Tag: marketing

The next book is another in my marketing theme How Brands Grow: What Marketers Don’t Know by Bryon Sharp. The core message of the book is that marketers have a lot of wrong ideas and empirical, scientific research can disabuse them of these notions, Sharp terms this new world “evidence-based marketing”.

I see a pattern in the form of the marketing books I have read: they both started with short form summaries of the main conclusions of the book. In Behavioural Marketing these were tweet sized chunks. In this book they are a set of laws: double jeopardy law, law of buyer moderation, natural monopoly law. The chapters discuss each of the laws in turn. The conclusion repeats these laws. It follows the scheme my PhD  supervisor taught me for academic presentation “tell them what you’re going to tell them, tell them, tell them what you told them”.

Sharp, who authors most of the chapters, works at the Ehrenberg-Bass Institute, perhaps it is following his own marketing advice but this seems to get mentioned in every chapter, several times.

The core of the laws is that much of the observed outcomes of brand performance really just come down to brand penetration (the share of the market that a brand has in terms of the number of customers and the frequency of purchase), and the essentially random behaviour of consumers. Loyalty to a brand is a bit of an apparition, brands maintain their position not because the same people buy from them repeatedly through loyalty but because consumers move from brand to brand somewhat at random but biased by the market share of a brand. So big brands gain and lose customers from sale to sale in greater numbers than small brands but as a proportion they do slightly better than small brands because an infrequent customer is more likely stumble on a big brand than a small one.

A second important feature is the distribution of buyer frequency. There are few frequent buyers of any brand, the shopper that buys Coca Cola 3 times a day, but there are many infrequent buyers – someone who buys Coca Cola a couple of times a year. Most brands have a distribution of consumers where the top 20% contribution at best half or so of the sales of the brand. This is rather less than the 80/20 Pareto law. It means that the long tail of infrequent buyers is important. Sharp is keen on citing the negative binomial distribution (NBD)-Dirichlet theory. It strikes me this isn’t a theory, it is a description of a distribution, the features of this distribution can be used to explain observed market behaviour.

Sharps conclusions on acquisition and retention are interesting, he argues that brands grow by acquisition rather than reduced retention. All brands lose some customers to their rivals at each purchase but acquisition and loss are generally proportional to the market share.

What surprised me was that even for brands with famous brand loyalty such as Apple, the degree to which “Apple buyers” buy other brands of computer is quite high i.e. 50% or so. And there aren’t really brand demographics, except in extreme cases such as rich people buying expensive items, and children watching children’s TV. There isn’t a “Ford buying demographic” and a “Renault demographic”. It isn’t even that there are generally customers committed to a particular price point, customers buying a premium brand are quite likely to buy a bargain brand as their next purchase. Niche brands turn out to be just a polite way of saying small.

Fundamental to all of this is that consumers care a whole lot less about brands, and their differentiating features, than you, the marketer, ever will. As consumers we just don’t care. And so the role of advertising is to shift the needle in terms of simple recognition. When it comes to my desire to buy a brown, sweet fizzy liquid to drink the special features of a particular brand are unimportant, the image a brand tries to project about itself is unimportant, the special demographic the brand tries to hit is unimportant. All that is important is that I remember the brand, the packaging and that it is on hand when I make my occasional decision to buy a brown, sweet fizzy liquid.

This seems to go against the idea of behavioural or targeted marketing, although perhaps the distinction is that the “bad” targeted marketing is based on demographics, age or gender, whilst the “good” targeted marketing is based on behaviour, abandoned baskets or opened emails.

I enjoyed this second foray into marketing, How Brands Grow is well-structured and relatively easy reading. I found the ideas it contained intriguing and the methodology is in tune with my background as a physical scientist.

Here I branch out into completely new territory with Behavioural Marketing by Dave Walters. I wanted to learn more about other areas of endeavour outside of data science, and I sit between the marketing services department and the internal marketing department at work. So here I am.

It turns out that actually I haven’t travelled so far, data is at the core of behavioural marketing. It is clear from how Walters writes, in 2015, that this concept has not penetrated to the heart of marketing which traditionally has been a more “creative” industry where crafting that perfect message has been the core of the job, rather than ensuring the perfect message is sent to the right potential customer.

Behavioural marketing is marketing that is responsive to the behaviour of the customer, so for example marketers might set up a set of emails that are triggered when you visit a website or put things in your shopping basket but then navigate away before buying (the so-called “abandoned cart”). The first such email might simply be a reminder that you had left items unbought, the second might ask if you have any questions and the third might be an offer of reduced price.

The book is divided into three parts which broadly cover an introduction to behavioural marketing, implementing behavioural marketing in an organisation and advanced topics (how to become immensely rich and successful as a behavioural marketer and what the very best behavioural marketing brands are doing).

During the preamble to these parts the “behavioural marketing manifesto” is provided, quite deliberately, in 50 tweet sized chunks for you to disseminate:

“Almost every sale begins with marketing. Get it right, then scale”

“Care more about what your audiences do than what they say”

These two quotes capture the core message of the book, the job is to measure what your prospects are doing (automatically) and then send them customised messages. The idea here is that process is largely hands off.

Behavioural marketing is based around customer journeys, Walters sees the big three here as acquisition, path to first purchase and path to repeat purchase. Much of the analysis of these journeys revolves around the behaviour of a customer on your website. So web analytics is important, when a potential customer first arrives on your site you will know little about them but you can tag them with an anonymous cookie so when you do find out who they are their behaviour prior to that point can be analysed.

Alongside marketing triggered by behaviour (visiting a webpage, abandoning a shopping cart) there is content triggered by segmentation. Segments of users may be something like “repeat purchasers”, “deep considered”, “raving fans” and “disinterested recipients”. They can be identified by scoring their behaviour, this scoring is achieved by attributing points to different actions and then thresholding based on point scores, or alternatively using actions to determine segments. I.e. someone making 10 purchases a year may be considered a repeat purchaser. Someone making 10 purchases a year, immediately after a product is launched and talking about it on social media may be considered a “raving fan”.

Another division is by channel, it is fair to say the key channel in all of this is email in terms of ability to act. Walters lists the channels as email, direct mail, sales (outbound telephone calling), call centre (inbound calling), social media, and mobile applications. Mobile applications are relevant here because they can do custom notifications and support messaging which is opted-in at time of installation. Walters sees the arrival of GPS on mobile devices as game changing, now marketing can be very location aware, with the importance of wearables growing in the future.

I quite liked the random facts scattered through the book, call centre calls average cost is $7.50, expect 20% attrition on a mailing list per annum, you need to do 30 hours per month per channel on social media to make it worth while, expect to see one marketer for every five to eight sales people in a mature organisation.

Unlike the scientific, technology and historical books I read this book has fairly direct instructions on how to execute your new behavioural marketing strategy in a real business. The closest in approach is Data science for business by Provost and Fawcett.

This book is pretty readable, although I found the style a bit grating in places. I felt it gave me a good overview of behavioural marketing. I know what it is now, and I’m pretty clear of what my key activities would be in making it work in an organisation. It probably means I’m going to pester my marketing colleagues now.