate that the process by which people link brands to attributes is probabilistic: “In simple terms this means a consumer may have certain memory links between brand A and attributes such as modern or tastes nice or high quality, but whether the link between brand A and any of those particular attributes is actually recalled, either in a purchase situation or a market research interview, has an element of chance.”

Not only should this get us questioning how stable people’s feeling towards brands are, but it should get us questioning all those consumer segmentation studies we’re subjected to. If people change their minds to this extent, what kind of confidence do we have in these studies if their outcome depends on what day and mood we happen to catch our respondents?

The notions of difference and positioning suggest that brand perceptions are something that are permanently ‘fixed’ in people’s minds.  As if people carried correspondence maps in their minds. But life and human nature is far more complicated than a Powerpoint slide and the evidence suggests that brand memories and associations are in fact they are not so stable. They are affected  – as is all memory retrieval – by situation, circumstance, cues, reminders, distractions, and chance.

 

If differentiation matters, why is brand knowledge not evenly distributed?
If brand differentiation were successful, we might expect that ‘brand knowledge’ i.e. the sum total of all perceptual associations held by people about the brand would be evenly distributed across all of a brand’s user base.

It isn’t.

Romanuik and Sharp (2008) have looked at the composition of brand knowledge across populations of buyers, calculating the extent of this knowledge using data for 28 product categories and 208 brands.

They found that half of all brand knowledge is concentrated amongst 20% of buyers and the remainder is spread thinly across the remaining 80% of buyers. This spread is the same for big and small brands, and brands with both high and low equity scores.

So it emerges that a few people will know a lot about your brand. And a great many more people know something about your brand.

Not surprisingly, this reflects the composition of a brand’s user base. A few people buy you a lot. And a great many more people buy you occasionally.

We talk about ‘brand image’ as if this were a universal phenomenon, and as if differentiation were something perceived equally amongst all buyers of a brand. Yet we forget that just as some buyers buy more, so too do some buyers know more. And the ones that know more are unsurprisingly, those who buy more.

However, the vast majority of a brand’s purchases come from the vastly greater number of people who don’t buy it very often and who have a much weaker relationship with the brand. So in contrast to the theory of differentiation which holds that increasing and strengthening brand knowledge is a Good Thing,  the vast bulk of a brand’s purchases will come from people who don’t have very strong or developed views about your brand.

Brand knowledge – like brand purchases – is spread thinly across a brand’s user base. Clearly then, deep brand knowledge is not necessary for the vast majority of people to make purchase decisions.

 

If brand differentiation matters, why do people not believe that brands are different?
If brands were clearly differentiated, we’d expect consumers to explicitly recognize this. We needn’t have to search for implicit differentiation in image attributes. 

So what happens when we actually ask people?

The analysis by Romaniuk, Sharp and Ehrenberg (2007) will raise a few eyebrows.

Here’s their analysis of data from Y&R’s Brand Asset Valuator for the UK soft drinks market. We can see that only about 1 in 10 current customers perceive the brand to be different or unique. About 15-20% state it is either:

ALL TABLES FINAL.011

And here is their analysis of data for the banking sector in Australia. Again, only about 1 in 10 current customers perceive the brand to be different or unique:

 

ALL TABLES FINAL.012

In fact the pattern is replicated across 17 categories, with an average of 17% users saying their brand is different, 10% users saying their brand is different, and an average of 17% saying it is either:

 

ALL TABLES FINAL.013

Millward Brown too have found very similar evidence. According to Nigel Hollis (2011), the BrandZ database (which contains information on over 6,000 brands collected over 10 years)  shows that the proportion of people willing to endorse any brand as “different from other brands of (a specific category)” is “low”. Among those that do consider a brand acceptable (which presumably makes them users), on average just 18% endorse that it is "different from other brands in the category". In other words, 82% do NOT endorse that it is "different from other brands in the category".

So when asked if brands are different, most people say they don’t believe they are. Now asking people direct questions in research surveys can undoubtedly be a reliable way of generating false trails, red herrings, post-rationalizations, bullshit, and downright lies. This may at least in part be a case of people asserting that they’re not capable of being influenced by marketing. But the number of people who say their brand is different is so low, that surely something else is also going on. And if brands were so obviously and self-evidently different, it would surely not be unreasonable to expect to see this reflected in people’s beliefs about brands.

 

If differentiation matters, then why do loyalty levels not differ much between brands?
The argument for differentiation tells us that it makes it hard to substitute with other brands. And that as a result, well-differentiated brands enjoy a more loyal customer base.

