The pricing model that built Notion, Figma, and Slack — and the one that quietly bankrupts most solo SaaS founders who copy it without the volume to make the math work.
Research-based overview. This article synthesizes public pricing pages, investor decks, and OpenView SaaS Benchmarks data on freemium conversion rates from Notion, Figma, Slack, Vercel, Supabase, Linear, Calendly, and Typeform. How we research.
Freemium is the pricing model that powers a generation of category-defining SaaS — Notion, Figma, Slack, Spotify, Dropbox, Canva, Zoom, Calendly. It is also the model that quietly kills more solo SaaS startups than any other pricing decision, because the math that works for a company with a million weekly signups falls apart at the volumes a solo founder actually reaches. The pattern is deceptively simple: build a useful product, give a meaningful slice away for free forever, charge users who outgrow the free tier. The companies that pull this off build moats; the companies that copy the structure without the underlying economics burn through runway supporting free users they cannot afford.
Look at the companies that scaled freemium to billion-dollar outcomes:
Four properties show up in every one of these examples. First, the products have a built-in viral or collaboration mechanic — you do not use Slack alone, you do not share a Figma file with yourself. Second, the marginal cost of supporting a free user is low — serving a Notion doc or a Calendly link costs fractions of a cent. Third, the addressable market is enormous — hundreds of millions of potential users. Fourth, the upgrade trigger is concrete and unavoidable at scale. Strip any one of those out and the model breaks.
Public OpenView SaaS Benchmarks data and disclosures from companies that have run freemium for years put typical free-to-paid conversion in a narrow band: roughly one to five percent of free users convert to a paid plan over the lifetime of their account. Best-in-class operators with strong upgrade triggers and a paid-feature surface that users feel the absence of land in the three-to-five percent range. The median is closer to two percent. Below one percent is common for products with weak upgrade triggers or commodity-feature free tiers.
Run the numbers and the implication is brutal:
| Conversion rate | Free users needed for 1,000 paying customers | Free users needed for $10K MRR at $20/mo |
|---|---|---|
| 1% (typical low) | 100,000 | 50,000 |
| 2% (median) | 50,000 | 25,000 |
| 3% (good) | 33,000 | 17,000 |
| 5% (best-in-class) | 20,000 | 10,000 |
Solo SaaS founders rarely reach the volumes the left column requires. A successful indie launch on Product Hunt produces a few thousand signups in a week. A consistent SEO and content strategy produces hundreds to low thousands of free users per month after a year of work. The free tier is doing free distribution work in the background, but the founder is also supporting tens of thousands of non-paying users who still send support tickets, hit the database, and occasionally show up in the comments. Freemium works at scale because a small conversion percentage multiplied by enormous user counts produces a real business; below the volume threshold, the same percentage produces a hobby.
Not all freemium is the same shape. There are four common variants and the choice between them is one of the most important early-stage pricing decisions.
The free tier includes a subset of features; the paid tier unlocks more. Notion uses this model — the free plan includes unlimited blocks for individuals but locks version history, admin controls, and team features. Figma does the same — basic editing free; shared libraries, design systems, and dev mode paid. The strength is that users discover paid features by trying to use them; the weakness is that founders who gate too aggressively kill the viral loop.
The free tier includes a usage allotment — events, users, storage, requests — and paid plans buy more. Vercel offers a generous free tier capped at specific bandwidth and build-minute allotments. Supabase offers free Postgres up to 500MB with a monthly active user cap. Stripe Tax is free under a transaction threshold. This is the cleanest model when marginal cost is real — you set the cap where supporting the user starts to cost money.
The free tier supports a small number of users or seats; teams above that size must pay. Slack’s historical free tier worked this way. Linear’s free tier supports small teams; larger teams pay. Seat-limited works well for inherently collaborative tools because growth in seats correlates with growth in value — once a team has integrated the tool into daily work, the upgrade is rarely contested.
The free tier is fully functional but adds branding, watermarks, or attribution; paid removes them. Calendly historically added a “Powered by Calendly” footer to free scheduling pages. Typeform showed Typeform branding on free forms. This works when users have a status reason to remove the branding — agencies, consultants, customer-facing pages.
