Send the right offer to the right customer at the right email address – that’s a key goal of a robust email marketing program.
How do you accomplish this? The answer shouldn’t be surprising: Know your customers.
But that can mean a lot of different things. There’s a lot of room on the spectrum between knowing everything there is to know about each customer (and tailoring communication to them individually) and knowing nothing about any customer (and blasting out mass communication blindly). Traditional segmentation techniques are clear: use data to drive a targeted communication strategy for particular customer groups.
Here at FreshAddress, we’ve been doing a lot of thinking about optimizing email communication, and we’ve found it helpful to think of data as measured across a few different axes:
- Quantity. More data means more columns in your database, less data means fewer columns. If you know that your customer is male, over 65, and lives in New York, you have more data than if you know only that he is male.
- Depth. Data depth is an attribute of a data point or set of data points. The deeper data you have, the fewer assumptions you need to make about a customer. For example, data showing that a customer has a strong history of purchasing your company’s blue widgets is “deeper” than data showing that your customer is a male and that males are more likely to purchase blue widgets than red widgets. (Those of you in the lead scoring space may see an analogy here to explicit and implicit scoring models.)
- Age. A data point’s age is inversely related to the assumptions you are forced to make if you want to use it in isolation. If you know a customer purchased a Hyundai Sonata yesterday, it is less of a leap to assume they still own it than if the customer purchased their Sonata three years ago. But notice the caveat “in isolation”; if you’re considering a set of many data points, then older data can reinforce more recent data if they suggest a trend (e.g., this customer is a loyal and longstanding Sonata purchaser who recently purchased another Sonata).
You can see how being explicit about your data in these terms can give you some perspective as you strive to deliver relevant messaging. For example, if you have only a few, old, shallow data points, you should be less confident about using this data to drive messaging strategy than if you have many, new (and supportive old), deep data points – all else being equal.
So what does all this mean? Before you jump headfirst into a segmentation approach, pause to make an honest assessment of your data, including how it stacks up on the metrics above. Doing so will either give you the confidence to continue down that road or suggest where you need to shore up or benchmark before you can proceed reliably.