The Sum of the Parts, Part 4: Segmenting Data

Welcome to the fourth in our series “The Sum of the Parts,” in which consultant Rob Galkoff discusses the elements critical to successful e-mail marketing. ?

It’s so much easier to just send out an e-mail to the whole database rather than spending the time to split it up into segments. But by sending a one-size-fits-all message to your entire house file, what do you learn about your customers, and are you maximizing the return on your mailing? Let’s face it, not all customers are the same, are we?

Our ultimate aim is to make as many customers buy from us as possible while making the highest achievable profit from the transaction. The elements of reaching that goal are the three C’s: customers, conversions, and contribution.

We need all three C’s to be as high as possible. But we all have different sets of customers who react differently to different offers and different communications. And we want to find a way of making costumers convert without giving away too much margin. So the reality is that we might need something that looks like this:

In the above table, we now have nine customer groups that will each provide a different conversion rate and, depending on these factors and the promotions you offer them, will give you a different contribution level.

Before we move on, let’s just clarify what promotion means. A promotion is not necessarily money off. It might be a free gift, a piece of exclusive information, or just a standard newsletter promoting you products. In our tests with clients, we often found that the 0-6 month house file performs better when there is no promotion—and this is not a one-off; it regularly happens.

From the above table (which is totally hypothetical, by the way), there are total sales of $363,180 and a contribution (before marketing costs) of $172,937, or a return of $0.77 per person on the database. The 3,255 customers who made a purchase represent an average response rate of 1.48% and an average order value of $111.58.

If we just did one campaign the end results might be very different:

By segmenting the data, we managed to increase the net contribution by 83%.

It’s important to be clear that you don’t just stumble across these segments. You have to test, test, test, and test some more. This is a skill, and it takes a huge amount of time to know when you’ve got it right. Is it the promotion, the seasonality, the product offering, the subject line?
There’s another huge consideration to be aware of: Those customers you recruit online aren’t necessarily as brand loyal as other customers. Some customers buy from you because you have the product they want at the right price at the right time and not because they have bought into your brand. Therefore, you might end up sending them e-mails and promotions that have a much lower conversion rate than those sent to other recent customers.

I can now hear you asking “How far do I go with segmenting my data?” The honest answer is that you can go on forever, but the return will diminish the deeper you go. The skill is to find the balance and maximize profits the best you can. Those with bigger databases either employ a consultancy to help them segment their data or have a computer program that can help them analyse the data. If you are using a top-tier e-mail service provider, then challenge it to help you segment your data; its staff are the experts and can help you.

Next time we’ll talk a little bit more about what you actually do with your data in the form of e-mails and testing.

Don’t forget to drop me an email ([email protected]) if you have any questions or comments.

Rob Galkoff is CEO of the Business Consultants. He was previously marketing director at multititle UK mailer Findel.