Good/Going/Gone: Simple Segmentation Scheme To Market To Your Customer's Next Move

Inertia and fear determine more marketing campaigns than most marketers are ready to admit. While companies talk the talk about customer-centricity and relevant communications, what comes out of their servers is that same old/same-old, one-size-fits-all, spray-and-pray messaging. We don’t know how to do individualized marketing, they’ll say, so let’s stick with what we do know.

When reading that each of their millions of customers should be getting individualized emails, the prospect of doing that actually frightens some of them and questions their competence as marketers. Their incorrect perception of the required effort (it’s much less than they think) feeds that fear and stifles their ability to move forward with better campaigns.

Individualized marketing doesn’t require individualized messages

Messaging is the culprit in that perception. Coming up with a million different messages is logistically impossible. (The companies that send out a million different emails do it by sending emails with different product offers, not different messages.) Even one hundred different messages is impossible at scale, and ten is very tough if you are a frequent mailer.



How about three? Three different messages are an achievable goal. Three we can do. If we segment the customer population into just three groups, we can come up with a message for each one. And that’s how the simple segmentation scheme Good/Going/Gone was born.

All customers fit into one of these three segments. Gone is for inactive customers, those that haven’t made a purchase for whatever is the time span of your active file. Going is the label for underperforming or fading customers. Healthy customers go into the Good bucket.

A single metric defines the segment

To keep it simple, we like to use only one metric to assign segment membership. Recency determines which customers are in the Gone bucket. We favor Risk Score (the probability that the customer will not make a purchase before becoming inactive) as the way to divide the active customers into Good and Going. Those with a Risk Score greater than 50% are in Going; under that number we put the customer in Good. Risk Score is perfect for this task. It’s an excellent measure of customer loyalty.  However if neither you nor your vendor know how to calculate it, recency (days since last purchase) is an acceptable substitute.

Using a simple segmentation scheme like this is an effective way to match your marketing message to your customer’s state of mind. While Risk Score defines the segment membership, you could expect that a Good customer will have purchased from multiple categories, has a shorter than average inter-order wait time, and a low recency.

Your Going customers are less healthy.  Their Risk Score is very probably worsening compared to the previous period. Recency will be higher than those who are Good. These customers are fading.

The Gone customers are of course already inactive. Typically there will be a wide spectrum of customer behaviors during the time they were active. Some might be one-time buyers.  Others could have had a healthy purchasing history in their prior life as an active customer.

Match marketing to segment characteristics

The basic idea behind a simple segmentation scheme is to have marketing actions appropriate to the segment. Good customers are buying regularly, but they are often stuck in a rut of buying from the same category selection. The best way to leverage their active purchasing is to present offers to buy of high margin cross-sell products (products they’ve never purchased from you before).  Those high margins create space to offer a discount. The discount doesn’t have to be large, but it signals to the customer your recognition of their status.

The Going customers need some inducement to resume their earlier buying patterns. It’s much easier to do this before they become inactive. First, calculate the customer/product purchase probabilities for each customer (or have your vendor do it). Then make an attractive offer on goods your purchase probabilities predict will be attractive to them. Whatever margin you sacrifice now is less expensive than the margin given up in reactivation efforts.

The Gone customers are absolutely a challenge, but one that must be accepted. Catalogers in particular need to keep dipping into their inactive files. You know who they are and how to reach them.  You know something about their purchasing habits. Any costs to rescue these customers are almost certainly less than the cost of acquiring a brand new customer.


The option to adopt a simple segmentation scheme like Good/Going/Gone means that sending more relevant emails is within reach of just about any company. If you have an appetite for even more relevance, you can individualize the offers to each customer in the segment, using segment membership to set the messaging. Combining segment messaging with individualized offers is a powerful, customer-centric program.
Next story loading loading..