In the last article, I showed how AI helped me navigate an unfamiliar domain: building in-browser vector search. We ended on a question: if the architecture costs $0/month in infrastructure, what do I charge?
Now I'm wearing a different hat. Not engineer. Business person.
Let me be honest about a few things. 1. I revised this article 3 times, almost rewrote everything. You'll feel it in the later section.
2. Subscription is the default model for tools like this. My target users are already paying subscriptions everywhere: ChatGPT, Claude, Notion, Figma. They're used to it. The math is well-understood: recurring revenue, long-term value, predictable cash flow.
From a pure business standpoint, subscription is the right answer.
So why didn't I choose it?
The Emotional Truth
This is my first product. My first one-person company. My first time charging real money for software I built.
I'm confident ChatShuttle solves a real problem. I use it myself. I know my target users will benefit from it. But I had a fear I couldn't shake:
Would they actually pay a monthly fee for this?
Not because the value isn't there. But because I'm new to this. I've never done it before. There's a shyness that comes with charging money for the first time. An embarrassment, almost. You're asking strangers to take out their wallet and trust you.
Subscription feels like asking for ongoing trust. One-time feels like asking once.
For my first product, asking once felt easier to stomach.
The Rational Layer
Underneath the emotional hesitation, there was also logic.
Usage frequency matters. ChatShuttle is not something you use eight hours a day like ChatGPT. It's a utility. You import a chat, search for something, restore a thread when you need it. Maybe a few times a week. Maybe less.
When users pay $10/month or $20/month, they compare that to the $20 they're already paying ChatGPT. If they're using ChatShuttle 20 minutes a week, that comparison doesn't feel fair. They start calculating: Is this worth half of what I pay for my primary AI tool?
That friction is real.
My architecture gave me freedom. Because I chose a local-first architecture, I have no backend server costs. No database hosting. No per-user infrastructure. The cost to add one more user is near zero.
That freedom meant I could be... opinionated. I didn't have to charge subscription to cover recurring costs. I could choose.
Subscription requires peripheral work. If I wanted to do subscription properly, I'd probably need to build or integrate billing infrastructure, churn tracking, failed payment handling, dunning emails, upgrade/downgrade logic. None of that is impossible. All of it is time. And as a one-person company, time is the constraint that kills you.
I wanted to spend my time on product, not billing dashboards. At least for now.
What this product means to me. ChatShuttle solves my own problem. That's where it started. But it's also more than that. This is my first one-person company. My first public build. I'm walking through the entire process: product, engineering, marketing, pricing end to end, in public.
The financial outcome is not the (sole) main priority. The learning, the completion, and the experience of building something real and putting it into the world, are equally important.
That context shaped my pricing decision.
A Reflection
I have an MBA from an Ivy League school. I've studied pricing strategy, customer acquisition, lifetime value. I know the frameworks.
But when I actually made this decision, I didn't follow a framework. I made it through a messy mix of rational analysis, emotional uncertainty, and personal values. I think that's the curse of one-person companies. I think that's the beauty of one-person companies.
The fear of rejection played a role. The shyness about charging played a role. The desire to finish my first product end-to-end played a role.
I don't know if this was the right decision. One-time pricing will almost certainly generate less total revenue than subscription.
But I'm at peace with it. For my first one-person product, I wanted to complete the entire journey: build, ship, charge, promote, record, and learn.
This is self-awareness, not strategy. I'm saying out loud how I made the decision so that I can learn from it later.
The Framework (Revised)
If I were to distill what I learned into something reusable, it would be this:
Step 1: Start from user value, not developer cost.
What does the user need? What problem are they solving? For ChatShuttle, the core value is: import your chat history, search it, restore context to a new AI. That's the baseline.
Step 2: Ask what makes that value easier, not just what costs you maintenance.
The original version of this article focused on "maintenance-heavy features = paid." That's a developer perspective. But users don't care about your maintenance burden. They care about their experience.
Pro features should be the things that make the core value more efficient, less restricted, smoother to use. Image restore means the new AI sees everything. Unlimited slots means no artificial barriers.
Pro also unlocks more use cases. With ChatShuttle Nexus, power users can connect their web AI conversations to their IDE workflow as a searchable AI memory layer. Your Claude or Cursor agent can now search your past ChatGPT threads for context. It raises the ceiling of what ChatShuttle can do for someone who uses AI across multiple surfaces.
The output was $8.99 one-time. About the price of a bubble tea in San Francisco.
Again, I could be totally wrong about this. Someone else with different constraints might land on subscription. That's fine. The point is to know what's driving your decision, emotional and rational, and own it.
What's Next
This article covered the business side: pricing and the messy reality of making that decision.
Next, I want to go back to the technical side. How do you actually capture a conversation from a web page? ChatGPT, Claude, and Gemini don't have public APIs for this. You have to scrape the DOM. And DOM structures change constantly.
That's the Snapshot feature, and it's the part that requires the most ongoing maintenance work.
Want to see what the free vs. Pro split looks like in practice? The FAQ covers what's included in each tier.