Marketing is out of touch with reality. How can we fix it?
July 27, 2017
Religie als merkstrategie: ik gelóóf dat het werkt!
September 4, 2017
Show all

Is this normal?! (Or: Byron Sharp for dummies)

Today, most brands measure most of their commercial efforts. Most of these measurements end up in dashboards or reports. Dashboards are usually straightforward and do what they must do: they simply show the current state of your KPI’s. All too often, reports are just like dashboards: they show some numbers, perhaps decorated with some ad hoc speculations. Here are some suggestions to let your reports do what they should: facilitate decision making.

A pile of bricks

Imagine you observe that 54% of the consumers that know about your brand state that you are ‘reliable’ and 36% state that you are ‘innovative’. Is this good? Should you worry? Does it make sense to simply compare this number with other brands? (Spoiler alert: no, it doesn’t.)

Or you might observe that in a year 10% of your customer base leaves you. Is that a lot? Should you do something about it? And you might observe that 18% of these darn disloyal customers go to competitor X. Is that a lot? Should you expect less?

As advertising icon Bob Hoffman stated: “data is just a pile of bricks until somebody builds a house”. Following Hoffman’s analogy, most reports show no attempt of stacking the bricks into some coherent structure.

Ehrenberg’s castle

But guess what, someone did build a house. A coherent structure you can use to your advantage. The late Andrew Ehrenberg and his fellow academics have, over the past sixty years, revealed and validated some very consistent patterns in consumer behaviour. They hold up in a broad range of categories, from beer to diapers and from credit cards to kerosene suppliers. Byron Sharp summarized this body of knowledge in the best seller How Brands Grow (and recently How Brands Grow 2 together with Jenny Romaniuk).

Understanding Ehrenberg’s behavioural patterns is vital: it means understanding whether your KPI’s are on par or not. It seems that slowly but surely, brands are starting to adopt this knowledge. For many, looking at brand KPI’s with these patterns in mind is uncomfortable, as most of us have learned to work with assumptions that are often the opposite of what Ehrenberg revealed.

Ehrenberg discovered, among others, that big brands not only have far more customers compared to small brands, but also that their customers are also slightly more loyal. This pattern is called the Double Jeopardy Law, in short ‘DJ’.

Besides a very well documented in the marketing sciences, DJ is a very useful pattern. It means that if you know the market shares and the average loyalty KPI of the category, you can predict the actual loyalty-KPI’s per brand quite accurately. That might sound like magic, but most of the time it works. Brand size is by far the most accurate predictor of brand KPI’s. Which makes sense, as people just have more experience with bigger brands and people happen to like what they are familiar with. Above, bigger brands are far more visible because of bigger marketing budgets.

The Double Jeopardy Law tells you what normal brand metrics look like. That means, for example, you can throw yourself a party if your brand turns out to have the most loyal customers, but if you are the market leader, well, that’s exactly what you would expect. It is, in other words ‘normal’.

On a more practical level, it means that as a marketer there’s not much you can do to improve the loyalty metrics of your brand, besides growing the entire customer base. This is the part that lots of marketers just refuse to believe. Retention is different from acquisition, right? But if you look at the data, it turns out that these concepts are very much related: if one goes up or down, so does the other.

The acquisition lever, however, is much stronger. Kantar Worldpanel demonstrated this pattern in their huge Brand Footprint study that included the performance of more than 10.000 brands. The figure below shows that growing brands grow mainly because of an increase in penetration and that declining brands shrink because of a decrease in penetration. Frequency, the loyalty measure in this graph, barely moves its needle.

Figure 1: Summary of Brand Footprint study (Source: Kantar Worldpanel)

Who you (really) compete with

So Double Jeopardy demonstrates that penetration is king and that you should probably pay more attention to light buyers. So, how about competition? Well, Ehrenberg also revealed consistent patterns in what brands consumers buy. It turns out that consumers simply buy other brands in line with their market share. This is called the Duplication of Purchase Law, or in short ‘DOP’. DOP tells you that the consumer is predictably polygamous and that all brands in the category are your competitor.

