<%@ Language=VBScript %> <% CheckState() CheckSub() %> Differentiation or Salience
Journal of Advertising Research

Journal of Advertising Research

Advertising Research Foundation
641 Lexington Ave., New York, NY 10022; (212) 751-5656

November/December 1997

Differentiation or Salience

Andrew Ehrenberg, Neil Barnard and John Scriven

Why is the market leader often ten times as big as the tenth biggest brand? The evidence says that it is not due to brand differentiation. Each brand's customers do not look at their brand as being very different. Instead, what differs is the number of people to whom each brand is salient.

In this paper we first critically review the notions of brand differentiation and of major sustainable competitive advantages. These are widely thought to be crucial. However, the users of competitive brands do not in fact see them as very different.

Next we elaborate on salience. This is the common factor in how many people are aware of the brand (by any measure), have it in their consideration set, regard it as value-for-money, buy it or use it and so on. When these numbers correlate across brands some brands scoring for many people on all the measures and other brands for only few that reflects the brands' different levels of salience. In practice this then correlates with distribution, shelf-space, market shares and sales.

Third, we consider the implication for advertising. This has to focus on for how many people feel each brand is salient at all rather than on either what it is they feel about the brand or how strongly they feel it.


It is often said that brands need to be differentiated in order to be bought consumers must have a reason.

Differentiation can take many forms: from the clear-cut physical or functional (a PC with a fax facility), through the less distinguishable (two kinds of tomato soup), the barely noticeable (eg, 'More nourishing'), the emotional (a mood or aspiration), to the 'distinguishing but irrelevant' (blue packaging, or stripes in dentifrice).

Differentiation which is successful in terms of sales asks to be copied, and generally it is (eg, PCs with Pentium chips; shampoos for oily hair). There are only rare exceptions (eg, patents). This advantage which is of sales importance is seldom sustainable. And any advantage which is sustained is seldom important.

Hence there tends to be ceaseless talk of innovation, to find new ways for brands to be differentiated.

The more realistic picture has, however, been summed up by Professor Tom Robertson (condensed):

Even a 'revolutionary' product change mostly differentiates only briefly (eg, laundry soap being replaced in turn by detergents, biologicals, greens, high-density liquids and then also H-D powders).

'Me-tooism' remains the dominant force in competition. Being competitive means cashing in on one's competitors' successes. Such imitation is not just restricted to minor product developments (baking soda in toothpaste or guaranteed-money-back long-term investments) but applies even more to the major product characteristics (all mainstream cars have to be speedy, safe, fairly economical, washing powders wash white or whiter, etc). This is not new.

The trends in our technology lead to competing products becoming more and more alike. (1920s guru James Webb Young)

Successful product innovations tend to benefit the consumer. The producer may hang on to some early-mover gains ('they have much greater long-term success than the [actual] pioneers' Golder and Tellis, 1993). But innovation is not the stuff of effective long-term competition, except defensively. One probably cannot afford to be left behind. Innovation does not justify our large on-going advertising budgets a technical advance like air-bags soon becomes old-hat and calls for little further publicity just some occasional reminders perhaps that 'We've got them too').

Differentiation between and within brands

This general similarity of brands makes people say that product differentiation must be very difficult. But that is not so. Striking differentiation is easy and everywhere. But the differentiation occurs within each brand, eg, telephones with or without memory buttons; cars with two, three, four or five doors; different pack-sizes and/or flavors; variable or fixed-interest loans; and so on.

In contrast, keeping brands differentiated from each other is difficult. Financial services are a commonly quoted example. They are typically accused (and also accuse themselves) of being unable or even 'unwilling' to set themselves off from their competitors' offerings. However, that is not because differentiating their insurance and loan policies, etc, is hard but because good ideas are copied.

The objective, of course, is to gain competitive advantage by building sustained customer loyalty with products or services meeting quite precisely the demands of closely-defined markets. (Sir Colin Marshall, 1995)

The lack of brand segmentation is therefore not always recognized: British Airways (Sir Colin's company) certainly segments travelers by price (with a 14-times factor between APEX-economy seats and transatlantic first class) and also by their destinations. But so do almost all other (big) airlines. There is little competitive differentiation, except crucially by being available (that that airline is usually almost the only one flying precisely that route at about that time).

