Thanks in part to McKinsey’s coining of the term “Big Data” last summer, company data have never had more visibility than they do now. At the same time, they have never faced more scrutiny. Most of this scrutiny takes the form of concerns about privacy, the consumer-facing threat of data. However, the larger data discussion has not--so far--broached the aspect that makes them the scariest assets a company has.
Once the near-exclusive domain of direct marketers, data has become the secret weapon du jour in-store, online and just about everywhere else. The New York Times reported in February on how Target developed data models so sensitive that it could tell whether a customer had become pregnant before she told anyone. Brands such as Virgin America and Axe have used Klout to evaluate customers.
So what makes data so scary? Intangibility.
Take a look at other corporate assets. An experienced factory manager can tell whether a given factory runs efficiently, even if he lacks familiarity with the specific product manufactured there. For that matter, even an amateur could probably recognize that a factory with broken lights, pools of oil and water on the floor and visible fire hazards needs help. Similarly, high- and low-performing personnel tend to distinguish themselves clearly.
Not so with data. Data sit quietly in servers waiting for targeting or information requests. Moreover, when an organization uses data, doubt enshrouds the actual circumstances of use. For instance, if an email list underperforms, does that problem stem from shortcomings of the list, bad creative, uninspiring offers or something else?
Assets you can’t see working inevitably raise more questions than answers. Even a company’s branding leaves more clues as to its efficacy than data do. After all, a simple Google search can tell you what people think of your brand. Data remain more elusive.
What’s a marketer to do?
The answer lies in elbow grease. Inventory your data--what you have, where you have it and how you use it. Then start running some top level reports to see what you know about your customers. The answers may surprise you. For instance, if, say, your data show that 40% of your Canadian customers have French as a language preference, that means one thing. However, if you have no locations in Quebec, that means something else. Subject all of your information to your own personal sniff test.
Don’t be scared! Unlike a run-down factory, ghosts don’t hide in data.