However, irrespective of the category one looks at, despite the rhetoric of fans, zealots, engagement and the like, loyalty levels in fact differ little between brands.

Here for example, is data from John Dawes’ analysis of the UK sportswear market:

ALL TABLES FINAL.014

From this we see that brands vary considerably in market share – Adidas has a market share of 33%, Umbro a share of 9%.

But there is much less variation in loyalty – so while it is almost four times the size of Umbro, the gap in loyalty is less significant: Adidas captures 68% of requirements, and Umbro captures 50%

Here is an analysis of six years’ of TNS buyer data in the coffee category from Charles Graham. We can see that although penetration and share values move in line, average purchase per buyer for any brand remains roughly constant:

 

ALL TABLES FINAL.015

The theory of differentiation holds that well-differentiated brands succeed in being uniquely appealing to consumers, and thus that differentiation is a key means of driving superior brand loyalty. An examination of purchase data however, makes it difficult to sustain this argument.  For whichever category we look to, the pattern is the same.  There is a wide spread in market shares. But large and small brands have roughly equal and predictable loyalties. All that effort in the name of differentiation and positioning does not translate to superior behavioural loyalty. Let alone ‘loyalty beyond reason’.

 

If differentiation matters, then why do brands not have unique and discrete user bases, but share their customers?
The fact that markets and brands are characterized by varying levels of size not not differentiation-fuelled loyalty, is further underlined by the fact that people are polygamous in their loyalties. Again, if people are happy to spread their purchases across competing brands,  and find competing brands to be acceptable substitutes, then surely differentiation must have failed.

Going back to Dawes’ analysis, look  for example, at how sportswear buyers spread their purchases:

ALL TABLES FINAL.016

 

So here we see that:

Adidas buyers allocated on average about 68% of their requirements to Adidas,14% to Nike, 8% to Reebok, 5% to Umbro

Nike buyers on average allocated 63% to Nike, 18% to Adidas, 9% to Reebok, 6% to Umbro

And here is Dawes’ analysis of the buying patterns for toothpaste brands in the UK:

 

ALL TABLES FINAL.017

In other words, brands share their customers with other brands, and they do so roughly in line with their market shares.  As Dawes notes: “It is difficult to think these brands are bought due to special or enduring positions when their customers freely buy competitor brands as well.”

If marketing succeeded in truly differentiating a brand, we’d see in the patterns of buying behaviours that people found other brands to be poor substitutes for it. However, irrespective of the category we examine, we see that the vast majority of buyers are in fact not loyal to a single brand.  Devoted loyalty – borne of the belief that other brands just aren’t as good, or just aren’t the same – does not exist. Instead, consumers are perfectly happy to buy from a repertoire of brands. So much for loyalty-building power of differentiation.

 

If differentiation matters, then why do different brands not appeal to different kinds of people?
William Moran (1990) has railed against the market research community’s devotion to multivariate analytical techniques, arguing that these “identify confusing, overlapping and totally ambiguous clusters of people… which more often than not, are not even related to any common market behaviour. These clusters quickly become the subject of a parlor-game mentality as the research challenge is to assign meaningful-sounding names to them.”

Anybody who’s had to endure the torture of those consumer pen profiles with their Cautious Cathies, Profligate Pips, Spendthrift Smittys, Desperate Dans and the like, and has had to struggle to relate those to anything resembling real human beings will share Moran’s frustration.

Indeed the suggestion that ‘our’ brand with its unique positioning appeals to a distinct and different group of consumers runs through much of marketing’s assumptions and sacred texts.

If differentiation succeeded, it wouldn’t be unreasonable to expect different brands appeal to different types of consumers. 

However, the analysis conducted of TGI data conducted by Kennedy, Long, and Ehrenberg (1990), comparing the user-profiles brands in each of some 40 industries, demonstrates that users of directly competing brands hardly differ in their profiles. 

Two batches of TGI data were analysed, with the same outcome. A first batch was for all brands in 13 industries with 100 attitudinal variables; the second hatch of data, some two years later, compared the top 10 brands in 30 other industries, with over 200 attitude variables.

The profiles of the different brands in each product category were compared  against the profile of the average brand in that industry. 

The findings are striking:

Overall, the individual brands percentage profiles deviated from each other by an average of 2 or 3 percentage points, which as the authors note “is… in effect zero”

Only around 8% of the individual deviations were more than 5 percentage points, and even these larger deviations averaged at only about 9. 

Just 2% of individual deviations were 10 points or more. 

