The Fallacy of Analytical Marketing

While researching for a post on consumer and business spending in times of a recession, Google search made me stumble upon a sales pitch for an Analytical Marketing company:

“In today’s hyper-changing market scenario companies need to invest in data-based analytical marketing without which their ability to respond to quick changes in customer behaviour will be limited!”

Excuse me? A lot of FUD – fear, uncertainty and doubt being spread.

Supposedly, Analytical Marketing will be the biggest competitive advantage enterprises will have in the next decade and that only those will be able to improve profit and growth who can organize and leverage customer information. The marketing pitch claims that ‘tomorrow’s business decisions are driven by models that encapsulate yesterday’s reality.’ Hmm, interesting. Apparently a fundamental shift in consumer behavior is taking place, which might or might not be true, but how would a model that ‘encapsulates YESTERDAY’s reality’ help tomorrow’s business decisions? If a shift is taking place then would yesterday’s model not be wrong for today and certainly for tomorrow? Yes, consumers are spending less money and some understanding who is still spending on what could help a business to market to better fitting classes of customers. I propose that the problem is much rather related to customer to business communication and not vice versa.

The sales pitch foolishly tries to present economic expertise but misquoted Keynes’ ‘Paradox of Thrift’ as an example for consumer irrationality and to prove that yesterdays analytic methods are out of date. Oops! Keynes stated that principle in 1936, therefore irrational customer behavior would not really be new, right? While Keynes’ ideas are opposed by economists such as Milton Friedmann, the ‘Paradox of Thrift’ is not about irrationality of consumers, but about the question whether less consumer spending in a recession will actually increase savings and if and how politics should deal with the consequences for the economy. The paradoxical part is not the relationship of savings to business investment but the fact that the problem of changes in consumer spending is supposed to be solved by the analytical mass marketing means that created it! Consumers are saying that they will be more careful when making buying decisions and that they won’t max out their credit cards anymore to buy stuff they don’t really need. Which means that they are buying less of the crap that has been marketed to them with the help of analytical tools.

Let me reiterate here that the problem of data mining and analytical marketing is actually threefold:

  1. First, you need some kind of model – no amount of data mining can give you that. It is a theory that you have to come up with. Some propose that the best way to define those is to employ mathematicians. What? Mathematicians are the most unimaginative people on this planet.
  2. Second, you need to collect data. But which data? In what intervals. From where? What do the data mean? Will a trend follow through? Do they correlate in any way? Mapping collected data to the model is pure guessing.
  3. Third, most of the data are of poor quality and there actually is NO WAY of knowing how good your data are. The more you try to ‘fix the data in the mix’ meaning in the data warehouse with sampling, averaging and merging the more meaningless they get.

And to top it all off, there is the claim that being able to work with analytical data and the fitting visualization tools will be a boost for your career, because you will dazzle some board member with flashy charts. That does not portray board members as very intelligent, right? It is however in line with ‘The Economists’ observation that CEOs of many large enterprises are hopelessly inept and have no idea what is going on. All those charts may be flashy, but they will still be wrong. Even if the data are correct the model that projects them into the future is theory. Data and charts won’t improve the bottom line and growth or improve the economy. How should they? What knowledge would you gain that makes people part with money they don’t have? Supposedly, the knowledge could help your customers to be more thrifty and save money. Damn, according to Keynes that would not help the economy either! The only thing that will improve your business is a motivating leader with an intuitive mind and a customer focus.

Ebay has 5 petabytes of data stored. I have not seen any sensible marketing out of that. Walmart also has petabytes but I still wouldn’t shop there. The others all know that Walmart is cheap! Let me ask you: Would it not be better to spend the money on IT investments that improve product quality, customer communication and service rather than on analytical mass marketing?

A serious change in customer behavior IS NOT a goldmine, it is a message and businesses should better listen!

2 Comments on “The Fallacy of Analytical Marketing

  1. why be so rude about mathematicians? Multidimensional curved spaces and imaginary numbers requires imagination, surely!

    I agree about lack of theory – but the walmart finding about Beer and Nappies/Diapers being unexpectedly frequent on the same ticket demands imaginative reply – which further data exploration – time of day, gender of purchaser helps confirm…
    There’s a role for Abductive reasoning – norms, exceptions and explanations which can help build up understanding and predictions. If you have a complex understanding of your data set and you can see changing salience of aspects of that understanding you can make informed predictions based on data and theory.


    • Stephen, thanks for the comment! Appreciated. I apologize for the my quip about mathematicians. I agree with you that they tend to be VERY creative when it comes to finding ways … i.e. Renormalization for quantum electrodynamics (QED) to make infinite integrals in perturbation theory work. The Walmart beer and diapers story is no more than anecdotal evidence and even here I would doubt the long term benefit of that knowledge considering the cost of acquiring it.

      You are right that it is possible to make predictions based on the law of large numbers but these have no causal relationship as to why a certain individuals may or may not behave in a predicted way. You cannot gain an understanding of complex data gathered from the behavior of individual agents … because the relationships are complex and not just complicated (Evolving Complex Systems. Anderson, Arrow, Pines – 1988).

      When the environment of a complex adaptive system evolves you can take the ‘changing salience of aspects’ and shove it. it will no longer make sensible predictions. Why do you think we have the crisis? The Black-Scholes model for future equity value is partially responsible. Despite the fact that it requires eight simplifying assumptions people (mathematicians mostly) blindly followed it. It cannot handle extreme moves, yielding tail risk, which can be hedged with out-of-the-money options and assumes instant trading at zero cost. Gap and volatility risks are substantial when one uses a stationary, continued trading process that does not exist in reality.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: