The truth about sunk cost fallacy

By Grace Segran

SMU Office of Research & Tech Transfer– Launched in 2005, Amazon’s membership service, Amazon Prime,has more than 200 million Prime subscribers worldwide at the beginning of 2021 – 50 million more than last year. For US$112 a year, members get free expedited shipping, access to Prime Video and Music, unlimited cloud photo storage, exclusive deals, discounts on groceries, and other benefits. For many, the program is a no-brainer – the return of investment from the benefits is more than the fee they pay.

Subscription programs are designed to keep customers engaged by giving access to exclusive benefits for a fee upfront. Other retailers have jumped on the bandwagon with similar programs such as Barnes & Noble’s B&N Membership,Sephora’s Flash,and Alibaba’s 88VIP,offering benefits that range from unlimited free shipping to member-only discounts and additional loyalty points.

There has, however, been a long debate on whether they really generate value for firms. “Some claim that the effect of subscription programs is largely overestimated. They believe that the programs do not really change members’ behaviour. Rather, they screen out the price-sensitive customers,” SMU Assistant Professor of Marketing Qi Yu told the Office of Research and Tech Transfer. “Others are worried that the additional sales are generated only by reducing (effective) prices, and the program might negatively affect a firm’s performance in the long term.”

These arguments provided the impetus for Professor Yu’s working paper “The Impact of Subscription Programs on Customer Purchases”.

Research areas of the new addition to the faculty at the SMU Lee Kong Chian School of Business include quantitative marketing, competitive strategies in retailing, and subscriptions.

Data collection and analysis

The team, consisting of Professor Yu and two other professors, was interested in addressing the following questions: Does a subscription program generate value for a firm? Is the subscription program effective in inducing customers to change their behaviour because of the economic benefits and/or the psychological drivers? How do these effects vary over time and across customers? What are the underlying drivers of any documented effects?

The researchers worked with a company which actively experiments with new business ideas. It launched a subscription program on its e-commerce website in December 2015. The program has a standard structure, and there is very detailed individual-level data, giving the researchers an excellent opportunity to study the subscription business.

Data obtained from the company’s management and data science team included very detailed information on customers’ transactions, program adoption, and program usage. “We applied a machine-learning-based method to evaluate the causal impact of the program at the granular level,” says Professor Yu. “We then examine whether the purchase patterns are consistent with the predictions by various mechanisms.”

Assumptions and findings

When the researchers conceptually compare the subscription programs with existing types of membership programs such as loyalty programs, one distinguishing feature is that subscription programs involve an upfront fee.

“In some cases, this fee makes up a substantial part of a company’s revenue and it is a stable cash stream. But in many other cases, the benefits of the program are so generous that the upfront fee doesn’t seem to make up for what the company has given out,” Professor Yu says.

Then why are companies doing this?

“We hypothesise that the subscription fee may lock in customers through a psychological process called ‘sunk cost fallacy’. The subscription fee appears to be an irreversible cost to the customers and should not affect subsequent purchase decisions (only the effective prices should),” Professor Yu argues. “However, in reality, not all customers are rational. They tend to hold the initial cost in a mental account and increase their purchases in order to utilise the program benefits and amortise the psychological burden of the cost. We find the data patterns support this hypothesis.”

The researchers found that subscription leads to a large increase in customer purchases whereby the effect of subscription is economically significant, persistent over time, and heterogeneous across customers. Interestingly, only one-third of the effect on customer purchases is due to the economic benefit of the subscription program and the remaining two-thirds are attributed to the non-economic effect.

 There were two surprising results. One is that consumers not only care about the prices they pay but they also keep track of the benefits brought by the subscription programs.

Secondly, one would expect the customers with the highest amount of past purchases to also be the most valuable customers for the subscription program. However, that was not the case.

“The high-value customers based on past purchases, while enjoying program benefits, increased their purchases only moderately, which is likely due to a ceiling effect,” Professor Yu avers. “Rather, the members who contributed the highest profit for the firm upon subscription were relatively less active customers in the pre-subscription period but were willing to explore new products, had repeated purchases, and were responsive to promotions so that they re-engaged with the firm post subscription.”

Applications

While this study examines one subscription program launched by a single firm, Professor Yu says that the insight that the effect of subscription goes beyond the economic benefits of the program applies to many other settings. For instance, subscription programs offered by companies in other industries, e.g., Grab, have similar effects. Credit cards with annual fees can also be viewed as subscriptions.

“Our results also enable us to derive recommendations related to the design of subscription programs. For example, given the sensitivity to sunk costs, it may benefit companies to make the initial payment more salient after customers become members,” she continues. “Our suggestion is in line with past work by Gourville and Soman (2002) that shows that making prices salient can make members consume a service on a consistent basis. As members’ enhanced status encourages purchases, managers can emphasise the exclusiveness of the program to boost sales.”

The researchers’ analyses also contribute to the ongoing debate concerning the amplified friction in the retail industry due to the rise of subscription programs (e.g., Amazon Prime), says Professor Yu. “Our results suggest that firms are able to lock in customers with subscription programs by creating a sunk cost fallacy which may potentially lead to increased market concentration.”

“The findings of this study apply to a wide range of subscription programs that share this fee structure,” adds Professor Yu. “With that said, the design of subscription programs may depend on the competitive landscape and would likely be country specific. For instance, shipping fees in many Asian countries are generally low. Therefore, programs that centre around free shipping like Amazon may not work as well.”

When the pandemic began, many consumers leaned into subscription-based products as they moved online. According to Professor Yu, this period is long enough for consumers to form new habits and the trend is likely to continue post-COVID.

Back to Research@SMU Aug 2021 Issue