One-time payment or subscription: how I actually decided

Before I had a single line of code written I had already decided Appray would be a subscription. Every app I admired ran one. Every "how to price your app" post I had read for the previous six months assumed recurring revenue as the only serious option and treated a one-time price as something you did before you learned better. I did not question it, I just copied it, the way you copy a lot of things when you are deciding thirty things a day and only have real conviction about three of them.

The spreadsheet that made the decision for me

I built the first spreadsheet to pick a number, not to question the model. Three tiers, monthly and annual, churn assumptions borrowed from a SaaS benchmark post that had nothing to do with an app someone opens four times and then puts down for a month. The spreadsheet produced a monthly revenue line that looked fine on a slide and meant nothing, because it assumed renewal behaviour from people using a tool that, by design, has nothing left to do once it has told you Build, Kill or Prove.

The second spreadsheet asked a different question: what does someone actually get in month two that they didn't already have in week one? For a project management tool the answer is obvious, new data keeps arriving, you keep needing the dashboard. For Appray the honest answer was close to nothing. You run your feature through the eight rounds, you get a verdict and the reasoning behind it, and unless you are starting a new project there is no reason to open the app again until you have another feature in mind. Billing someone monthly for that is billing them for a service they are not using most months, and eventually they notice.

What stopped me wasn't guilt, it was the cancellation

What actually moved me wasn't an ethical objection to subscriptions, plenty of them are earned. It was picturing the exact moment someone cancels. They open Settings, find Appray, tap cancel, and the reason they type, if the flow even asks, is going to be some version of "I only needed it once." That is not a complaint about the app. It is a correct description of how a verdict tool should work. Building a pricing model that depends on people forgetting to cancel a thing they used correctly is not a business decision I wanted to defend to myself at month six.

I went looking for how other indie developers had worked through the same question, and the honest threads on Indie Hackers say roughly the same thing from different angles: subscription fatigue is real, and it bites hardest on tools with no ongoing service component, no sync, no content that updates, nothing that costs you money to keep running for that user after the sale. One founder put it plainly, they'd rather charge once and be done than run a recurring relationship with no recurring reason behind it. That matched what my second spreadsheet was already telling me.

What stayed free, and why

The base app gives you the full interview, the full verdict, saved history and a Markdown export of the result, for nothing. That wasn't a marketing decision either, or not only. If the product's whole claim is that it helps you decide before you waste months building the wrong thing, putting the actual decision behind a paywall undercuts the claim before anyone has a reason to trust it. The free tier has to be the real thing, not a demo of the real thing, or the pitch is dishonest.

Pro, at 4.99 dollars once, is for people who are doing this seriously enough to want more than one project running, a cleaner PDF layout for a verdict they're going to show someone else, and a way to compare two runs against each other. None of that is required to get the honest answer out of the app. All of it is useful once you're doing this more than occasionally, which is exactly the group I am comfortable asking money from.

Picking the number itself

I tested 2.99, 4.99 and 9.99 against myself more than against anyone else, by asking what I would pay without resenting it for a tool I'd use three or four times total before either building the feature or dropping it. 2.99 felt like it was pricing the work down to nothing. 9.99 felt like I was pretending Appray was something with ongoing upkeep it doesn't have. 4.99 was the number that didn't make me flinch either way, which is a worse method than I'd like to admit and still better than copying whatever number a competitor had picked.

Where I still have doubts

I will not pretend the revenue math is settled. A one-time unlock means every Pro sale is the only money that person will ever give me, so the lifetime value of a subscriber, had I shipped one, would likely beat a one-time buyer on paper. If I ever add something genuinely ongoing, a yearly refresh of interview questions, continuing support for a running comparison across many projects, I would reconsider, and I would say so plainly rather than sliding a subscription in quietly on top of a model people bought into as one-time. Changing the deal after someone has already paid once is its own kind of mistake, and I've seen enough of it done to other apps to know exactly how it reads from the other side.

For now the model matches what the app actually does: one decision, one price, no ongoing bill for a problem that got solved the day you opened the app.