Hence airlines have to distinguish the virtually undifferentiable by their labels (AA, BA, TWA, UA, VA, etc); by the amount and perhaps the creativity of their publicity (mostly ads featuring airplanes, stewardesses, sunsets, beaches or price discounts-although BA/Saatchis were different); by painting their Jumbos differently; and by claimed differences in service which might be noticeable to those frequent fliers that do comparison shopping. Branding by 'distinguishing but irrelevant attributes' to make otherwise undifferentiated brands distinctive is still essential: 'A feather can tilt the balance' (Sutherland, 1993).

Distinctive brands

A brand for us here means a named version of a given product or service. Successful branding involves taking responsibility for enduring quality. There are variations on this theme, for example:

Sometimes a house-name is used as an umbrella for just about everything made by a manufacturer (eg, Heinz, Ford, Kellogg's, etc) and sometimes not (eg, P&G, Unilever, GM, etc). There are different ways of making good coffee.

Sometimes the house-name is primarily what is publicized. A specific variant a flavor, for instance may be featured but is just used creatively as an example of the brand. At other times the variant (eg, a particular car model) is the prime advertising theme.


Salience is broader than any single measure of brand performance. It depends on virtually all the different possible measures of performance correlating. Compared with Brand B, if Brand A has more salience than B, it has more people who:

For directly competitive and substitutable brands, all the different measures tend in practice to correlate well. Being in the consumer's consideration set is perhaps the simplest single conceptualization of 'salience.' The measures then also go with consumers' propensities to buy the brands and with their actual brand buying and usage (ie, with the actual repertoire) and with sales; and hence in turn also with how many users say subsequently that they like the brands ('Familiarity leads to liking').

Brand A being 'salient' to more people than B is then usually also linked with whether A has, if anything, wider distribution, more shelf-space and display, more sales-people, more promotions, more word-of-mouth, more media mentions, more advertising and probably bigger absolute profits.

There are also remarkable feedback loops and marketing-mix synergies in these relationships. The bigger brands are so much bigger because the promise of more advertising had slowly, and up to a limit, led to more shelf-space and display, to higher and more profitable sales, hence to bigger advertising budgets (and possibly less price-cutting) and to more shelf-space, etc, again. The benign spiral includes that the more marketing activity and display there is for Brand A, the more noticing the brand can again gain. Just seeing the brand around can reinforce its memorability and thus again its salience.

The evidence is that customers of Brand A are usually little more 'loyal' to A than customers of Brand B are to B (by any measure of loyalty see JOAB 1 and 8 and references there). If A is much bigger than B it has far more customers who are loyal. Typically, for instant coffee brands in 1992 (Ehrenberg, 1997) the top brand Folger's had 11% of US households buying it loyally, on average 3.3 times, and the much smaller Brim had 0.2% buying on average 2.1 times. Compared with the 50-fold difference in penetrations, these are rather similar degrees of repeat-loyalty, with a predictable Double Jeopardy trend (3.1 to 2.6). Similarly, the two brands' 'Share of Category Requirements' were also very similar at 48% and 42%. The same picture was obtained for the other brands and also ten years earlier for US instant coffee, when brand shares as well as total market size were very different, and also in 50+ other product categories over the years.

'Salience' is therefore not about how strongly the users of a brand feel about it not ten times as strongly about A than about B (eg, ten times 'Kinder to the Hands,' etc). For some one hundred attribute beliefs about brands in nine US and UK product categories, 48% of users of the top brand say 'Kind to the Hands,' etc, about it and 49% about the smallest of the eight itemized brands (with less than a quarter of the number of users see JOAB 7 and 8).

Nor does the much bigger Brand A have to be 'valued' much more by its many users than Brand B is by its fewer users (apart again from the typical but relatively small 'Double Jeopardy' trends with market share JOAB 1 and 7). There is also no systematic sign of big differences in 'brand equity' or the like, or in 'strengths' which can be associated with particular brands. As we see the evidence, there are no strong brands and weak ones but only big brands and little ones.