So what emerges from this analysis is that when it comes to their user-profiles, brands rarely differ from the average brand in their category, and when they do so it is not by very much.

This, the authors of this analysis conclude: “Brand segmentation generally does not exist – substitutable brands usually compete in what for them is a single unsegmented mass market, whatever its overall structure may be.”

We’re accustomed to the rhetoric of ‘our consumer’…. the ‘Hellmans consumer’, the ‘Budweiser consumer’… the ‘Huggies consumer’ and so on. But these are fictions. Or worse, they are notions of corporate vanity. For despite the amount of time and money poured into consumer segmentation work, it would appear that within a given category or sub-category, different brands and their supposed ‘brand positionings’ do in fact not appeal to different kinds of consumers. They appeal to and are bought by the same people. It is categories that are segmented. But not brands. Again, so much for differentiation and positioning.

 

If brand differentiation matters, why do people not exercise a high degree of brand evaluation
If differentiation mattered, we could expect consumers to invest both attention and time making purchase decisions. As Kapfefer (2004) has written: “Positioning is a crucial concept. It reminds us that all consumer choice is a comparison.”

But people don’t spend that much time or that much attention to making purchase decisions. 

In contrast to the implications of Kapferer’s assertion, Alan Swindells (2000) has written eloquently of what he calls the ‘myth of consumer choice’:

“For any individual consumer the world is made up of a myriad of vague ideas, thoughts, images and feelings. They come together in a loose network of meanings which shape consumer behaviour. Some of it is organised and structured but most of it is disorganised and unstructured. Some of it is thought about, rationalised and is articulate while most of it is vague, inarticulable and taken for granted. In any consuming society brands are a part of this taken for granted world.”

Swindells reminds us that many of the mechanics of consumption are in fact invisible to consumers themselves, that consumer ‘decision making’ is in fact exceptional occurrence, and much of marketing communications is processed in a relatively passive way, rather than actively thought about.

This perspective has been echoed in the more recent work of Robert Heath on the role of low-involvement processing in learning about brands.

Heath distinguishes between high- and low-involvement processing.

High involvement processing requires high attention. Not only can we not maintain if for long periods, but brands simply aren’t important enough to demand it: “If we were to process the price at high involvement we might form a whole raft of new conclusions and interpretations about it: we might question how it compares to the prices of other similar products, make a judgment about whether the difference is important, decide whether the product is worth it, and work out if we have enough money to buy it.”

Instead, we employ low involvement processing. While it happens consciously, it takes place at very low levels of attention. Thus we ‘note’ things about our environment without actively trying to learn and store information. 

As Heath (2000) puts it: “Low involvement processing is of particular importance to brand learning because, in circumstances where consumers regard brand information as being of little importance, there is little likelihood of high attention levels. Low involvement processing means certain brand information is processed and stored whether or not the consumer wants it to be.” 

The triviality of ‘learning’ about and ‘deciding’ between brands is difficult to capture in consumer research precisely because people are not very aware it’s happening.  

In timeless Testing to Destruction Alan Hedges provides us with a vivid pen portrait of a consumer and her contacts with the floor polish market market through an imaginary week. Given the ability of so much consumer research to exaggerate the consciousness of consumer decision-making and brand-learning it’s worth quoting from at length: 

“Let us suppose that this market (for simplicity) contains only three brands. Let us imagine they are called Lewis’s,Whistle and Glo. New Glo is our brand…she has never tried Glo. Her impressions which are vague and hazy in the main, are a compound of her own experience, odd remarks from neighbours, past advertising seen or half seen, the look of the products and their packages… She gets out the Lewis’s and sets to work. She sees the pack. She doesn’t really look at it, it is so familiar… but perhaps she half notices an odd word, and another flake of meaning joins the mixture in her mind…. She puts the polish away, noticing that the pot is nearly empty, and making a mental note to get some more… At lunchtime, flicking through a women’s mag, her gaze wanders over an ad for Whistle for a few seconds. She doesn’t think about it, but somewhere her mind ticks up the impression that she has heard a lot about Whistle recently… Later that afternoon she goes shopping. On the way she passes one of our new Glo posters. She doesn’t even glance at it, although later in the week she will notice the Glo on the supermarket shelf and wonder for a split second where she saw it before…”

Given that it didn’t exist in 1972 when this was first published, the internet is absent from the account. But you get the picture. And it’s worth reading in its entirety. It’s a truer portrait of consumer behaviour than all those achingly trite consumer portraits and typologies we have to endure.

