E-commerce businesses send a variety of incentive emails (e.g. 10% discount, free shipping, etc.) to attract users to purchase items. Incentive emails can boost sales in the short term but it may lead to undesirable consequences in terms of margins. For example, it would be unwise to train users to expect only incentives which would harm the overall brand image. The best audience to send incentives are obviously to the right users. But we need to verify who these ”right users” are.
In this post, our team reveals how Retention Science chooses the right users for incentives. The post is two-fold:
- we show whether incentives are increasing sales
- we illustrate which users redeem incentives.
Do incentives boost sales?
In this section, we identify that incentive emails boosts sales but we need to verify if this is true.
One way to check the causality is that we randomly divide users into two groups. Incentive emails are sent to one group and non-incentive emails to another group, then we check the differences of sales between the two. The random selection reduces the difference of characteristics between the two user groups, and that leads to an unbiased comparison on average. Unfortunately, we cannot ask our clients to give us full access to start a campaign like this. Instead, we do an analysis based on our clients’ current data:
- purchase with incentive emails
- purchase with non-incentive emails.
Even though a random selection is not possible, it gives us an idea. Put simply, after everything is set as similar as possible we can check the differences.
To find similar users between these two groups, we need an arbitrary score…
Get full, free access to the rest of this paper. Preview: Click to Expand