26th of January 2018
The value of Pareto’s bottom 80%
By Professor Byron Sharp and Dr Charles Graham
According to marketing’s infamous version of the Pareto law, the bottom 80% of a brand’s customers only deliver 20% of that brand’s sales, i.e. 80/20. If this were true then, ignoring the Heavy Buyer Fallacy for a moment, one might almost be forgiven for neglecting such consumers in order to focus marketing attention on the heaviest 20% (“loyalists”) who deliver the bulk of a brand’s sales.
In “How Brands Grow” (Oxford University Press, 2010) we reported on the Pareto law, based on a more detailed research report for Ehrenberg-Bass Institute sponsors (Sharp & Romaniuk 2007) titled “There is a Pareto Law, but not as you know it”. The report title gives it away, there is a lawlike pattern, but it isn’t 80/20, it’s more like 60/20. Which means that light buyers, the bottom 80%, deliver almost half of a brand’s current sales, so definitely not something to ignore – and later in this article we’ll explain why these light buyers are of even greater strategic importance than they look.
But old myths die hard, especially when there is a consulting industry that has made a living from them. Recently dunnhumby released a report that claims that the Pareto law is closer to the classic 80/20 than the 60/20 we reported.
dunnhumby analysed loyalty card data from one single retailer (presumably their owner Tesco). So their data captured only the purchases that consumers made in Tesco (a supermarket chain that has about 28% grocery share in the UK).
dunnhumby’s chart shows average Pareto shares of 62/20 for top-5 brands, and up to 73/20 for smaller brands. When calculated over a five year period it was 66/20 for larger brands and up to 79/20 for smaller brands. Indeed… one might wonder why dunnhumby would make a fuss about results that don’t look very different from those that we reported. But them doing so gives us the opportunity to point out errors, and publicise and reinforce the strategic implications of the Pareto law; we also provide some guidance on how to calculate Pareto Share.
We (and Schmittlein et al 1993) showed that Pareto Shares are more extreme for smaller brands and with longer analysis windows; dunnhumby, apparently unaware of the past academic literature, found the same. This does NOT mean that smaller brands enjoy higher loyalty. It also highlights the care analysts need to take when comparing different windows of time. In very short periods (or infrequently bought categories) the Pareto Share can look mild, as everyone who has bought the brand has pretty much only bought once hence the top 20% are responsible for not much more than 20% of sales. In longer time periods the Pareto Share is higher as heavier buyers make repeat purchases and more light buyers enter the data set by making a single purchase. Over very long time periods there is the problem that some customers effectively “drop out” (move house, exit the category, die) this makes these customers look very light, inflating the value of the remaining customers.
But there is something else that you really have to watch out for… the quality/completeness of the sample, i.e. does it capture all of a household’s buying? The dunnhumby analysis was limited to (loyalty program) data covering households’ purchases in a single retail chain (dunnhumby’s owner Tesco). This is probably why even their single year Pareto Share looks a bit high. In any period their Pareto Share calculation will be contaminated by people who have moved house and are mainly now shopping at another chain. These households will now look very light when in reality they are simply buying the brand at a different store. This inflates the Pareto, decreasing the apparent value of the bottom 80%. Even households who move to Tesco will be misrepresented. This bias gets worse over time. It’s a reminder that a flawed sample isn’t fixed by making it bigger.
Since “How Brands Grow” was published much more Pareto research has been done. Jenni Romaniuk calculated Pareto shares in developing countries from Nigeria to Malaysia, China to Brazil (see “How Brands Grow: Part 2” (Oxford Uni Press 2016)). While Charles Graham conducted two landmark analyses using Kantar data: continuously reporting households over multiple years, not purchases in just one supermarket chain, but in forty categories, for nearly 600 brands across the whole UK market between 1998/2004 and again 2009/2014 with closely replicating results.
We found that in the average brand’s five-year customer base:
- In one quarter the Pareto share is typically only 40/20, rising to over 50/20 for an analysis period of one year. Over a five year window it rises to just over 60/20, still far lower than the old 80/20 rule
- 80% of buyers are super-light – five purchases or fewer in five years.
Like dunnhumby we also found variances between brands: but we saw that small brands systematically attracted far more once-only buying than leaders – up to half of all customers for the smallest brands.
A typical 60/20 Pareto ratio of course means that heavy buyers are heavier and therefore important (dunnhumby’s headline was “new study proves heavy buyers are important”!). But the critical message is that most brand buyers are light/occasional customers yet they still account for 40% of long-run turnover. It will therefore take serious marketing attention to keep them coming back. Firstly, because there are four times as many of them, which demands clever use of broad-reach, low-cost media. And secondly, because the brand is not so mentally nor perhaps physically available for them – the challenge of improving this availability is something every marketer should be thinking about every day.
There is another fact to consider. If one looks at the future, for example by asking “what will my current bottom 80% of buyers be worth next year?” the answer is always “more than they were worth this year”. And this year’s top 20% of buyers will be worth less than the 60% of sales they generated this year (i.e. about 45%). Those who are statistically minded will recognise this regresssion-to-the mean. Given that most marketing plans are about the future, not the past, it’s best then to consider the 60/20 law as an over-statement, this year’s heaviest buyers will be worth less in the future (and our lightest will be worth more).
On top of this, if a brand wants to grow, then these light buyers (plus the many many non-buyers) will be the primary source of growth. Which makes the old “target your loyalists for efficiency” strategy look like a near complete dead-end.
Graham, Charles; Sharp, Byron; Trinh, Giang; & Dawes, John (2017) “The unbearable lightness of buying”, Report 73 for Corporate Sponsors, Ehrenberg-Bass Institute, University of South Australia Business School.
Schmittlein, D. C., Cooper, L. G. & Morrison, D. G. (1993) “Truth in Concentration in the Land of (80/20) Laws”, Marketing Science, 12:2, 167-83.
Sharp, Byron & Romaniuk, Jenni (2007) ‘There is a Pareto Law – but not as you know it.’ Report 42 for Corporate Sponsors, Ehrenberg-Bass Institute, University of South Australia Business School.