Tldr : Stop using recurring revenue as justification for negative unit economics
I think we have bastardized recurring revenue too much. Recurring revenue is really cool because each year you start with existing flow rather than a clean slate.Assuming that you have 100% retention, if you had $100K revenue in the first year, you will start with $100K. If you grow 100% again, then you will end up with $200K revenue at the end of this year.
But people stretch what this recurring revenue should be doing. Instead of it being a hedge against starting off clean and having to sell everyone again from ground up. People use it to stretch cash flows. The best assumption with which you can conduct a recurring revenue business is to assume that there will be nothing that is recurring. But people blow money on customer acquisition because they think there will be a payback, even if it takes 6 months or more. You have to design your business such that it can have recurring revenue, but you cannot assume that it will recur. Thus, the first payment should be enough to have positive cash flow. You cannot lower your subscription value and hope that the customer lifetime will be long enough that you will get payback and profits.
You can look at a lot of retail businesses. They do provide subscription services, but they do not use such a low entry price that you can put them in negative cash flow. They still recover their margin right at the first transaction, right at the second transaction, and every transaction that will take place. They will give up some of the margin to have the recurring nature, but they won’t put themselves at the risk of a multi-period payback. If you are doing business, anything greater than 1 as payback period should scare you. You are not in fucking real estate. You are not building infrastructure. That’s why we had the SaaS crash of 2022. Because you can put up this farce of long payback periods as long as the interest rates are low, but you will die under pressure if it goes away.
I don’t know how we bastardized the notion that Hey, we don’t have to start from scratch every year to Hey, we can just lose money right now; we’ll get it back with recurring revenue anyways. Like for infrastructure, long payback periods make sense. The infrastructure is not going anywhere. It’s going to pay you. But you are creating software, something so fucking fickle, like it is so vulnerable to changes that the business you have today might evaporate tomorrow. You are vulnerable to rapid obsolescence. So, why would you want to front-load a lot of cash burn and have a long payback period? Design for recurring revenue, but don’t depend on it
And I bet that, it’s not like you cannot create something with a payback period of just one and not have customers or something like that. It’s an offer issue. I despise long payback periods.
Plus, what exactly are you trying to do with high payback periods, high CAC, and cash burn? The hope in the long term is that the CAC collapses because you put your product in the hands of people who’d appreciate it enough that they tell about it to others. That the CAC spent to generate trust among users that hey, we can help you. And those users then translate that into positive word of mouth, and we won’t have to spend as much on ads and thus CAC will collapse. You are relying on the fact that you can spend your way into making people talk nice about you. That the way you spend money on ads is putting your product in the hands of people who not only want it but would pay for it and will talk about it in the future. From the start, you’re spending unsustainable amounts on CAC.
We’ll spend unsustainably high CAC now → put product in users’ hands → they’ll love it and tell others → CAC drops → we become profitable
You really don’t see anything that’s wrong with this? You’re essentially saying “our product is so good that once people use it, they’ll sell it for us” while simultaneously saying “our product isn’t compelling enough to charge profitably for upfront.” The high CAC isn’t putting the product in the hands of people who genuinely value it enough to evangelize. Its putting it in the hands of anyone willing to try it at a subsidized price. You’re building a user base of tourists and cheapskates, not passionate advocates.
How do you make software that is cash flow positive on the first payment?
Well, go back to the traditional ways: either charge a lifetime fee or get annual payments, so that you are cash flow positive right at the first instance. The service cost is not much when it comes to software. It’s CAC. If your average order value is greater than CAC, which you can get by having only annual plans, then you are good. Don’t chase revenue because exits are not as easy as before. You might have good revenue and good multiples, but people might not want to acquire you. So chasing revenue is not going to work. Chase profit.
This way you can actually have a group of evangelists who might give you positive word of mouth, so that you will have lower CAC. You have to put a filter that says “we are not allowing free users or monthly payment people around our app because they are going to most probably use and bounce, and all the CAC we spent won’t be used properly. If we target just the people who will actually pay for annual plans, there’s a greater chance that they will talk better about us, and we will be cash flow positive. That’s a Plus! Higher barriers create better communities.
Plus, after studying the AER model, acquisition does not really contribute that much to the growth in the long-term. It is coming from retention. So, if the funnel at the start is collecting people who want subsidies to even try it, how are you going to bet that the retention is going to happen? You deliberately have to put high barriers while onboarding people so that you can optimize for retention. Long payback periods is fundamentally an acquisition tool with not much thought given to retention. If growth comes from retention (which AER proves it does), then acquisition should optimize FOR retention, not against it. Don’t start unless you and the buyer have high trust. Demand higher prices and ask for longer commitments. If someone needs subsidy and low time commitment, then most probably they are not fit for your business. You should spend resources to find someone else. You have a real opportunity cost here.
From claude
Annual-only acts as a filter:
- Weeds out tire-kickers and feature tourists
- Self-selects for customers who actually have the problem you solve
- Attracts people committed enough to bet on you for a year
- These are the same people likely to care enough to tell others
The Compounding Effect:
- Charge annual → only serious buyers sign up
- Serious buyers actually use the product (they paid real money)
- They get real value (because they’re the right customer)
- They tell others who are also serious (like attracts like)
- Your CAC drops because word of mouth kicks in
- You’re profitable AND growing organically
The True Cost: It’s not just that they don’t pay enough. It’s what they consume:
- Support time (they have questions but no commitment to figure it out)
- Product roadmap (you build features for people who’ll leave)
- Mental energy (you’re serving people who don’t really care)
- Cash (their CAC never gets recovered)
- Positioning (you become known as “the cheap option”)
The Opportunity Cost: Every hour spent on wrong-fit customers is an hour NOT spent on:
- Finding customers who will pay full price
- Serving customers who are already committed
- Building features for people who’ll stay
- Creating content that attracts serious buyers
- Deepening relationships with evangelists
You’re Trading Gold for Copper:
- One annual customer at $5K > ten monthly at $50
- Not just in revenue, but in everything else
- The $5K customer refers others, engages deeply, stays for years
- The $50 customers churn, complain, and tell no one