Why I Don't Offer a Free Plan
Free plans attract people who'll never pay, create disproportionate support, and dilute focus from the users who actually value the product. For a solo builder, a free plan is a tax on attention. A free trial, yes — a free plan, no.
The growth playbook treats a free plan as table stakes: widen the funnel, let people in for nothing, convert some slice of them later. I don't run one — on Delivery Sheet or anywhere else — because a free plan attracts people who'll never pay, creates support out of all proportion, and pulls my focus toward users who don't value the product and away from the ones who do. For a solo builder whose scarcest resource is attention, a permanent free tier is a tax I can't afford. This isn't anti-free dogma; a free trial is often the right call. It's the forever-free plan I avoid, because it quietly optimizes for the wrong users. I'd rather have a handful of people who need the product enough to pay than a crowd who don't.
The appeal, and the trap
The pitch is seductive: more signups, more top of funnel, more growth. And the number does go up. But the number that goes up is the count of people who wanted something free — a very different population from people who'll pay to solve a real problem. You optimize the vanity metric and inherit its users.
The hidden cost — and why it's worse solo
Every free user costs you support, infrastructure and — the expensive one — attention, against zero revenue. A funded team books that as growth spend. A solo builder is spending the one thing no amount of money buys back.
Free users aren't free. They file support requests, consume infrastructure, and — most costly — pull your attention toward feedback from people who, by definition, don't value the product enough to pay. A funded team can file that under acquisition cost. Solo, it's your own finite attention going to non-buyers — the exact attention your paying users and your other products are waiting on.
Free trial, yes. Free plan, no.
The distinction is everything. A free trial is time-boxed: full value for a set period, then a decision. It qualifies buyers — someone who tries and pays actually needed it. A free plan is permanent, and it accumulates non-buyers indefinitely. A money-back guarantee does the same job as a trial: it removes the risk of trying without leaving behind a standing population that will never convert. I want people to try with confidence. I don't want a permanent tier of users who cost me and never pay.
When free genuinely makes sense
To be fair to the other side: a free plan earns its keep when the product needs scale to work at all. Network effects, where more users make it better for everyone, or virality, where free users bring paying ones, make free users part of the mechanism rather than a cost. If that's your product, free users create value and the math flips. But a focused tool that solves a specific problem for a specific person doesn't get better because more non-buyers use it — so the honest move is to charge the people who have the problem.
Pricing as a filter
There's a deeper reason I charge from the start: price is a filter for who actually has the problem. Someone willing to pay is telling you the problem is real and worth money to them — the same signal I look for before I build anything at all. Free removes the filter and floods you with feedback from people who don't have the problem badly enough to act. Charging keeps the feedback, the users, and the attention pointed at the people you're actually building for.
What usually goes wrong
- Chasing signup counts. Optimizing the vanity metric with a free tier and inheriting a population of non-buyers.
- Confusing trial with plan. Reading "let people try" as "free forever," when a time-boxed trial does the job without the standing cost.
- Free on a focused tool. Offering free where free users create no value — only support and diluted focus.
- Free where you needed scale. The opposite error: charging from user one on a product that genuinely needed network effects to work.
Offer a trial or a guarantee so people can try with confidence, charge the people who have the problem, and skip the permanent free plan unless free users are genuinely part of how the product works. Do that and your scarce solo attention stays on the users who actually value what you built. Free isn't generosity when it costs you the focus your real customers need. It's a product decision like any other, and for a focused tool the answer is usually: charge.
Part of Building AI Products. See also how I decide subscription or one-time pricing and why I built Delivery Sheet. The newsletter sends one practical build lesson every two weeks.