Minor differences

Competitive brands do, however, differ in what we call 'minor' functional properties. Possibly also in emotional features they differ (eg, some people say that brands possibly have different 'personalities'). From an advertising point of view the point is that minor functional differences seem seldom or never to feature in the brand's advertising. They are noted by consumers after the brand has been tried rather than before.

An example is that Muesli A and Muesli B may differ in their degree of sweetness, in having nuts rather than almonds, in the amount of raisins, etc. Muesli A may well be preferred by some consumers and B by some others. In the longer term, that can be competitively important to sales. But this is a result of product development and hardly ever featured in the advertising, either as new news initially or for reinforcement.


What then are the advertising implications of this widely established double finding that:

  1. the number of people for whom the brand is salient differs greatly from brand to brand.
  2. what brand-users feel about their brands is mostly much the same.

The answer is that advertising (and marketing) could change the first (because it can vary so much anyway) but hardly the second (because it generally doesn't). If Brand A has ten million people to whom it is salient, and Brand B one million, A will be bought ten times more.

Advertising a well-established brand to experienced consumers can seldom, if ever, imbue it with new, highly differentiating attributes, or with more loyalty or liking than before, because these things are features of the product type and generally vary little from brand to brand anyway. Nor does it seem a priori likely that consumers' attitudes can be changed readily just by seeing a few advertising messages. In contrast, simply to publicize the brand well can lead to more people being aware of the brand, feeling assured about it, bringing it into their consideration set, etc. That is in line with these measures differing greatly between large brands and small ones. They must, therefore, already have been influenced differentially by the marketing mix in the past, so that there is scope for doing it again.


The informative role of advertising is something very different: telling consumers about a new brand, a price-cutting 'sale,' or a new feature such as airbags or a left-handed corkscrew. It is telling people something new, instead of trying to get them to feel differently about something (a brand) that is already familiar.

But new news for established brands or even a new brand altogether is rare and requires relatively little informative advertising. The new feature (eg, a new flavor, bottle-top or airbags) is soon noticed and assimilated by users of the product. It therefore needs little on-going publicity (extensive repetition of mere facts even becomes irritating and rightly so). The bulk of tonight's TV advertisements will therefore not be informative but will publicize established and mostly 'well-known' brands.

'Here I Am' publicity

The brand's name and package are the distinctive features by which it is known. Hence the overwhelming importance, we think, of publicizing the brand name.

Many ads, and perhaps most, do just that, either as:

The advertising copy has to be distinctive to be noticed and to leave memory traces. But in practice this is not in order to attach a different 'message' or 'image' to the brand (or certainly not necessarily). To change peoples' attitudes or feelings greatly and/or lastingly is, we think, quite widely accepted to be very difficult or near impossible as stressed by Gordon Brown, for instance, although it is also still widely talked of as the goal (an irony pointed out by Bill Moran). Nor should the aim be to make the advertising itself memorable: With few exceptions (eg, well-established slogans) few people recall a brand's advertising from two or three years ago, although one may recognize the old ads again when one sees them. ('Coke is it' was ten years ago; more recently 'Always Coca Cola.')

Instead, the aim is, or should be, for the advertisements to help publicize the brand itself, to leave idiosyncratic memory traces for it, and, possibly, longer-term memory associations with it.

The closer and more substitutable the brand is with its competitors, the easier it is for creative and effective 'Here I Am' publicity to maintain, reinforce and/or nudge the brand's salience and consumers' purchase propensities. There are then no differentiating functional (or emotional) values which the advertising would have to overcome (like greatly preferring shredded wheat to a muesli, say). Only consumers' habitual brand-choice propensities (and their consequent feelings for Brands A or B: 'I use it therefore I like it') may inhibit moving between directly competitive brands.

The less differentiated brands are, the more readily can advertising nudge choice behavior. But the greater, therefore, also is the need for defensive reinforcement of your brand.

To use this explains great scope for advertising to try to nudge your brand's impact and the even greater need to defend your customers from your many competitors' encroachments. Most advertising and marketing for established brands seems in practice to be geared to counter the competition and defend one's market share from its close competitors.


