There are four hidden costs every founder pays when they pick the free tier of a tool they'll grow out of. None of them show up on the pricing page. All of them show up in month four when you're trying to leave.
This isn't an argument against free tiers. We use a lot of them. Vercel Free, Supabase Free, Cal.com Free, PostHog Free are genuinely good. The problem is treating "free" as the deciding factor instead of as one input. Here's what you're actually paying.
Cost 1: switching cost paid forward
The cheapest tool is the one you don't have to leave. The most expensive tool is the one you joined because it was free, then outgrew, then had to migrate while running a business.
The pattern repeats. Mailchimp Free was great until they cut the free tier from 2,500 to 250 contacts in 2025. PlanetScale Hobby was great until it disappeared in March 2024. Slack's free tier got compressed twice in three years. In each case, the founders who joined because it was free paid the migration cost when free changed.
The honest math: if you're under 1,000 customers and just starting, pick the tool you'd want to be on at 10,000 customers, not the free one you'd outgrow at 250. The migration tax is paid out of your time, not your subscription line.
Pick for the version of you that's a year ahead.
Free is fine if you'd still pick this tool at scale. Free is expensive if you're choosing it because you can't afford the better option yet.
Cost 2: feature compression
Free tiers compete for the upgrade. The best way to win that competition is to gate the features you'll need most as soon as you grow up.
Look at any modern free tier and you'll see the pattern. Beehiiv Free publishes, but custom domains, custom fonts, and basic analytics are paid. Kit Free sends, but automations are paid. Notion Free works, but the AI features that turn it into a real workspace tool are gated. Free tiers don't compete on free; they compete on how soon you'll need the paid feature.
The cost is invisible because it's an opportunity cost. You don't ship the automation, the custom domain, the analytics view, because you're on free and it's not available. Six months later you wonder why your stack feels thin.
Cost 3: data and ownership lock
Free tiers almost always have weaker export. The export is gated by tier, by API access, by retention policy.
The clearest case is email. Free email tools let you import a CSV and send. Try to leave with three years of subscriber engagement data, sent campaign history, automation flow definitions, and you'll find the export is structurally lossy. CRMs follow the same pattern. CMSes follow it. Note-taking tools follow it.
This is the shape of cost 3: you can put your data in faster than you can take it out. The longer you stay on the free tier, the more data accumulates, the heavier the export problem gets. Founders who think they can swap tools cheaply because they're on free are reading the pricing page, not the terms page.
Cost 4: vendor pinning
Free tier users don't drive the roadmap. Paid users do. So when the vendor decides between two product directions, your free-tier preferences get the lighter weight.
This shows up two ways. First, the features you want get deprioritised in favour of features paying customers want. Second, when the vendor needs to monetise more aggressively (most do, eventually), the free tier compresses while the paid tiers don't. Docker ran this play in 2025. Twitter/X ran it on the API tier in 2023. Heroku ran it in 2022.
The pattern is consistent enough to be a default expectation. If a tool's main user base is on free, the squeeze is coming. The only question is when.
The 'free is fine' criteria
None of this means free is bad. We use plenty of free tiers and recommend them. The criteria for when free is fine:
- You'd still pick this tool at scale. Free is just the starting tier of a tool you'd happily pay for at growth.
- The vendor's main revenue is enterprise, not your tier. Tools selling to big customers can keep free tiers stable for years (PostHog, Cal.com, Cloudflare). Tools selling to your tier squeeze it.
- Your data is exportable in a structured form. Test the export before you depend on it.
- The feature you need most is in the free tier. If you're already wishing for a paid feature on day one, you're on the wrong tier.
The audit pass
Pull up your stack. For every free tier you're on, ask:
- Would I still pick this tool if it cost the paid tier price?
- Has the free tier compressed in the last 18 months?
- Is my data exportable cleanly today?
- Is the feature I'm actively missing on the paid tier?
If the answers are yes / no / yes / no, free is genuinely free for you right now. If you're answering the other way on any of them, you're paying a hidden cost the pricing page didn't quote.
If you want a quick scan across your whole stack, paste your tools into Stack Audit and we'll flag the patterns.