I remember sitting in a business school marketing class. The professor started talking about data analysis. Statistics. Correlations. Math. Hey wait – this class was supposed to be about marketing. You know, understanding customers. Touchy-feely insights. Creativity. Thinking of clever ways to sell stuff. There wasn’t supposed to be any math!
Just then I got a dark, unnerving, queasy feeling in my stomach. I felt my body getting warm. It crept over me like the shadow of death – “marketing” wasn’t going to be the safe haven my liberal arts education and I had hoped for. What if the precious “business intuition” I’d been cultivating (and paying outrageous tuition to enhance) was ultimately an illusion? What if the real future of marketing would one day belong to math, leaving me nothing but insecure, commoditized, 2nd rate “soft skills” in the desperate hope suckers somewhere could be convinced those “insights” were still worth paying for? I was no math wizard. I was no statistics genius. I wasn’t a computer scientist. I’m surprised I didn’t throw up right then and there.
I bring this up because, as it turns out, I wasn’t alone. A recent Harvard Business Review blog shared a survey of nearly 800 marketers in Fortune 1,000 companies.[i] They found that marketers generally loathe anything quantitative. Most over-rely on their “gut” intuitions and struggle with statistics. Meanwhile those who embrace data often don’t know how to use it and are dangerously distracted by it. For example marketers, on average, depend on data for just 11% of all customer-related decisions. When asked to articulate how they made decisions, more than half drew only from personal experience or their subjective intuition about customers. In other words, when asked how big marketing decisions were made, most said the equivalent of “it just felt right,” and they were content (even proud) in doing so. When you pair this blasé attitude towards math (and glorification of intuition) with the reality that over 80% of new product and service launches fail… there might be a problem.
Some marketers need to exalt their intuitions because, apparently, it’s all they’ve got. When tested for statistical aptitude with 5 questions ranging from basic to intermediate, nearly half (44%) got 4 or all of the answers wrong. Only 6% got all 5 questions right. Meanwhile the marketers that did use data (11% in this study) frequently couldn’t draw actionable insights from their analyses and tended to be severe underperformers. “Every time they see a blip on the dashboard, they adjust – and end up changing direction so often that they lose sight of the end goals. In management positions, these people can wreak havoc by creating endless fire drills and preventing anyone from sticking with projects long enough to achieve the best results.”[ii] Misusing a tool can be just as dangerous as not using it all.
Meanwhile some marketing departments are beginning to hire data scientists instead of more traditional marketing MBAs. It’s still early and there remain far more intuition-types than “quants” in most marketing organizations. Still, analyzing Internet behavior, customer buying patterns, shifting market trends and solving complex market problems through predictive analytics isn’t something esoteric in the backroom anymore. It’s gone mainstream and is gaining speed.
Data scientists have long used used math, algorithms and computing power to unravel mysteries of the universe. Nuclear blasts are simulated using supercomputers (and a whole lotta’ math) to predict where radiation clouds will float. Cities are simulating floodwaters to plan evacuation routes. Proctor & Gamble used empirical simulations to model everything from product design, process design and supply chain management to the aerodynamics of Pringles potato chips. “We build and test the first prototypes…virtual ones.”[iii] Data analysis and quantitative models have gone into Whirlpool appliances, golf clubs and even Speedo swimsuits.
Here’s the kicker. Now data scientists are even – with accessible data, open source software and cheap computing horsepower – simulating entire business models and marketplaces. You heard that right… they’re not just simulating products anymore, now they’re simulating entire business models, markets and how those systems interact. They’re modeling dynamic, multivariate business ecosystems and are predicting important market behavior with unheard-of accuracy. Like a war game simulator or computer-modeled car crash, business simulation is now letting some innovators meaningfully test-drive new business ideas quickly and cheaply – without writing a single marketing plan. It wasn’t intuition that did this. Regrettably (for today’s aspiring Nick Drapers) it was math… and this is just the beginning.
Some of you are thinking “wait a minute, intuition is still valuable,” or “there are some things math can never replace.” You may also think I’m being too hard on marketing overall. Yes, you’re probably right. I’m not advocating blind surrender to all things quantitative or the total banishment of intuition. Yet whether you’re a math-lete or a marketer who majored in 11th century Nepalese poetry, it’s time to ask yourself some hard questions. What skills will you need, or, how can you otherwise align yourself to better catch the quantitative wave instead of being swamped by it?
Put differently, how can quantitative tools and data science become more accessible, easier to use and more intelligently targeted so marketers can better improve their decisions through math (yes, math)? It isn’t just a matter of “more data” or “more dashboards.” There’s plenty of that stuff already. Rather, how can marketers with less empirical sophistication begin to leverage number crunching in ways that were previously only available to those with greater quantitative skill or resources? Given their connection to customers and markets, if the math revolution is coming, perhaps marketers (yes, marketers) can even lead the way.
[i] Spenner & Bird, Marketers Flunk the Big Data Test, Harvard Business Review Blog Network (August 16, 2012). http://blogs.hbr.org/cs/2012/08/marketers_flunk_the_big_data_test.html
[ii] Id.
[iii] Thomas Lange, Director of Modeling & Simulation, Proctor & Gamble. See Lyndon, Modeling & simulation speed process development, Automation.com http://shar.es/7HQxh via @sharethis