From Amazon Prime to Panera Bread’s Unlimited Sip Club, loyalty programs have become commonplace in a wide variety of industries. However, despite their popularity, it’s not always clear whether these programs are actually profitable. The authors analyzed data from 24,000 customers at a large Asian retailer to explore the various factors that can drive a program’s profitability, and identified three takeaways to help managers ensure their subscription programs are a net positive: They recommend that managers should go beyond averages and analyze trends on an individual level, measure the underlying factors that drive changes in revenues, and remember to account for increased costs associated with program benefits. Ultimately, they find that loyalty programs can be quite effective — but only if managers carefully track revenues and costs for different customer segments, and experiment and adapt accordingly.
In February 2020, Panera Bread announced the Unlimited Sip Club and rocked the coffee world. For just $8.99 a month, members could get unlimited refills of their favorite coffee or hot tea at any Panera location. With the average American consumer spending more than $2,000 a year on coffee, this sounded like an incredible deal for coffee drinkers — begging the question, how could Panera possibly justify such generosity?
Loyalty programs like this are common in industries as wide-ranging as retail, food and beverages, health and wellness, and more. Proponents argue that increases in sales justify the costs — for example, while Panera may lose money on its drinks, if loyalty members buy a croissant every time they get a coffee, the program could still be quite profitable for the company — but these effects can be hard to quantify.
For example, the average Amazon Prime member spends more than twice as much as the average non-member. This may seem to imply that the program is responsible for a substantial increase in revenues, but the difference in spending could also be driven by self-selection (that is, customers who already intend to spend more may be more likely to choose to become members). The costs to the company of providing membership benefits can also be substantial, whether that’s a free cup of coffee, exclusive product offers, free shipping, or other perks. In light of this complexity, how can managers ensure their loyalty programs are truly profitable?
To explore this question, we conducted a large-scale study in partnership with a major Asian retailer that was launching a new, paid loyalty program in 2015. We analyzed 15 months of transaction data for more than 24,000 customers, about half of whom had joined the program, and identified three key takeaways for managers:
Go beyond averages and analyze trends on an individual level.
On average, we found that customers spent more than twice as much per month after subscribing to the retailer’s loyalty program. However, this increase was not evenly spread out across all the members in the sample: Some customers contributed substantially to the increase in revenues, while others did not. This illustrates how a better understanding of whose spending is likely to increase the most after joining a loyalty program can help marketers better target prospects going forward.
To do that, managers must go beyond average results to track and analyze changes in purchasing patterns on an individual level. That means setting up data collection and analysis systems, allocating the necessary resources internally, and getting buy-in from the relevant stakeholders. And importantly, this all should happen before a company introduces a new loyalty program, to ensure managers have the benchmark data that will be necessary to measure the program’s impact.
Don’t just measure changes in profits — measure what drives those changes.
In addition to drilling down into individual-level spending, our research emphasized the importance of developing a nuanced picture of the various changes in consumer behavior that may be driving shifts in overall profits. For example, we found that post-subscription, customers bought a wider variety of products: Approximately 75% of the increase in revenues came from new products that customers had not previously bought, suggesting that one factor contributing to the program’s success was that it encouraged members to start exploring products beyond their typical purchases.
Conversely, we also found that while the total number of purchases increased after the introduction of the loyalty program, average basket size decreased. This was likely because membership included free shipping, meaning that customers were no longer incentivized to group purchases together into a single shipment. This change in member behavior could have substantial cost implications for the firm, as shipping costs increase when members spread out their purchases between more shipments — and if members incur greater costs, higher revenues may not translate to higher profits.
When evaluating a program’s impact, don’t forget about the costs of serving customers.
This last example highlights the importance of paying close attention not only to changes in revenues, but also to the various costs associated with a loyalty program. That means both the cost of the additional goods sold and the costs of any benefits included in the program, such as free shipping or member-exclusive offers.
After we factored these costs into our analysis, we found that net profits increased a lot more for some customers than for others after becoming members. Specifically, we found that 14% of the members in our study contributed the highest profits despite incurring high costs for the retailer, 46% generated a sizable increase in profit without as much cost, and the remaining 40% generated minimal profits and incurred substantial costs.
Interestingly, the 14% of members who contributed the highest profits after subscribing were not the ones who had purchased the most prior to subscription. Rather, they were customers who had been less active before subscribing, but who had demonstrated interest in exploring new products, made repeated purchases of similar items, and were responsive to promotions. In contrast, the customers who had purchased the most pre-subscription only moderately increased their spending while still enjoying all the benefits of the program, suggesting that the retailer might have been better off if those customers hadn’t joined.
This analysis highlights the importance of monitoring changes in both revenues and costs on an individual level. Rather than assuming costs stay constant with respect to the value of goods sold, managers should track how costs change for different customer segments after the program is introduced — and then target loyalty program promotions at the customers who are likely to generate the highest net profits.
If implemented well, paid loyalty programs can be highly profitable. Indeed, after launching its popular subscription program, Panera announced that food purchases attached to coffee orders grew by 70%. But to ensure that a loyalty program is actually a net positive for the company, it’s critical to track revenues and costs for different customer segments — as well as the underlying factors that may be driving both — and experiment and adapt accordingly.
All authors contributed equally and are listed in alphabetical order. We thank several managers at the participating company for their efforts in making this research possible.