It’s hard enough to get lots of people to buy our brand, so why do marketing consultants insist on telling us we need to go further and get people to love our brand as well. Do they think we are an insecure bunch?

Loyalty is strongly correlated with tendency to agree to ‘brand love’ survey questions but…… most lovers still buy other brands, and almost all of a brand’s buyers don’t love it.

This turns out to be the real finding from a new piece of academic research (Rossiter & Bellman (2012) “Emotional Branding Pays Off”, Journal of Advertising Research). But that’s not how it’s presented, instead the Murdoch University press release gushes that new research shows people who love a brand purchase 38-60% more, “[emotional branding’s] effectiveness has been virtually unknown until now…the payoffs are substantial”.

Specifically the research showed that people who ticked “I regard it as ‘my’ brand” also reported that this brand made up more of their category buying; more than for buyers who didn’t regard it as their brand. But all this shows is what we’ve known for 40 or so years, that people who buy a brand more are more likely to agree (on a market research survey) to positive statements about that brand – that stated attitudes reflect past behaviour. Or more succinctly: attitudes reflect loyalty.

“I regard it as ‘my’ brand” was, by far, the most common of the ‘emotional attachments’ they measured – with about 20% of the buyer bases of particular brands of beer, instant coffee, gasoline, and laundry detergent ticking this box. It was also the most strongly associated with higher share of requirements (behavioural loyalty). I’m not surprised because it is most like a direct measure of (past) behaviour. If I mostly buy this brand of coffee then I’m much more likely to tick “I regard it as ‘my’ brand”. Whereas if I buy another brand(s) more then I’m hardly likely to tick that I regard this one as my special brand.

So reasonably we’d call this particular survey question (“I regard it as ‘my’ brand”) a measure of reported behavioural loyalty, so it would have to be highly associated with any other measure of reported behavioural loyalty. But Rossiter & Bellman in classic sleight- of-hand name this question ‘bonding’, which they say is a measure of an emotion (not a self-report of behaviour)! Naughty naughty.

Their measure of ‘brand love’ was more respectable, it asked if brand buyers agreed “I would say that I feel deep affection for this brand, like ‘love’, and would be really upset if I couldn’t have it”. Interestingly, hardly any of any brand’s buyers ticked this box.

• Just 4% of the average beer brand’s (male) buyers and average laundry detergent’s (female) buyers,

• 8% of the average instant coffee brand’s (female) buyers,

• and a mere 0.5% of the average gasoline brand’s (male) buyers.

Restricting the samples to the specific gender that represents the main weight of buyers reduced the proportion of light and lower involvement category buyers. This would have increased the incidence of brand lovers yet it was still about as low as is possible.

Rossiter & Bellman wrote that these results “reveal the difficulty of attaining strong attachment-like emotions”. Hmmm, these results also reveal how successful brands largely do without these emotional attachments.

With so very few of any brand’s buyers agreeing that they feel deep affection for the brand we would expect that these few rare ‘lovers’ would be quite different from the average buyer. We’d expect that they would be the heaviest, most loyal in the buyer base. And these lovers were more loyal, though it was far from absolute (100% share of category buying). In fact, lovers only reported buying the brand about half the time (50% share of their category buying). So while behavioural loyalty is associated with propensity to agree to ‘brand love’ questions…… most lovers still buy other brands.

Rossiter & Bellman interpret their results differently. Their article title says emotional branding pays off, though the article does nothing to investigate marketing practice. It’s wrong to use cross-sectional survey data to imply a massive payoff from building brand love. Such a mis-interpretation ignores research going back decades that shows, over and over, that past usage largely determines propensity to react to attitudinal type survey questions (see Bird & Ehrenberg 1970; Barwise & Ehrenberg 1987, Riquier & Sharp 1997; Romaniuk & Sharp 2000). They even conclude that a brand’s emotionally attached customers must be its most profitable. The truth is that the data shows none of these things, their well written article turns out to be highly misleading. Yet I fear that this will not stop unscholarly academics mistakenly citing the article, and many believers of this old theory citing this article as ‘evidence’ to support a focus on building love. Beware of such nonsense.