The pseudoscience of a 12 month payback on CAC
For most of the time I have been a VC the prevailing wisdom has been that companies, whether B2B or B2C, should aim for a 12 month payback of CAC.
To be clear on definitions here:
CAC = fully-loaded Customer Acquisition Cost (marketing, sales, onboarding)
Monthly contribution margin CM = Monthly revenues — monthly variable costs (COGS, account management)
Retention in month i= Rᵢ
12 month value (‘LTV’) of a customer = CM₁ R₁ + CM₂ R₂ + …+ CM₁₂R₁₂
Aim for CAC payback < 12 month value.
CAC payback time = the number of months required for customer value to exceed CAC.
Simple.
But why 12 months? It is a nice round number. But, as with the food pyramid it is based more on frequent repetition than any real science. For sure there is some logic that growing a business rapidly ends up requiring a lot of capital if CAC payback time is long. However there are important differences between a business with low retention versus one with high retention or where user revenues increase over time (net negative churn). In the former a CAC payback time of even 6 months might be desirable, in the latter a payback time of 24 months might be perfectly acceptable
There are examples out there of successful businesses with CAC payback times of well over 12 months. For example much of the wealth management and pensions industry. For a typical personal pension year one revenues are tiny. On a salary of £40,000 someone contributing 8% into a pension would put in £3200 in year one, so average assets under management for the year would be £1600. Taking net revenues for the pension provider at 0.75% of assets, year one revenues would be only £12. Yet a personal pension provider would be willing to pay much more to acquire that client. [note: I originally wanted to include data from some public companies here, but none that can I find publicly report their customer acquisition costs or unit economics…]
A more recent example are the two big insurtech IPOs of 2020, Lemonade and Root. Lemonade’s payback time on CAC is ~3 years given their retention rates according to my back-of-envelope maths (with help from Twitter). According to this analysis it is as long as seven years, but I think this is a bit too tough in its assumptions. According to their S-1 Root’s CAC payback time is also ~3 years (CAC of $332, $909 average premium, ~20% margin, 51% retention to year 2 and 65% to year 3*).
This has inspired much debate in the insurance industry, in particular as Lemonade and Root’s retention rates are not great. Despite this Lemonade and Root’s combined market cap in December was $9B. Both have built strong brands and are demonstrating improving underwriting performance and an ability to improve revenues per client. We also don’t know what long term retention will be: maybe the 50% of customers who remain after year two then stay forever. The standard in the personal lines insurance industry is a multi-year payback on CAC, with insurers offering teaser rates in year one but then ‘price-walking’ customers onto a higher rate at renewal (although the UK regulator is now cracking down on this).
Looking outside of fintech the mobile games industry is another interesting example of different payback times. Great mobile games keep players for many years. Candy Crush is an 8 year old game on mobile and (as of 7th Dec 2020) top 2 grossing in the US iOS app store. While there is plenty of user attrition in the first few months, the users who remain stay forever and on average pay more over time. I know of very successful mobile games which worked with a 24 month payback on CAC (example). On a smaller scale one of our regrets at the Wooga board was not spending more earlier on marketing for our best games, given their strong retention and improving ARPDAUs.
On the flip side in mobile games, if you make hypercasual ‘throwaway’ games (e.g. Voodoo or Kwalee) with a day 30 retention of 5–10% and day 90 of 1–2% you had better make sure you are paying back CAC very quickly. Good hypercasual studios aim for a maximum 30 day payback on CAC.
To be clear the shorter time it takes to pay back your CAC the better (and if you can grow organically even better). However if you are in an industry where this isn’t feasible, and where users have long lifetimes, then you should think carefully about whether 12 months is right for you.
Would love reactions to this and other examples of companies targeting payback times of more or less than 12 months
As further related reading I can recommend this recent post by Christoph Janz: https://medium.com/point-nine-news/why-your-ltv-might-be-higher-or-lower-than-you-think-f35539291701