This is important, as most marketers distract themselves by the ‘positioning’ of themselves relative to that of their competitors. If brands are positioned more like us, then they must be a ‘closer’ competitor. If brands are positioned very differently, then you don’t have to worry about those. But DOP demonstrates that the bigger the brand, the more customers you will share with it.

The funny thing is that a DOP analysis is nothing more than a plain and simple cross-table, as you can see in Table 1. The table shows (anonymized) data from Dutch DIY chains. In the first column, you see the relative number of consumers that visited the DIY chains in the last 12 months. The other columns show you the relative number of visitors of one DIY chain that also visited other DIY chains. The pattern is straightforward: of the people who go to the smaller chains, most go to the bigger ones as well. But of all the people who go to the bigger chains only some go to the smaller ones.

This pattern is called the ‘Natural Monopoly Law’: bigger brands have a greater proportion of light users than smaller brands. And in every category, the light users outnumber heavy users by far. So, this pattern also suggests that paying attention to light buyers is a smart thing to do.

Table 1: example of a DOP table (source: The Commercial Works)

Usually, marketers don’t look at DOP tables, but at the much more sophisticated brand map: a plot with the relevant brands in the category and a bunch of brand attributes as depicted in Figure 1. The figure is a visual representation of a cross table with attribute perceptions of the various brands included.

Figure 1: example of a brand map (source: The Commercial Works)


The brand map in this case suggests, for example, that Brand E and F are close competitors. If we look at the duplication of purchase table, however, you will see that all brands share the most customers with brands A and B; indeed, the biggest brands in this category.

Be aware that a brand map shows some (usually small) variation in the perception of consumers; a DOP analysis shows what they buy. The latter shows a consistent relationship with your P&L, the former doesn’t.

When to pay attention

Besides giving you a more realistic picture of who you really compete with, DOP is also very useful as it tells you whether that 18% of your customers that end up at competitor X is something to worry about. It probably isn’t. It’s likely to be ‘normal’. That’s why it’s called a ‘law’.

The example above is not a carefully selected case to prove my point. With few exceptions, I observe the same patterns in all categories. Which makes this way of looking at your (purchase) data so extremely useful: it gives me a framework of norms.

Of course, you do come across exceptions. This is when it gets interesting. When you do, it doesn’t mean the marketing laws do not apply in your category—an often-heard excuse (“I told you our category was different!”). It means that something else is happening that makes your brand metric either go up or go down. Usually deviations from the norm have annoyingly obvious explanations. For example, one retailer has more shops than the other, or a brand bought more shelf space. Or perhaps a brand spent far more on advertising than its competitors. It is important to identify these reasons, as it makes sense of the dynamics in your category.

Selling more stuff

Once you understand that brands will submit to the laws of consumer behaviour, you’ll develop a better eye for nonsense. And in the world of marketing, there’s plenty of it. For example, every now and then you might read about a brand that claims how they have become a Love Brand and have an exceptionally high share of loyal customers.

Or you might read about the new marketing holy grail that will increase the loyalty of your customers. You will notice that these claims are mostly anecdotal and not accompanied with data. It’s very hard to break a pattern that has been consistently observed and documented for so long (Ehrenberg published his first paper when Elvis and The Everly Brothers were dominating the charts!)

Although some marketers get depressed by the idea that some fundamental laws limit the impact of their efforts by definition, there is a bright side to it: marketing life becomes more plain and simple. You will realize that many things you spend your precious time on just don’t matter. That’s a cost reduction right there. You will realize that it doesn’t make sense to update most of your metrics every quarter. That’s another cost reduction. More importantly, the metrics you keep will tell you if you are selling more stuff or not. You might have forgotten along the way, but that was the main reason to start measuring in the first place.

Start stacking your bricks

To go back to my original statement: reports shouldn’t look like dashboards. They should help you to decide to go left or right or perhaps refrain from action. Start stacking your bricks into coherent structures. DJ and DOP are some of the most fundamental constructions that will help giving your reports real business value. Just by giving a clear answer this one vital question: ‘is this normal?!’

Comments are closed.