Ten Questions Before You Pick a Business Model

This series has walked through why subscriptions took over, the value gate, freemium math, tier structure, billing cadence, AI pricing, the trust cost of dark patterns, and what games teach about monetization. This last note is the framework that ties them together — the questions I run a product through before committing to any of it.

The ten questions

Does the product have ongoing infrastructure costs? Yes points to subscription or usage-based; a pure one-time purchase probably won't survive. No opens up one-time or freemium.
Does the value increase over time? If customers get more the longer they stay, subscription fits. If not, a subscription has to re-justify itself every month.
Is usage predictable or variable? Steady usage suits a flat subscription; highly variable usage (especially AI) wants usage-based, or your caps become a liability.
Is the user an individual or a business? Consumers want freemium, ads, or a low subscription; businesses tolerate — and expect — per-seat or enterprise pricing.
Is the product habit-forming? High habit potential makes freemium work, because people convert after the habit forms. Episodic use tends to produce a big free tier and thin conversion.
Is there a network effect? Strong network effects make free or freemium almost mandatory — every user makes it more valuable. Without them, you can charge from day one.
Can advertising actually support the economics? Only at massive, broad-audience scale. For niche or B2B, ads will never cover costs — don't kid yourself.
Is it solving a painful problem or providing entertainment? Painful problems carry high willingness to pay (premium/subscription). Entertainment skews toward free, ads, or impulse purchases.
Is it ready for a paywall, or does it need distribution first? If you need reach, stay free or freemium until product-market fit. If you're ready, early revenue is validation.
What are your unit economics? Low marginal cost per user makes free tiers sustainable. High marginal cost (AI, media, anything compute-heavy) means every free user is an expense you must design around.

Four case studies worth stealing from

OpenAI — free as a distribution engine, then a subscription ladder (Free → Plus → Team → Pro → API) that captures value at every willingness-to-pay level. The lesson: you don't pick one model, you build a staircase.
Spotify — the freemium masterclass, but it took ~16 years to turn a profit. The lesson: the playbook works, but only at a scale and cost structure most of us will never have. Admire it; don't copy it blindly.
Adobe — financially successful, then sued by the FTC for the dark patterns it used to hold onto customers. The lesson: retention through lock-in and opacity is a liability that eventually comes due.
X (Twitter) — tried to force a free user base into paying by degrading the free product, and converted under 0.15%. The lesson: the subscription has to be pull, not push. People upgrade because they want to, never because you made free miserable.

What this means for a solo founder like me

Run my own products through the ten questions and the answers are consistent: real (if small) infrastructure costs, a genuinely habit-forming use case, individual users, high willingness to pay for the pain of lost focus, and low-enough marginal cost — because Amethyst is local-first — that a free tier stays sustainable. That points cleanly at freemium with a paid subscription, which is exactly what ResistGate and Amethyst run.

The condensed advice I'd give another indie builder from all of this:

Start paid from day one if you solve a painful problem and can find even ten users willing to pay $50+/month. Free growth with no monetization is just expensive experimentation.
Choose monthly over annual early — lower friction, faster truth about churn.
Add a free tier only when you have traffic. Below ~10,000 users it mostly means low revenue.
Treat lifetime deals as a specific, limited move, never the default.
Avoid ads unless you have millions of users, and never promise unlimited AI.

The lesson: there's no best business model, only the one your product's costs, users, and habits actually point to. The ten questions don't hand you an answer — they stop you from copying someone whose situation isn't yours. Choosing the wrong model is one of the most reliable ways to kill a good product. Choosing the one your own answers point to is most of the battle.

I'm Orlando. I build browser extensions and AI tools from Venezuela, including ResistGate and Amethyst. This closes the From Free to Premium series — nine notes on building a business model that funds the work without betraying the people who use it.