We have argued here that the sales success of a competitive brand is primarily due to how many consumers regard it well, or 'well enough,' or see it as 'salient' (eg, ten million versus one million). Precisely how well they regard the brand, and why they say so in typical attitude surveys, usually differs relatively little from one competitive brand to the next. Otherwise the brands would hardly be very substitutable or competitive in the first place. Any differences which do occur are at the margin and can therefore hardly account for the vast differences in market share that are so common.

The currently popular terms 'Strong Brands' and 'Brand Equity' might seem to say the opposite. They imply that brands have very differing values. But the terms have never been tangibly defined.

What is Brand equity anyway, and how do you measure it? (Paul Feldwick, 1996)

Brand equity still means different things to different people. (Katherine Jocz [in Sood, 1995])

Coca-Cola and Pepsi-Cola do not have to be seen as markedly differentiated in order to be worth something on the stock market, this being dominated by their more or less steady market shares over a succession of years. Brands which are suppose to be 'strong' seem in practice only to be big. And the main factor in a brand being- big is, as we have said, how many people regard it as salient and buy it, and not that they typically think one brand to be very different or much more valuable than the other.

How many people find a brand salient (eg, in their consideration set) can be aided by creative 'Here I Am' publicity making the brand more widely known in attention-getting ways which also leave a long-term memory trace for the brand and potential associations. 'Telling a good story well' is what matters, more than just which particular good story it is, apart from its broad consistency with the brand's history. (If there was only one possible approach, there would hardly be so many different agency pitches.)

In contrast, and despite all the rhetoric of differentiation and innovation, important differentiation between one brand and another is seldom sustainable for long. Nor indeed would that always be desirable: effective differentiation would merely segment and hence deliberately limit the market (eg, focusing on left-handed corkscrews, let alone exclusively on the young or with-it). In our view the essence of most competitiveness is not to be different but to compete with success by being good at imitating that, and perhaps being marginally better. It is a case of survival of the fit enough, not of the fittest.

This is not because differentiation is difficult. Most brands carry a wide range of product variants or line extensions (different sizes, flavors, etc). But competitive brands then still seek to resemble each other closely in all these respects, as in most others.

When brands or specific brand variants are very similar, publicity can work more easily to help make a brand salient to more people. In fact, advertising seems to have most scope when brands are similar, so that their advertising is almost the only variable factor that does distinguish the brands as such.

Our emphasis on sheer salience rather than brand image or values may be surprising to many in marketing, given the traditional and still on-going marketing rhetoric. But the empirical test is that highly sophisticated 'Here I Am' publicity simply to boost an established brand's salience or its awareness in the broadest sense is where most advertisements are in practice focused.

Some advertising does feature a particular claim or a product feature. But this is usually not to establish that feature in its own right but occurs only in order to tell a creative story about the brand.

This paper is based on one of eight reports in the two-year South Bank project on 'Justifying Our Advertising Budgets' (JOAB). This aims to help companies establish better how their large-scale advertising can be justified. Technicalities and references are given in the Notes section.

JOAB participants in the United States and Great Britain have included: BMP DDB Needham, British Gas, BP, British Telecom, CBS, Coca-Cola, Colgate, General Motors, IDV, IPA, Philip Morris, Procter & Gamble, Prudential, Publicis, Scottish Courage, Seagram's', Shell, Unilever and United Biscuits. We are also indebted to Helen Bloom, Pam Mills, Colin McDonald and other colleagues for helpful comments.


Agres, S. 'Leading and Lagging Indicators of Brand Health.' In Brand Equity and the Marketing Mix: Creating Customer Value, S. Sood, ed. Cambridge, MA: Marketing Science Institute, 1995.

Brown, G. 'Big Stable Brands, and Ad Effects.' Admap, May 1991.

Carpenter, G. S., R. Glazer, and K. Nakamoto. 'Meaningful Brands from Meaningless Differentiation: The Dependence on Irrelevant Attributes.' Journal of Marketing Research 31, 3, (1994): 339-50.

Ehrenberg, A. S. C. 'B.E. or Not B.E.' In the eXperts reports on the 43rd ARF Annual Conference. New York: Advertising Research Foundation, 1997.