And there’s empirical evidence for this.  Consider for example, that a study of wine shoppers in Australia found that the average time spent in front of a shelf in a retail outlet was less than a minute, and the total time browsing in the store was about four minutes. Involved shoppers spent up to 15 minutes buying wine, but these were the minority of wine buyers. 

We might hate to admit it. But for all their pleasures and usefulness, brands for the most part are part of our ‘taken for granted world’. It’s a world characterised for the most part by routine and triviality. 

And for good reason. If people don’t think brands are that different or that unique (as is evidenced both by their brand perceptions and buying behaviours) why should they devote previous time and energy to something so trivial?

While the language of differentiation and positioning suggests that people pay attention to and care about the differences between brands, people don’t actually spend much that much time and that much attention making purchase decisions. And – as is evidenced by the fact that most people aren’t devoted loyalists, their willingness to buy from a reprtoire of brands, the fact that in general they don’t regard brands in a given catgeory as offering anything particularly different from  all the other brands – they do so because they know that in general brands aren’t that different.

 

Back from the rabbit hole
All this probably feels like first stumbling upon the existence of our quantum universe – a place that plays by very different rules. Very, very weird. The evidence is admittedly not complete. Significant progress in our understanding is made difficult by the reluctance of clients to publicly share proprietary data. However, the existing data is plentiful (and there is more where this came from), and it should at the very least cause us to question our unwavering faith in the notions of differentiation and positioning.

Now if people don’t need brands to be that differentiated in order to buy them, it rather begs the question what the role of creativity is. My own view is that if the implications of this data are to be believed, then things get a lot more more interesting. But that’s for Part 2.

 

 

 

 

Sources

Jennifer L. Aaker, ‘Dimensions of Brand Personality’, Journal of Marketing Research, Vol. XXXIV (August 1997)

Neil Barnard, Andrew Ehrenberg, ‘Advertising and Brand Attitudes’

Tim Bock, ‘Improving the display of correspondence analysis using moon plots’, International Journal of Market Research, 53, 2011

S.B. Castleberry, N. Barnard, T. Barwise, A. Ehrenberg, F. Dall’Olmo Tiley, ‘Individual attitude variations over time’, Journal of marketing management, 10, 1994

Martin Collins, ‘Analyzing Brand Image Data’, Marketing Learnings 8 (March 2002)

Martin Collins, Response to Tim Bock, ‘Improving the display of correspondence analysis using moon plots’, International Journal of Market Research, Vol. 53, Issue 5, 2011

John Dawes, ‘Brand loyalty in the UK sportswear market’, International Journal of Advertising Research, vol. 51, No. 4, 2009

John Dawes, ‘Predictable Patterns in buyer behaviour and brand metrics: implications for brand managers’, in M.D. Uncles ed., Perspectives on Brand Management

Andrew Ehrenberg, Neil Barnard, John Scriven, ‘Differentiation or Salience’, Journal of Advertising Research, November/December 1997

Andrew Ehrenberg, ‘Repetitive advertising and the consumer’, Journal of Advertising Research, Vol. 40, No. 6, November/December 2000

Andrew Ehrenberg, ‘What Brand Loyalty Can Tell Us’, Admap, October 2004, Issue 454

Andrew Ehrenberg, Neil Barnard, Rachel Kennedy, Helen Bloom, ‘Brand Advertising As Creative Publicity’, Journal of Advertising Research: Vol. 42, No. 4, July/August 2002 

Peter Field, ‘The creation of buzz and fame’, Admap, April 2008, Issue 493

Charles Graham, ‘Is your brand’s share trying to tell you something?’ Market Leader: Quarter 1, 2010

Robert Heath, ‘Low Involvement Processing – a New Model of Brands and Advertising’, International Journal of Advertising, Vol. 19. No.3, 2000

Robert Heath, ‘Low-Involvement Processing Part one: A neuroscientific explanation of how brands work’, Admap, March 2000 

Robert Heath, ‘Low-Involvement Processing Part 2: Seven new rules for evaluating brands and their communication’, Admap, April 2000

Robert Heath, ‘How The Best Ads Work’, Admap, April 2002, Issue 427

Robert Heath, 'Ah Yes, I Remember It Well!', Admap May 2004, Issue 450

Robert Heath, ‘Emotional Persuasion’, Admap, July/August 2006, Issue 474

Robert Heath and Paul Feldwick, ‘Fifty years using the wrong model of advertising’,  International Journal of Market Research Vol. 50, No. 1, 2008

