WHEN ANALYSTS GO APE

Photo of Cliff Courtney
Date: 2019-04-23 13:35:18
By: Cliff Courtney
Date: 2019-04-23 13:35:18

There’s a bit of madness in retail marketers’ chase for numerical fortification: A desire to build arguments as impenetrable as Kevlar and as logical as an anal-retentive Vulcan. This breakneck chase for certainty presumes that with an infinite data set, sharper algorithms, and predicative analytics, we can totally crack consumer behavior and drive retail KPIs with zero marketing slippage. Ha!

Here’s something else that’s true… given an infinite amount of time, a monkey will actually type Hamlet.

So here is what monkeys and data scientists have in common: They’re both theoretically right, but getting it all the way wrong.

The fact is, beyond available data, beyond the 20 petabytes of data Google crunches each day, or Amazon’s 1B gig algorithm model, the way consumers not just think, but feel about an experience, whether on their phone or on the shelf in front of them, can defy measurement.

Consider that sales data might easily tell you that 26% of sauce buyers choose Ragu, 21% choose Prego, 15% choose Newman’s Own, 12% choose Bertolli, and 26% choose other. But shopping lists don’t name actual brands, they just say sauce and 74% of consumers make their decisions within 3 feet of the shelf. Doesn’t this mean we can modify behavior at the moment of truth by making a connection that defies data? And ultimately, aren’t we always forgetting that any data set solely reflects what was, not what could be?

Respectfully, it is true that with enough data and planning, retailers can focus on the wide side of the barn, as it were, knowing that the eccentricities of some consumers will indeed mean slippage in the form of a few wildlings whose behavior flies in the face of the even predicative analytics. The masses, though, will follow the predicative analytics model and in theory the retailer wins. But there’s a chasm between winning incrementally and winning big, and the later comes not just when the data says “this will work”, but when the surprise and delight of the offering galvanizes the consumer in a way that perhaps we didn’t see coming, yet we were able to goose with intention. From breakthrough people (TV stars like Joy of Painting’s Bob Ross or Donald Trump) to breakthrough retail (silly bands or slime), phenomenons happen outside the realm of the calculable. In the case of slime alone, sleepy Elmer’s glue was purchased by Newell brands in 2015 for $600M after decades of flat sales. A single, slimy quarter later, they were up 9% on the kind of trend spike that would have caught Nostradamus asleep at the wheel #ohthatslime.

Amazon’s legendary algorithms know that someone who loves the movie Bohemian Rhapsody would likely respond if they were served an interstitial for a Muse greatest hits download. But are algorithms factoring in the decay rate of burnout on a movie typically viewed more than once and today is the number one choice of airline travelers watching in-flight movies? (Monkeys now need in-flight ethnographies).

Then think about true meaning of retail surprise and delight, which remains the quintessentially underused retail weapon, and often where what is and what could be pass like ships in the night. Conventional analytics would suggest that inflatable floats belong in the pool department. Yet imagine Joe and Jane shopping at a mega box like Sam’s Club: They’re picking up some basic household commodities, when suddenly they turn a corner and come face to a ginormous inflatable pink flamingo large enough to hold ten people. It’s huge, it’s shocking, and the fact that it almost makes no sense where it’s placed in the store makes it even more compelling. Analysts don’t respond well with things that make no sense. But consumers do. They smile. They’re engaged. The entire Sam’s Club Brand is elevated beyond commodity and membership, and into the realm of fun “destination brand.” And all engineered through more of a serendipity coefficient, and less through retail dogma.

The Chinese proverb says, “when men speak of the future, the Gods laugh.” I get the feeling the biggest yucks are reserved for analysts who don’t just speak, but lean on data the way a drunk leans on a light post… not for illumination, but for support. Don’t get me wrong, I believe in the power of data and predictive analytics alike. We’ve built retail empires with both. Moreover, I feel a whole lot better when my big idea bets are hedged with data. But if you’re not also considering X factors, the improbably and profitable unknowns, then you’re just monkeying around.

Share

x