Not rushing the results can create a better outcome

A couple of weeks ago I cut and harvested a couple hundred chickpea plants.

After a few of hours work, I had about half of them out of their pods and harvested but it was slow work. I boxed up the rest of them, set them aside, and came back to them last weekend.

In that time, the pods dried more and the work went twice as fast.

By letting them sit for a time, I was able to work more effectively. The pods did some of the work for me.

Sending marketing messages to your customers to convince them to buy again can run into a similar problem.

If they've just ordered, messaging about buying again might fall flat. You'll have a reduced response rate from them, which could hide your response rate from other customers.

But if you waited a bit of time, they might be primed and ready for another purchase.

This is why segmenting is so valuable. If you withhold the more salesy and pitch emails from recent buyers, you'll see a better response rate from the non-recent buyers. Instead, send the recent buyers more customer service or softer sales messages.

(If you're running an event, by all means send that to everyone though. Events break many of the rules, those rebels)

Sending to segments is especially important if you're testing your messages (e.g. A/B tests). Otherwise you have to be very careful that you split up your recent buyers so they are equally represented in both groups (control, experiment). Too many recent buyers in one segment could skew the results badly.

In practical terms, I'd recommend holding back on customers with a Recency of 5 based on the RFM analysis Repeat Customer Insights. That would be the best way to have the analysis self-adjust and you don't have to worry about specific date thresholds (e.g. 30 days post-purchase VS 60 VS 25 post-delivery).

Eric Davis