Ease
While we’ve spent the last few weeks of this newsletter getting into the concept of loyalty as a merchandising lever (and how that can change the way you view the impact of your program), the success of any loyalty program starts with making sure it’s used by customers.
In most cases, you’ll want to look at points breakage here. (Breakage, if you’re not familiar, is the percentage of points earned that aren’t redeemed.) And that is one way to look at it, because it gives you a good view into the financial liability of the program.
Viewing in that fashion, though, can create tension between Finance (higher breakage rates means less redemptions, which means lower costs to fund the program) and Marketing (lower breakage rates means more redemptions, which means higher costs to fund the program). If you’re a marketer running a loyalty program, this misalignment might sink your collaboration before you can even prove whether your loyalty program rewards generate incremental revenue.
Another way to look at loyalty program participation, though, could be to look outside of breakage.
What if you were to look at the percentage of loyalty program purchases that included a points redemption versus the percentage of loyalty program purchases that didn’t include a points redemption?
This view, different from breakage, gets rid of the Finance-Marketing tension and puts more of an emphasis on participation at the purchase event level, as opposed to the points level. And since programs have varying degrees of value for points, this normalizes the measurement around purchase behavior, thereby putting less emphasis on points.
To get to this view, we normalized points to their dollar value, then looked at how many dollars a loyalty program customer needed to redeem a loyalty program reward.
We then used that dollar value as a way to rank a program by “ease of redemption” (or, put plainly, how much money a customer needs to spend to redeem points), and bucketed by quartiles. The 25th percentile represents the “easy to redeem” programs, the 50th percentile represents the “medium effort to redeem programs,” and the 75th percentile represents the “hard to redeem programs” in the dataset.
What you can see from the above is that, under the “easiest” programs, it’s often far easier to redeem rewards than even the “medium” programs.
That’s important.
Because when you start to look at participation rates, you notice that ease doesn’t translate in the way you’d expect. You might assume, for instance, that participation rates and “ease of redemption” would be inversely proportional, perhaps linearly so. Except we don’t see that.
Instead, what we see is that participation isn't tied so tightly to ease of redemption. While it's true that variable dollar off programs (which are easier to redeem) have higher participation rates, you can see by isolating to individual program types that participation is not fully dependent on ease of redemption.
This is an important point.
As with any merchandising lever, there are tradeoffs between adoption and business impact. And finding the right lever isn't always about optimizing for adoption.
The takeaway here?
While variable points programs can be an effective way to drive participation, they won't necessarily produce much boost to average order values. (In a previous newsletter, we showed that these programs deliver AOV lifts of 4% and basket size lifts of 5%, lower than percentage-based discount programs.)
That said, such a program can build habits around a loyalty program, and so those lower business impact metrics are buoyed—to a degree—by the fact that more customers participate.
However, within Fixed discount (percentage or $ off) programs, where we've historically seen more of a AOV/basket size benefit, there doesn't seem to be much correlation with ease-of-use vs. participation. This implies that brands leveraging these types of programs can further drive margin by making thresholds-to-redeem somewhat higher.
Based on these findings, we plan to dive deeper into whether higher participation rates—with lower AOV/basket size benefits—or lower participation rates—with greater AOV/basket size benefits—delivers optimum results.