How to Prepare a Successful Integrated Digital Marketing Program, Part 2
Less is sometimes more. In theory, a good analytics practitioner can take a huge, undefined mass of consumer data and find the salient points. However, the "data by the ton" approach has several immediate pitfalls. For one, processing large amounts of data takes valuable time. Secondly, sometimes data remains unavailable due to technology, security or, more often, process barriers. Lastly, the larger a batch of data is, the greater the chance that inaccurate data is hiding in it.
Focus instead on relatively small, relevant data sets. Part of the danger of large data sets lies in the presence of irrelevant data. For retailers, for example, purchase data generally represents the most valuable data because it represents what people actually bought. Similarly, publishers use browsing behavior to determine what content may interest visitors.
Unfortunately, not all marketers have purchase or browsing behavior data at their disposal. What then? Simply put, you need to ask yourself, "which data can best predict behavior or interest?"
If a clothing retailer didn't have purchase data, for example, it could use gender to determine which items to feature on its website when a cookied user visits. On the other hand, a hardware retailer might not have much use for gender data, but would benefit from knowing whether the visitor lived in a house or an apartment.
Always test results. While data conveys a sense of mathematical certainty, it's only that — a sense. Test various data sets against one another and against a control. From personal experience, we can think of several smart marketers who learned that all the data they had couldn't do any better than the static offers in their emails.
Just because you can collect a piece of information doesn’t mean you should. As discussed, not all data is actionable; that which isn’t is nothing more than overhead. Data can be labeled “smart” if it enables you to build strategies around customers and segments, and if it gives you insight into what a person might — or might not — be interested in at a particular point in time. It’s data that makes engaging with your company more relevant and satisfying for your customers.