Marketers Don’t Need to Be Data Scientists
Consider your average FDA food label. The nutrition information on these labels is comprehensive and accurate, but the label itself is not at all well-suited for the job of making intelligent food choices -especially when you are a busy parent racing down the super-market aisle with your screaming kids.
Marketers today are often working with data that is similarly hard to parse. There’s been a lot of hype has been about “Mad Men” becoming “Math Men.” But this is the opposite of how we should be thinking about it. We need to help marketers use data to do their job better, not ask them to change jobs. In fact, the more precise the targeting algorithm, the more a campaign requires brilliant creative. When marketers over-invest in algorithms without any increased investment in creative, the result is that the right people are being targeted at the right time, but with a pretty unimpressive message.
What marketers have today is the equivalent of a food label full of calorie counts and mystery ingredients – we have classifications for talking about who owns data (1st party, 2nd party, 3rd party), how data were collected (explicitly, implicitly), and what sort of insights data reveals (descriptive, predictive). But what we need is a more user-friendly solution that tells us in clear terms “this will make you fat” or “this is really healthy.” Otherwise, the expensive analytics platform you’ve invested in will just go unused, just as nutritional labels are often ignored. In a recent study, almost four of ten respondents said they didn’t use analytics tools that their company had adopted because they didn’t understand “how to use analytics to improve the business.”
If, for example, you’re running a social marketing campaign, your community manager probably doesn’t have time to schedule a meeting with the “analytics department” to see what’s trending and why. She is optimizing for viral traffic which moves very fast. She doesn’t need data, she needs answers: is this the right headline, what should I put at the top of the page right now, what article should I promote with my limited paid media budget? Those answers should be readily available in her natural process – not trapped in an analytics report.
A company that is putting data into the right context is UPS. The shipping giant crunches thousands of data points to optimize package delivery routes. The output they give to their 55,000 drivers and route supervisors – turn right, turn left – is far less complex, but much more useful, than what food manufacturers give the average consumer. Especially for a driver with a truck full of deliveries to make, winding his way through a city in high traffic conditions. These algorithms enable drivers to be productive while also minimizing emissions from their vehicles.
Remember that analyzing data isn’t the point. The point is better marketing. And marketing decisions are still made primarily by people, not machines (even if, increasingly, it’s people operating machines). It’s not that these people are innumerate and can’t understand math, but that they have a lot more on their plates than just analytics. The point of collecting massive troves of intelligence and having great data sciences is to help these people accomplish what they need to accomplish – just in a more effective and convenient way.
As much as Big Data holds great promise for marketing, churning out more and more analytics will not unto itself create better marketing and maybe worsen it. Let’s put math in the service of the job to be done.
Harvard Business Review
Posted from WordPress for Android