Why Email Marketers Should Stay Clear Of Big Data

Why Email Marketers Should Stay Clear Of Big Data

"The so-called "Big Data" aficionados have done a masterful job in branding this meme. The promises keep coming. After the "behavioral targeting" meme crapped out, a new term was coined and voila! Big Data was born. "

Why Email Marketers Should Stay Clear Of Big Data 

 While those without entrepreneurial expertise rely primarily on
predictive techniques, expert entrepreneurs tend to invert these.
In particular, they use an effectual or non-predictive logic to
tackle uncertain market elements and co-construct novel
markets with committed stakeholders.
--From "Marketing Under Uncertainty" – Stuart Read et. al.


The so-called "Big Data" aficionados have done a masterful job in branding this meme. The promises keep coming. After the "behavioral targeting" meme crapped out, a new term was coined and voila! Big Data was born.

I do not want to bury the answer asked in the title so let me get to that and then explain. First and foremost, the answer to the title's question is simple; Big Data does not work...and worse, it cannot work. More data is not better...only better is better. Do not confuse one for the other.

Big Data and all the scientists in the world cannot solve the problem of "spurious correlation". These are variables that are absolutely statistically valid, but plain wrong.** As one aggregates more data, more noise occurs. But when noise is dressed up as signal, here lies the real problem because we marketers spend a lot of money on signal. Marketers need to eliminate noise or we'll go broke.

Some people swear by Astrology. And I have no illusions that this short essay will persuade anyone any more than me trying to persuade an astrologist that the stars do not dictate my life.

But eschewing Big Data is not theory for me. I nearly went broke when I chased statistically valid results that pointed the way to new variables. Acxiom and Experian helped me chase these by the way...plus many other service providers along my journey to discovering that it was all Bull S***.

Ever notice the only people touting Big Data are the folks who sell that crap? No kidding. They have no skin-in-the-game for actually using it themselves. They are not mining for gold, they are selling the pick axes and shovels to the folks trying, nay even desperately trying to tame uncertainty.

I have tremendous empathy for marketers who have not yet made peace with uncertainty. Predictive methodologies are akin to the old Homeric Siren songs. They are almost too difficult to resist. Predictive methodologies make a lot of sense on the surface...and they more often than not have the imprimatur of PhDs in physics or statistics.

We would do well to remember that University of Chicago's Myron Sholes won a Nobel Prize for his predictive algorithm valuing derivatives. He believed so much, he went into the private sector to apply it and folks rushed to give him over $7 billion to invest. Well, he ended up going up in flames SPECTACULARLY and lost billions. Beware of those touting credentials. It is a tip off that they are full of s***.


Now, email marketers should especially stay clear of the charlatans touting their voodoo science. Email is God's gift to marketing. It is (or should be) the least expensive marketing channel in your arsenal. You need not add unnecessary costs to your company. Do not be fooled by service provider Powerpoints or case studies. Instead, read Taleb's books, Fooled By Randomness, The Black Swan and Antifragile.

Lastly do not fear uncertainty. Embrace it. The future is inherently unpredictable. It is better for you to develop a way to cope with it rather than to run away from it.


** A simple example is as follows from William Burns:
1. Get data on all the fires in San Francisco for the last ten years.
2. Correlate the number of fire engines at each fire and the damages in dollars at each fire.


Note the significant relationship between number of fire engines and the amount of damage. Conclude that fire engines cause the damage.