Every freemium model needs an upgrade trigger — the specific moment that converts a free user into a paying one. The most reliable trigger is the “limit hit” moment: the user tries to do something the free tier does not allow. The fourth Figma editor invites the fifth. The Notion team needs version history to recover a deleted page. The Slack team tries to search a message from four months ago. The Vercel project crosses the bandwidth cap. The user is already deep in the product and the upgrade is framed as “keep doing what I’m doing, with the cap removed” rather than “decide whether this product is worth paying for in the abstract.”
Triggers that fail: feature flags hidden behind menus the user never sees; usage caps so generous they are never hit; free tiers functionally identical to paid because the founder was afraid of looking stingy. The free tier needs to be useful enough that users adopt it and rough enough at the edges that they feel the upgrade when their usage matures.
The conditions under which freemium is the right pricing model are specific and not in the founder’s control:
When all five conditions are present, freemium is the dominant strategy. When fewer than three are present, freemium is a trap.
The mirror-image conditions are where founders walk into the trap:
The vast majority of solo SaaS products have at least two of these conditions present, often three. The honest read is that freemium is the right model for maybe ten to twenty percent of solo SaaS launches and the wrong model for the rest.
The closest alternative is the time-limited free trial: seven, fourteen, or thirty days of full-featured access, after which the user is required to pay or be downgraded. Freemium uses feature or usage pressure — the user converts because they hit a limit. The conversion window is indefinite. A free trial uses time pressure — the user converts because the trial is ending. Trials typically convert at fifteen to thirty percent — an order of magnitude higher than freemium — but absolute volume is lower because users who would have stayed on a free tier never sign up for a trial in the first place.
For most solo SaaS, the free trial produces better economics. A seven- to fourteen-day trial with full feature access, paywalled at the end, with a paid tier in the $19-99/month range, is the closest thing to a default model for B2B solo SaaS. The trial filters for serious users; the paywall produces a clean conversion event; the support load is bounded.
A hybrid pattern that has gained traction is the reverse trial. The user signs up and gets fourteen days of Pro-tier features for free. At the end, they are auto-downgraded to a permanent free tier rather than paywalled. During the trial, the user gets to feel Pro — advanced features, higher limits, polish. After the trial, the user is on the free tier and feels the absence of those features acutely. The upgrade decision now has both a recency anchor and the long conversion window of freemium. Linear, Vercel, and several newer infra SaaS use variants of this model. It is the right answer for solo founders who want a freemium-shaped funnel without committing to indefinite free support.
The honest default for solo SaaS founders evaluating freemium is to skip it. Start with a seven-to-fourteen-day free trial of the full paid product, a single paid tier in the $19-99/month range, and no permanent free option. The reasons: volume is the bottleneck and freemium needs volume that solo founders take years to reach; free users still consume support time; infrastructure costs scale with users, not revenue; and the price-discovery problem is solved faster with a trial than with a free tier. Examples of solo and small-team SaaS that succeed without freemium are easy to find — paid-only B2B SaaS at $29-199/month with a fourteen-day free trial, productized services with one-time pricing, vertical SaaS for trades and professional services charging $99-499/month from day one.
For deeper coverage see the complete guide to SaaS pricing and the solo founder pricing playbook. The companion piece why most SaaS pricing is wrong covers the mistakes that bleed into both freemium and paid-only models, what is product-led growth covers the distribution strategy that often accompanies freemium, the true cost of running an AI SaaS covers the cost-side argument against freemium, and what is SaaS positioning covers the strategic context.
Freemium is a pricing model where a basic version is free forever and upgrades happen because users hit feature or usage limits. It works when the product has network effects, viral loops, low marginal cost per user, a large addressable market, and concrete upgrade triggers. It fails when any of those conditions is absent, which is the situation most solo SaaS founders are in.
For the founders who do have the right conditions, freemium is one of the best distribution moats in software. For everyone else, a fourteen-day free trial paired with a single paid tier produces better economics, faster validation, and a lower support burden. The right default is paid-first; freemium is a graduation, not a starting point.
The stack, prompts, pricing, and mistakes to avoid — for solo founders building with AI.