An increasing number of goods and services that were once sold as one-off purchases are now offered as subscriptions. Newspaper subscriptions and gym memberships have existed for a long time, and 'software as a service' is now commonplace. But the model has spread much more widely: consumers can now subscribe to meal kits (such as HelloFresh), razors (such as Dollar Shave Club), and a growing range of other products. Why are firms that once sold products outright increasingly choosing to sell subscriptions instead?
That is the question addressed in this 2025 article by Liran Einav (Stanford University), Ben Klopack (Texas A&M University), and Neale Mahoney (Stanford University), published in the prestigious American Economic Review (ungated earlier version here). They start by noting that the rapid growth in subscriptions is often attributed to the rise of digital products, and the convenience of a subscription for consumers. However, Einav et al. focus their attention on a third factor:
Because subscriptions are automatically renewed, consumers who are inertial may continue to pay for subscriptions they no longer value... If consumers do not fully anticipate their inertia at sign-up, this may create supply-side incentives to offer subscriptions to exploit inertial consumers, amplifying the growth of subscription offerings.
My ECONS101 students will be familiar with this explanation for subscriptions, because we literally covered this in the lecture today. Einav et al. test for the extent to which inertia matters using transaction data from "a large payment card network in the United States between August 2017 and December 2021". Their final dataset includes over 800,000 accounts, and about 870,000 account-service pairs (each account-service pair is a set of observations of a payment card account that subscribes from one of the ten largest subscription services).
Einav et al. exploit the fact that when a card expires and is replaced, consumers typically have to update the billing information for their subscriptions, prompting them to either update or cancel each subscription. To the extent that card replacement decreases the retention rate of subscriptions, this provides evidence of customer inertia. If consumers cancelled subscriptions whenever they stopped making use of them, then there would be no difference in subscription retention between months with card replacements and months without.
Unsurprisingly, Einav et al. find evidence of customer inertia, and the effects are large and consequential for firms selling subscriptions:
We use the estimated model to perform counterfactual exercises that assess how much more quickly consumers would cancel their subscriptions if there was no inertia, which corresponds to fully attentive consumers (inattention model) or default cancellation every month (switching cost model). We find that seller revenues (or equivalently average subscription durations) are significantly higher due to subscriber inertia with important heterogeneity across services. Specifically, in the inattention model, we find that inertia increases seller revenues by 87 percent on average, with increases that range from 14 percent to more than 200 percent depending on the service. In the switching cost model, inertia raises revenue by 120 percent on average, with a range of 17 percent to 259 percent.
So, there are strong incentives for firms to engage in the selling of subscriptions, and to take advantage of customer inertia in subscriptions. However, many consumers are clearly spending more on subscriptions than they need or necessarily want to. Think about yourself as an example - how many subscriptions do you have right now that you rarely use and probably should cancel? I don't have any, but that's only because writing this post made me think about this and cancel one that I was no longer really using!
Subscriptions can provide important benefits for consumer though, including reducing transaction costs (it is simpler to pay a monthly subscription than to buy goods or services individually over and over), and reducing service interruptions (because a subscription makes it more likely that the consumer won't run out of the good they are buying a subscription for). However, we might still be concerned that customer inertia makes some customers with subscriptions worse off overall. So, Einav et al. then turn to evaluating what the most appropriate policy response is. They focus attention on a rule requiring firms to provide consumers with an active renewal decision at regular intervals. Using their two models, Einav et al. find that:
In the inattention model, we find that requiring active choices at a six-month frequency would reduce the excess revenue from inattention by 45 percent. The switching cost model makes a similar quantitative prediction; moving from default renewal to default cancellation once every six months would reduce excess revenue by 48 percent.
Those are quite substantial effects, which again illustrates just how much consumers are giving away to subscription firms for subscriptions that they no longer make the best use of and should be cancelling. What becomes clear from this paper is that one important reason why firms that previously would have sold products instead prefer to sell subscriptions is that consumer inertia can make them substantially more profitable.
Read more:
