How brands grow: work in progress

The baying and mooing of marketing sacred cows being slaughtered can be heard emanating from the pages of Professor Byron Sharp’s book How brands grow: what marketers don’t know. Deeply held marketing ideas and widespread practices are denounced and derided – marketers likened to medieval doctors, with their principles and practices compared to those of blood letting.

The book is an accessible compendium of the work of the late Professor Ehrenberg and his colleagues. It has been reviewed in Market Leader (Quarter 2, 2011) but, given its iconoclastic nature and that it is based on more than 50 years of scholarship, it requires much attention and some challenge by the marketing community.

Loyalty programmes don’t work

I am a partial convert to Ehrenberg’s ideas and have been since the early 1970s. However, in Sharp’s book there are areas where in my view his ‘evidence’ fails to convince and there are key issues that he doesn’t address at all.

A central Ehrenbergian assertion, with only minor caveats, is that a brand’s share is determined by the number of users is has. To grow, it must get more users; it can’t grow by inducing existing users to use more. Ehrenberg/ Sharp have studied hundreds if not thousands of brands in different markets and geographies and nowhere do they find brands with greater loyalty than their competitors. Their deduction is that ‘more from existing users’, an oft-stated marketing strategy, never works.

Although there are some major unanswered questions about this, I have always felt comfortable with putting penetration as the header objective if growth is to be achieved. It seemed to me to be a corollary of the Parfitt and Collins (1967) trial and repeat-purchase model of how new brands become established in the first place.

I also have no problem with his evidence that light brand users are of great importance for both current and future sales and therefore relationship marketing and CRM programmes that ignore them are a bad idea.

And I agree with his evidence and argument that loyalty programmes have weak effects on brand share and are unprofitable. I have always thought the provenance of these programmes suspect. They have usually originated in the Harvard Business Review or the like before being picked up and sold around the world by the large management consulting and IT firms but their evidential foundation has always appeared dubious to me.

It is therefore particularly gratifying to read the demolition job done on the article (Reichheld and Sasser in HBR 1992) that claimed that companies can boost their profits by almost 100% by retaining just 5% more customers. Sharp shows that although this may be true, arithmetically it is nonsense and can never happen.

The chapter on price promotions is particularly worth reading. He also demonstrates how advertising can be working even if sales don’t increase. And I have always found the Ehrenberg assertion that the main role of advertising is to reinforce memory structures compelling. And there is much more good in the book – it is essential reading.

Two fingers to Orthodoxy

Sharp’s audacious objective is to overturn the precepts that underlie today’s marketing orthodoxy, and the reader is likely to have the disturbing feeling that much of what they believe is being proved untrue. This is not the case. The impression is created through a combination of destroying straw men, exaggeration and going beyond the evidence. There is also a major flaw in the whole project that is not acknowledged.

The first object of attack is the ‘ideology’ that ‘differentiated brands should sell to different groups of people’. He presents academic evidence to show that brand usage is not related to customer personality type and presents many tables for a large number of brands showing that brand users don’t differ much by demographics, media habits or general values, by which he means TGI type statements – for example, ‘I can’t bear untidiness’ or ‘children should express themselves freely’.

This may disturb some marketers but this is a straw man. The concept of target group – ‘different groups of people’ – means people who want or who are attracted to slightly different functional or symbolic things, not people who are different in the ways being described (although there may be correlations with demographics etc).

The professor may think he is addressing this point when he attacks Philip Kotler for saying that different flavours of soft drinks, such as Coke, Lilt, Fanta, cater to different people. A position he demolishes by showing that large proportions of Lilt and Fanta buyers buy Coke.

This is another ‘straw man’ created through a combination of Kotler’s exaggeration and an unwillingness to embrace the spirit behind Kotler’s idea. Sure, the buyers of Coke, Lilt and Fanta are not completely different people but I am sure there are people who have a hankering (a need-state) for Lilt more often than for other brands and/or who would buy it more often if it were available or available in the right size or simply people who are in the mood for one or other of these drinks at different times. If this is the case, is it so silly to conceptualise these people as Lilt drinkers? I think not.

Major heresy

Fire is then turned on the most sacred of sacred marketing cows – the importance of brands. The attack is made on seven or eight fronts. The first attack is on the named versus blind product test, a cornerstone in the argument for branding importance which has received a significant boost in recent years with reports that brands stimulate brain activity in the hippocampus and other areas concerned with cultural knowledge. This is dismissed as being a symptom of familiarity rather than anything else but he presents no substantial evidence.

Ehrenberg/Sharp acknowledge that there is such a thing as brand loyalty but they assert that this is a phenomenon of the product sector, a habit adopted by humans to make life simpler. This is undoubtedly true. They quote experiments that show that under test conditions, consumers show loyalty to trivial differences – for example, loaves of bread labelled as L, M, P and H.

However, to establish the case that brands in the normal sense of the word don’t contribute to loyalty, it would be necessary to show that under similar conditions – for example, perfect distribution – the levels of loyalty were no higher than to the trivial alphabetic brands. This has not been done.

Furthermore, the most important point about brands – that they command higher prices – is not addressed.

Qualitative research among brand users who show emotional attachment to brands is dismissed as myth originating from interrogating small samples of biased customers who are paid for interviews and reported by biased researchers.

Despite this scornful dismissal, Sharp acknowledges that brand fanatics do exist but asserts without proof that all brands have them in proportion to market share and that even for iconic examples like Apple and Harley Davidson they have little impact on loyalty. The customer advocacy effect on recruitment is said to be negligible due to the small number of fanatics but again no evidence is presented.

Glaring weakness

The glaring weakness of the Ehrenberg/Sharp project is that it is a static analysis. It compares competing brands with different market shares at fixed moments in time. They have never studied how brands become established in the first place and the mechanisms by which the largest acquired the greatest number of users – so the book is not about how brands grow.

Studying how growth has actually come about gives an insight into growth that a static analysis can’t give.

At one point in the book, Kotler is taken to task for saying that brands should target specific segments but, according to Professor Sharp, segments don’t exist.

Let’s take a dynamic case – how Dove deodorant grew. Dove deodorant’s share plateaued a year after launch. A new scent variant was introduced but it cannibalised the original and the total hardly increased.

In year four, a variant for sensitive skin was launched followed two years later by one ‘with added moisturiser’ and then one called Invisible that didn’t leave marks on clothes. Each benefit was advertised and market share grew and stayed up with each introduction and there was negligible cannibalisation. By year ten, market share had doubled.

An Ehrengbergian snapshot, say over a six-month period, is highly likely to reveal that buyers of the variants have bought the original too and their analysis would conclude that there are no segments. I would submit that the brand grew because there were consumers who had a predisposition towards a product for sensitive skin, another group who wanted moisturiser and a third interested in Invisible. The new scent found no significant segment and, therefore, failed to grow the brand.

Throughout his long career, Andrew Ehrenberg strove to make himself clear, believing that this was the way to win acceptance for his ideas. But this was the wrong diagnosis. Professor Sharp’s book is admirably clear but despite what could be described as a tour de force I am still only a partial convert. The problem isn’t lack of clarity, the problem is that at its heart the analysis is one dimensional and therefore incomplete.

David Cowan is founder of Forensics.

[email protected]

‘How brands grow: what marketers don’t know.’ Professor Byron Sharp, OUP (2010)

Their deduction is that ‘more from existing users’, an oftstated strategy, never works

Fire is turned on the most sacred of sacred marketing cows – the importance of brands