Unlimited Free Forms Were the Whole Marketing Budget

Tally, a form builder run from Belgium by Marie Martens and Filip Minev, charges nothing for the things every competitor meters: forms and submissions are unlimited on the free plan. About 2 percent of users ever pay. That sounds like a business with a hole in it, yet Tally reached $5 million in annual recurring revenue in April 2026, bootstrapped, with a team of 11, according to a detailed April 2026 breakdown built on the founders' own numbers. The model works because the free plan is not a cost center with a conversion problem. It is the entire marketing budget: every free form published carries the product to the people who fill it in.

Three and a half years to $100K a month

The founders have published their revenue history on Tally's own blog, which is where the figures below come from (self-reported, March 2024, checked 3 August 2026).

Growth to that point was 100 percent organic: no ads, no sales team. Since then the curve has steepened, not flattened: $4 million ARR in October 2025, $5 million in April 2026, with over a million users and more than 100,000 new ones arriving each month. Note the shape, though. The first $1,000 of monthly revenue took seven months. The first year of a genuinely free product is a year of working for free.

The one discontinuity in the curve was earned the ordinary way: the Tally 2.0 relaunch in September 2023 went through Product Hunt, took product of the day, week and month plus a Golden Kitty in the bootstrapped category, and lifted weekly signups from about 2,300 to about 4,400, a 91 percent jump that never fully receded. A launch spike on top of an organic engine compounds; a launch spike instead of one just decays.

What free actually sells

Look at Tally's pricing page (checked 3 August 2026) and the design becomes clear. The free plan includes the features that make a form work: logic, calculations, file uploads, payments, integrations with Notion, Google Sheets and Zapier. Tally Pro, at $24 a month, sells almost nothing functional. It removes the "Made with Tally" badge, adds custom domains, and unlocks team features. Business, at $74, adds data retention controls and email verification.

In other words, the main thing a paying customer buys is the removal of Tally's growth loop. Every free form is filled in by people who are not yet Tally users, and each one sees the badge. The respondents of today's free forms are next month's signups. This is the same discovery that powered Carrd's one-person, $19-a-year business: when the product is visible in use, users are distribution, and you can price paid tiers almost entirely as branding and workflow.

The 2 percent math

Run the numbers and the model stops looking generous. Roughly 12,000 paying customers by the end of 2025 against over a million users is about a 2 percent conversion rate. That would sink a product with meaningful per-user costs. Forms are close to free to host, so Tally can carry 98 free users per customer the way a newsletter carries subscribers.

The team side is where the discipline shows. Two founders ran the product with no full-time employees for roughly three years; the first hire was a customer success manager, not an engineer, because a million free users generate support volume (around 35 tickets a day by late 2025) whether or not they pay. An experiment with two additional full-time engineers in 2023 was reversed within the year; the founders decided the product moved faster small. At $5 million ARR and 11 people (three of them support, three developers), revenue per team member is about $454,000, a figure most funded SaaS companies never reach. The company also got a distribution gift it did not plan: by early 2025, ChatGPT had become Tally's number one referral source, users asking an AI "what should I use for forms" and being handed the free option. Free-first products win recommendation engines for the same reason they win word of mouth: there is no reason not to suggest them.

Where the model breaks

Three honest caveats before you copy this. First, marginal cost must be near zero. Unlimited free video hosting or AI inference would have bankrupted them; unlimited forms did not. Second, the product must be seen by non-users when used. A bookkeeping tool has no badge to show anyone; a form, a landing page, or a scheduling link does. Third, you need the runway to survive the flat years. Martens and Minev launched in summer 2020 and did not reach $10,000 a month until February 2022. A 2 percent conversion rate only pays when the top of the funnel reaches hundreds of thousands, and that takes years or goes viral, usually years.

If any of those three fail, the boring alternative is better: charge from the first customer, as we argued in your first price is too low, and treat free tiers as trials, not marketing. If all three hold, Tally is the current best evidence for the free-first playbook, alongside TypingMind's paid-only counterexample. The decision rule fits in one line: give the product away only if the product itself, in use, recruits the next user. Otherwise the free tier is just a discount with extra steps.