Alan Hedges, Testing to Destruction

Nigel Hollis, ‘It Is Not a Choice: Brands Should Seek Differentiation and Distinctiveness’, Millward Brown Points of View, April 2011

Jean-Noël Kapferer, Strategic brand management: Creating and sustaining brand equity long term, 2004

Rachel Kennedy, Stephen Long and Andrew Ehrenberg, ‘Competitive Brands' User–Profiles Hardly Differ’, Market Research Society Annual Conference, 2000

McDonald and John Scriven, ‘Why marketing needs marketing science’, Admap, October 2006, Issue 476 

William Moran, ‘Brand Presence And The Perceptual Frame’, Journal of Advertising Research, October/November 1990

Rosser Reeves, Reality In Advertising, 1961

F. Dall’Olmo Roley, Andrew Ehrenberg, S.B. Castleberry, T. P. Barwise, Neil Barnard, ‘The variability of attitudinal repeat-Rates’, International Journal of Research in Marketing, 14, 1997

Christoper Riquier, Byron Sharp, ‘Image Measurement and the Problem of Usage Bias’, 26th EMAC Conference, The University of Warwick, Warwick Business School, U.K., 20-23 May

Jenni Romanuik, Byron Sharp, Andrew Ehrenberg, ‘Evidence concerning the importance of perceived brand differentiation, Australasian Marketing Journal 15 92), 2007

Jenni Romanuik, Byron Sharp, ‘Where knowledge of your brand resides: the Pareto share of brand knowledge’, in Report 44 for Corporate Sponsors, 2008, Ehrenberg-Bass Institute for Marketing Science

Jenni Romanuik, Byron Sharp, ‘Using known patterns in image data to determine brand positioning’, international journal of market research, Vol. 42, No.2, 2000

Jenni Romanuik, Andrew Ehrenberg, ‘Do brands lack personality?’ Report 14 for Corporate Members, March 2003

Jenni Romanuik, Byron Sharp, ‘Conceptualizing and measuring brand salience’, Marketing Theory, Volume 4(4), 2004

Francesca Roley, Lucy Rink, Patricia harris, ‘Patterns of attitudes and behaviour in fragmented markets, Journal of Empirical Generalisations’, Marketing Science, Volume 4, 1999

Byron Sharp, How Brands Grow: What Marketers Don’t Know, 2011

Byron Sharp and Kate Newstead, ‘Loyalty is not the Holy Grail’, Admap, September 2010

Jaywant Singh, Chris Hand, Hsin Chen, ‘Differentiation in a branded commodity category: Tapping into behavioural data’, Kingston University

Alan Swindells, ‘The invisible mechanics of consumption’, Market Research Society annual conference, 2000

Julian Viecli, Robin Shaw, ‘A Model of Brand Salience’, in Mark Uncles, ed. Perspectives on Brand Management, 2011

Maxwell Winchester, Byron Sharp, ‘Arguments against brand positioning’, Marketing Science Centre, University of South Australia

 

Advertisements

14 comments

  1. Martin

    Thank you, Will, that’s very kind. But the truth is I am merely recycling other people’s data, thinking, and work…. the thanks really should go to them.

  2. Will

    I think some of your posts have doubled the sales of Byron Sharp’s work, certainly – but it’s very handy to have the differentiation myth busted with a collated post of all of the evidence. So – thanks again.

  3. Carmen

    down the rabbit hole indeed..thank you for critical reasoning, great data and great research examples .. it is a rare combination (looking forward to part 2)

  4. @eriksebastian

    Magnificent! ..my mind just went through a paradigm shift.
    I’ve long questioned a lot of my day to day work and this post articulates my criticism far better than any other I’ve read to date.
    Very much looking forward to post number 2.

  5. Tofan

    Thanks a lot for the post Martin, this really open my mind of how I usually see think about the brand development.

  6. Chris Alders

    Hi Martin, I am currently investigating a similar hypothesis but with a different angle of approach. ‘In short: Brands mostly exist in the subconscious network of the brain, therefore our focus should be on finding the ‘receptors’ that trigger the interest of that subconscious network’. Therefore shifting from away from traditional value-based branding.
    I read your article with great interest and I am looking forward to the conclusions you draw. Cheers, Chris Alders

  7. SomeOfJess

    The fact that I own 4 identical pairs of stretch leather trousers is further evidence that sameness is in fact more important when making purchasing decisions than differentiation. Or that I am obsessive. Or both.

