When Do Subscription Models NOT Work?

On the phone a few days ago, a founder gave me an interesting piece of wisdom: subscription businesses don’t work when it costs more to retain a customer than it does to acquire one.

I was thinking more about this tidbit today, and it struck me just how many components of it are worth exploring. So, let’s do it! First, I want to unpack what it means for customer acquisition to actually cost less than customer retention. And, I want to further explore whether this is the only condition under which “technology-enabled” (I hate that term) businesses don’t work.

So, the statement itself. To say that it costs more to retain customers than acquire them is borderline heretical. And, generally, it’s wrong. The probability of selling to an existing customer is 60%-70%, while the probability of acquiring a new one is only 5%-20%. However, under the right set of circumstances, customer retention can increase and eclipse customer acquisition. Let’s think about those circumstances:

• Retaining customers has to be really hard, making it expensive. This is most likely to happen because a customer no longer needs a product service. He or she has used a company for a specific “job to be done” or task, and now that the job is done, convincing the customer to buy again is nearly impossible.
• Acquiring customers has to be really easy, making it cheap. This can be complicated to analyze, given the price of a product will undoubtedly impact customer acquisition costs. However, assuming some standard margin on top of COGS is applied to generate price (that conforms to industry pricing standards), we can still apply some logic to this. Products and services with relatively low downside risk that promote some positive outcome and do not require a lot of work, all while being fairly priced, will have low customer acquisition costs.

So, are there actually any products like this? Yes, there are, but they are really rare. The key here is that we’re not considering products that wouldn’t be considered subscription-worthy anyway. That is, products that don’t “run out” and therefore don’t require a subscription, are not kosher. As an example, you wouldn’t subscribe to shoes or musical instruments. Those are items that can theoretically be used forever, rendering a subscription counterintuitive (although maybe Allbirds will prove me wrong...).

For subscription-worthy products, we need to identify items that, once consumed and completely finished, fulfill a specific task and are likely unnecessary after fulfilling that task. At the same time, customer acquisition costs have to be low for the reasons aforementioned. Off the top of my head, the products that stand out are: performance-enhancing or wellness products tailored to a specific time period or life stage. Products in this broad category include cleansing or dieting foods and supplements, as well as consumables to prepare someone, for example, for running a big race or having a baby.

In these rare cases, where companies offer something that gets totally consumed and solves a specific and short-term “job to be done,” retention costs can be greater than acquisition costs because, simply put, retention costs can be infinite. If I buy protein shakes to help me prepare to run a marathon, I have little incentive to continue the regimen after my marathon is over.

Importantly, this is only one rare condition in which consumer goods are not fit for subscription. In these cases, there’s no sense to subscription because there’s no appeal to a long-term product commitment. There are a bunch of other, more obvious cases where this is true for all businesses (including “tech-enabled” ones that live online), the most obvious of which is one-off purchases, which actually make up more of the “things” you buy than you might think.

Overall, I think this case of customer acquisition costs being less than customer retention costs is an edge case, but an interesting one to explore. Hopefully I can think of some more cases where it holds true.