Why We Dropped Our SaaS Free Plan
October 25, 2020
We ran a free plan on PingPing for a few months back in 2020. It gave us the biggest signup week the product had ever had. Then we turned it off. Not because free plans are a dumb idea, but because the numbers sitting underneath those signups told a completely different story than the signup chart did. If you’re a SaaS founder weighing the same thing, free plan versus a time-boxed trial, here’s what actually happened and the call we’d make again today.
What our free plan actually offered
It wasn’t a crippled demo. Free accounts ran the real product: 30-second checks, status pages, every notification channel we had. The only ceiling was two monitored sites instead of five.
The bet was the textbook one. Take the barrier to entry down to nothing, let people watch the product do its job, then convert the ones who grow past two sites. Launch week backed it up. Best growth we’d ever posted, and it wasn’t close. We stayed thrilled for about three weeks.
Why our free plan didn’t convert
A free plan only converts if your typical user reliably outgrows it. Ours didn’t. Most people who signed up monitored a single site, maybe two, and the cap we’d built the whole funnel around was a wall they’d never walk into. They got everything they needed at zero, so they paid zero.
So we did the obvious next thing and made the free plan worse. Free accounts dropped to 5-minute checks. The status page went away. A thinner free tier would push people toward paying, or so the reasoning went.
It pushed nobody anywhere. The people who were happy on the original free tier were mostly just as happy on the slightly worse one, and they were the exact people who were never going to pay us a cent. All we’d managed to do was make the product meaner for the people who liked it. A pricing cap only works when it sits on a line your real customers actually step across. For a one-site monitor, that line barely exists.
The abuse you don’t budget for
Then the signup graph started looking strange. One domain pattern kept surfacing in new accounts, again and again. We tracked it to a code-collaboration platform where people were spinning up demo subdomains and bolting them onto free PingPing accounts automatically. A good chunk of it was plainly scripted. The “sites” were hello-world pages with no earthly reason to be monitored by anyone.
Free monitoring, it turns out, is catnip for people who will never be customers. We’d quietly become one step in somebody else’s automation pipeline. So we blocked the domain. They moved to bit.ly redirects pointing at the same demo URLs. We blocked the bit.ly hostnames. They kept coming, so we blocked the upstream IP range outright.
None of that registers as a problem on a growth dashboard. It registers in the infrastructure bill, and in the afternoons you burn chasing redirect chains instead of shipping anything real. When you’re a small team, that tax is basically the whole story.
Free plans aren’t “almost free”
There’s a tidy little argument for free tiers: each extra user costs us next to nothing, so what’s the harm? It was wrong for us in two separate ways.
Monitoring isn’t cheap at the margin. Every URL on a free account is real traffic leaving real nodes around the world every 30 seconds, indefinitely. Multiply that by a few thousand free accounts and the bill stops being a rounding error and starts being a line item you argue about.
The bigger miss is that the cost which actually scales with users usually isn’t compute at all. It’s the human work. The support question. The abuse review. The false positive someone wants explained at 11pm. The hour you lose staring at a funnel that won’t make sense. None of it gets cheaper just because the account behind it pays nothing.
Why a 14-day trial worked better
We shut the free plan off at the end of September and brought back a 14-day trial: full product, no card up front. Conversion ticked up almost right away.
A deadline forces a decision. The whole “I’ll get around to upgrading” cohort can’t really exist on a trial, because there’s a date, and after the date there’s no free account to coast on. The trial closed the abuse hole on its own, too. Nothing automated bothers farming a tool that switches itself off in two weeks. And there’s a signal buried in it. “Pay if you decide to keep it” reads like a company that plans to still be here next year, and plenty of founders trust that more than “free forever,” having watched enough free-forever tools wink out the moment the runway ran dry.
Was 14 days the right window? For us, yes. The people who converted mostly did it early, well inside the two weeks, so a longer trial would have added procrastination, not evaluation.
What we’d tell another SaaS founder
Map the wall before you build the plan. Write down, in one plain sentence, the thing a happy customer eventually runs into and pays to get past. If you can’t write that sentence with a straight face, a free plan won’t convert. It’ll just grow you a friendly audience that never reaches for a card.
Assume some slice of your free signups aren’t customers and never will be, and work out how you’ll catch the abuse before you launch rather than during the cleanup. Keep an eye on support load per signup too, not just the signup count. Growth that drags an equal amount of support behind it isn’t free growth. It’s an invoice you haven’t added up yet.
And 14 days isn’t sacred. Some products want seven, some want thirty. Pick the window that matches how long an honest look at your product really takes, then leave it alone long enough to read actual data off it.
We don’t regret running the experiment. We regret how long we left it switched on. If you’re stuck on the same question, give yourself a fixed window, watch the numbers underneath the signups rather than the signups themselves, and make the call with your eyes open.
Curious where we landed? You can start a 14-day PingPing trial with every feature unlocked and no credit card. Takes about a minute.