  8. Lestermarky

    Thanks for sharing this Martin, I think it’s great that you’ve brought this work to a wider audience.
    This has definitely made me think a lot, a few questions I’m tinkering with in my head before I’m willing to give up on differentiation:
    1) Could a lack of perception of differentiation among consumers simply be evidence of either ineffectual marketing practice, or that differentiating brand associations are largely unconscious?
    2) A number of the pieces of evidence above suggest differentiation is a weak force (instability of perceptions, concentration of perceptions among loyalists, etc.) is this the same as proving that it does not have large effects? In choices people care very little about, isn’t it conceivable that both things are true?

  9. Stassnl

    Hi Martin, I’m a also a big fan of Ehrberg and Sharp.
    There is only one thing they are not taking into account. That is the impact of volume per purchase. This is off course less relevant for buying cars, but for a lot of FMCG brands this could have serious impact on loyalty. I will send you a study on this.
    Your comments on differentiation are also very familiar. Most marketeers focus on being different, but the fail in being relevant. And in my opinion that is what it all about

  10. adrian

    hey martin. very stimulating material. i’m currently working with a client that applies ehrenberg’s ‘strategies of desire’ and ‘salience theory’. but i still have a few question that trouble me. i’m trying to link this material with one of your previous ones, maybe you can help me make the right connections
    SO
    – differentiation is rather a myth – and so is brand loyalty
    – success depends a lot on light users
    – the way they perceive your brand (as individuals) is highly unstable in time
    – and decision-making happens in a low involvement mode
    – consequently the buying decisions are influenced by a huge and unpredictable variety of stimuli

  11. adrian

    BUT
    – in a previous post (“infidelity, inertia and unsegmented markets”) you were mentioning how difficult it is to alienate consumers and that marketers shouldn’t worry about it
    – and that there’s a lot of inertia in consumer behavior, people tend to make the easiest, familiar choice, mainly due to behavioral anchors, hardwiring being responsible for habitual behavior
    The questions are:
    1. if most decisions are habitual, which are the ones that tend to easily react to new stimuli? or maybe the inertia only refers to the difficulty for brands to move both in and out of a repertoire? and once inside, the choice between brand A, B or C becomes hard to predict, control and influence?
    2. if the single individual’s perception about brands changes so fast as a reaction to stimuli, do we have any data proving that shopping behavior is sensitive to the slightest changes in perception? if true, then advertisers might have a much bigger responsibility on their shoulders
    3. and if true, then the marketer’s paranoia that the wrong ad will ‘alienate’ consumers – isn’t it justified? the ‘alienation’ might not be permanent but a short-term, behavior-altering ‘alienation’ can potentially make lots of difference in front of the shelf, isn’t it?
    Cheers,
    Adrian

  12. Fredrik Hallberg

    Hi Martin!
    This was an excellent post that propelled me to dig deeper into the rabbits hole. Support for these findings can also be found in related articles building on chaos – and network theory and from some of the recent nobel price winners in behavioural sciences. However on a more simple level the following quote from your post raises a simple but important question.
    “The vast majority of a brand’s purchases come from the vastly greater number of people who don’t buy it very often and who have a much weaker relationship with the brand.”
    Doesn’t this contradict the most basic and most common cited characteristic of the pareto principle i.e 80% of your profits come from 20% of your customers or that 80% of your sales come from 20% of your clients?
    Since most of the empirical evidence which gives aditional support to your article is built upon and powerlaws, there seems to be an inherent contradiction here.
    Where powerlaws rein the bigger gets bigger and earns its loyalty not from high engagement or high involement but simple because it is already popular.
    If powerlaws gives support to the article it cannot at the same time be irrelevant for the findings. Or can it? Please help me out here.
    Nevertheless, this was an excellent read. Best post i’ve read in recent years by anyone.
    Br
    Fredrik

  13. mweigel

    Fredrik,
    Glad you enjoyed, thank you.
    I am indeed contradicting that piece of inherited wisdom!
    And the evidence is plentiful:
    http://mweigel.typepad.com/canalside-view/2010/12/sooner-or-later-though-you-always-have-to-wake-up-corporal-jake-sully-avatar-the-received-wisdom.html
    http://mweigel.typepad.com/canalside-view/2012/02/the-participation-paradox-how-to-survive-it-how-to-prosper-from-it.html
    http://mweigel.typepad.com/canalside-view/2011/07/iconic-brands-the-illusion-of-love-and-why-size-matters.html
    The data actually confirms your point about power laws. Brand don’t get big because of loyalty.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s