Feldwick, P. 'What Is Brand Equity Anyway, and How Do You Measure It?' Journal of the Market Research Society 38 (1996): 85-104.

Golder, P. N., and G. J. Tellis. 'Pioneer Advantage: Marketing Logic or Marketing Legend.' Journal of Marketing Research 30, 2 (1993): 158-70.

Guadagni, P. M., and J. D. C. Little. 'A Logit Model of Brand Choice Calibrated on Scanner Data.' Marketing Science 2 (1983): 203-38

Jones, J. P. When Ads Work. New York: Lexington Books, 1995.

Laforet, G., and J. Saunders. 'Managing Brand Portfolios.' Journal of Advertising Research 34, 5 (1994): 64-76.

Marshall, Sir Colin. Foreword to Market Segmentation by M. McDonald and I. Dunbar. London: Macmillan Press, 1995.

McDonald, C. How Advertising Works. Henley on Thames: NTC Books, 1992.

McDonald, M., and I. Dunbar. Market Segmentation. London: Macmillan Press, 1995.

Mitchell, A. 'Interview with the Chairman of Unilever.' Marketing, March 24, 1993.

Moran, W. T. 'Brand Presence and the Perceptual Frame.' Journal of Advertising Research 30, 5 (1990): 9-16.

Porter, M. E. Competitive Advantage: Creating and Sustaining Superior Performance. New York: Free Press, 1985.

Robertson, T. 'Get Your Retaliation in First.' Annual 'Master Class' of the Marketing Society. London, February 22, 1996.

Sood, S., (ed). Brand Equity and the Marketing Mix: Creating Customer Value. (Conference Summary). Cambridge: Marketing Science Institute, 1995.

Sutherland, M. Advertising and the Mind of the Consumer. St. Leonards, NSW: Allen & Unwin, 1993.

Weilbacher, W. M. Brand Marketing. Lincolnwood, IL. NTC Books, 1993.

Van Westendoorp, P. H. On the Road to Accountability. Amsterdam: Genootschaft voor Reclame, 1996.

Young, J. W. How to Become an Advertising Man. Chicago: Advertising Publication Inc., 1963.

JOAB Reports 1 to 8 (South Bank University Working Papers)

  1. Ehrenberg, A. S. C., and J. A. Scriven. 'Added Values or Propensities to Buy?' Admap, September 1997, 36-40.
  2. Barnard, N. R., and A. S. C. Ehrenberg. 'Advertising: Strongly Persuasive or Nudging?' Journal of Advertising Research 37, 1 (1997a): 21-31.
  3. Ehrenberg, A. S. C., J. A. Scriven, and N. R. Barnard. 'Advertising and Price.' Journal of Advertising Research 37, 3 (1997): 27-35.
  4. Ehrenberg, A. S. C., and N. R. Barnard. 'Advertising and Product Demand.' Admap, May 1996, 14-18,
  5. Ehrenberg, A. S. C., N. R. Barnard, and J. A. Scriven. 'Differentiation or Salience.' Journal of Advertising Research 37, 6 (1997a): 7-14. (This paper)
  6. Ehrenberg, A. S. C. 'How Do Consumers Come to Try a New Brand?' Admap, March 1997, 20-24.
  7. Barnard, N. R., and A. S. C. Ehrenberg. 'Attitudes to Brands.' (In preparation, 1997b).
  8. Ehrenberg, A. S. C., N. R. Barnard, and J. A. Scriven. 'Justifying Our Advertising Budgets.' 1997 (Submitted to the Journal of Marketing).


Differentiation or salience

Few marketing or advertising texts ask, let alone answer, why the market shares of big and small brands differ so much. Even fewer say that it explicitly turns on the number of people to whom the brand is salient and not on how well users of a brand regard it (because in practice that hardly differs from brand to brand).

Brand or product differentiation

The traditional view is that if brand differentiation does not exist functionally, it must be achieved by advertising. This view seems to be held even by writers otherwise close to the approach propounded here (eg, Moran, 1990; Sutherland, 1993; van Westendoorp, 1996; Weilbacher, 1993). The vogue also for increasing physical product differentiation came greatly to the fore in the '80s, eg, there were at least 13 versions of Crest toothpaste (Weilbacher 1993). But it is now declining slowly, partly because of its great stockholding and logistic costs. In the literature various different kinds of reasons or feelings for consumers to choose a brand are put forward and also in market research reports, eg, rational or emotional responses to attributes which are thought to be 'valuable' to the consumer (eg, see Agres, 1995; Jones, 1995; McDonald, 1992; Guadagni and Little, 1983; etc, for varied discussions).

But advertising is commonly also positioned on attributes that are 'distinguishing but irrelevant' (Carpenter et al, 1995), such as P&G promoting Folger's instant coffee for its 'flaked coffee crystals' which would give extra flavor. (This is a good story but 'irrelevant,' since the greater surface area of crystals would only affect the flavor of ground coffee, not that of soluble.) The claim is distinctive but not true, ie, distinguishing but irrelevant.

Differentiation between and within brands

In recent years cars (of a given type) and financial services have in particular been said to be unable to differentiate themselves. Yet typically there is much differentiation here within each company's products, fixed and variable interest rates for financial services, loans for five, ten and 25 years, with or without profits, differing commission rates. Even the companies' salesmen often fail to explain or understand these differences satisfactorily. But the different financial companies' range of offerings is much the same, and so there is little or no brand differentiation.

James Webb Young's famous 'products becoming more alike' was in his How to Become an Advertising Man. This was the substance of his lectures in Chicago and within JWT in the 1920s.


Laforet and Saunders (1994) have recently surveyed companies' varying branding practices, as did Weilbacher (1993). They typically note the current move toward umbrella brands. (This is in line with the arguments in JOAB, as long as 'other things are equal,' like one's historical investments in a brand name.)


Moran (1990) and Sutherland (1993) are exceptional in emphasizing the role of 'salience,' or something close to it. But they have not yet developed the extent to which it works almost exclusively through the sheer number of people to whom a brand is salient.

McDonald (1992) reports a more traditional 'salience' model, among others. Salience here is about a brand standing out more than for the others (ie, being 'more salient') and not about for how many people it is salient or all-round acceptable.

In an exceptionally explicit statement at the 1995 MSI Conference on Equity (Sood, 1995), Don Lehmann listed factors similar to those noted here as making up 'salience.' But again he did not yet distinguish that what in practice seems to matter is not just how highly (or 'how salient') people rate Brand A on these factors compared with Brand B but simply how many people rate each brand as being salient to them at all.

Publicizing the brand

As a thought experiment, let us suppose that Brand A is publicized and a near-identical B is not (with other things in effect being equal).

Then more people will (slowly) become aware of A; more may (and/or slowly will) think of buying it (if they want the category); and more can and slowly will actually be buying it then; be loyal to it, start saying that it 'Tastes Nice,' is 'Nourishing,' or 'Kind to the Hands'; say that they like it; choose it in a named product test; and so on. Other things being equal, the trend can hardly be the other way round (ie, B's performance improving), even with 'weak' advertising copy for A.

The advertised Brand A will therefore (slowly) increase its sales, will slowly get more distribution and shelf space, more promotions, media mentions, more sales and revenue and still more advertising, etc. There would also be many feedback loops and synergies: Brand A will become salient to more people, and more will actually be buying it (higher sales).

'Here I Am' publicity

An advertisement for Brand A might just say 'A' (as in a sports stadium). Often it does much more creatively, to attract more attention and leave more memory traces: 'Buy A,' 'A, the Outstanding Coffee,' 'A, Made from Selected Beans' or 'New' (although 'New' by itself will hardly leave a long-term memory trace for A).

There is probably little disagreement that such publicity would 'work' (ie, slowly lead to higher sales for A), other things being equal. But if Brand B manages to remain in good distribution and keeps shelf space (eg, by slotting allowances and a good sales force), B need not be overwhelmed.

However, if Brands A and B were about equally active in their advertising and other marketing-mix factors, neither would win over the other (although both might well do well). This is the normal position in competitive marketing, with advertising being mainly a part of defensive brand maintenance hence, 'Running hard to stand still.'

Advertising Research Foundation