Why users keep restarting your free trial
Some users restarting your free trial are abusing it. Others are telling you the trial or the price is wrong. How to tell which, and what to do about each.
The same twelve rows keep pulling you away from the chart you were pasting into your investor update. Trials started: 800 in August, 1,000 in September, 1,200 in October. That's up 50% in two months, the best thing you've had to show anyone all year. You only opened the October list to find a customer worth quoting.
One workspace is called reports-test-12. You search for reports-test and get twelve results, one a week since early August. Each ran your full 7-day trial, each exported the same report on its last day and none entered a card. Someone has had a whole quarter of your $49 plan (3 × $49 = $147) without paying you a cent, using twelve different email addresses.
Your analytics tool has a chart for trials started. It has nothing for users restarting free trial periods, because every restart looks like a brand new customer. So you go through the list by hand. This company is made up, but by 1:15 a.m. its numbers look like this:
| Month | Trials started | From 125 repeat users | From everyone else |
|---|---|---|---|
| August | 800 | 100 | 700 |
| September | 1,000 | 300 | 700 |
| October | 1,200 | 500 | 700 |
| Total | 3,000 | 900 | 2,100 |
Everyone else is a flat line at 700 a month. All 400 extra trials between August and October came from those 125 people, so the growth you were about to send your investors doesn't exist.
The conversion rate goes next. Your dashboard says 210 paying customers from 3,000 trials, which is 7%. None of the 125 ever paid, so the figure for everyone else is 210 ÷ 2,100 = 10%. You've spent six weeks and $8,000 on a contractor rewriting onboarding, to fix a 7% that was never there.
Each trial costs you about $9 in AI usage, compute and email:
900 repeat trials x $9 = $8,100 this quarter
October alone: 500 x $9 = $4,500
At October's pace: $4,500 x 12 = $54,000 a yearThat same October, 70 new customers at $49 brought in $3,430, so the people who'll never pay you cost $1,070 more than the people who just started to. If your company spends $30,000 a month, $54,000 is 1.8 months of runway. Those repeat trials also sit in the same report queue as your customers, which is why one of them wrote at 8:37 a.m. last Tuesday asking why a 3-minute report took 40. You told them it was growth.
At 1:20 a.m. you delete the chart and look at your options. You could cut the trial to 3 days, demand a card or pull the trial altogether, and each of those lands on the 700 people a month who are the actual business. The update goes out at 9. Whoever runs reports-test used your product every week for three months, which might make them the most motivated user you have, and nobody has ever asked them why they won't pay.
Why users keep restarting a free trial
A trial says "once", and the only thing enforcing it is an account, which costs nothing to make again. A new account arrives with no history, so the trial starts. The reasons people do it vary, and they point at very different fixes.
| Who's restarting | How it tends to look | What it means | What to change |
|---|---|---|---|
| Ran out of time | One restart soon after the first | The trial is shorter than your setup | Fix setup, or extend once |
| Needs you now and then | A few restarts a year, around the same job | Your price assumes monthly use | Add a pass or a smaller plan |
| Short on budget | Restarts, and says price is why | Your cheapest plan is out of reach | A smaller plan, or one offer |
| Former customer | Back after months with a new project | Nothing's wrong | Let them back in |
| Colleague | A different person, same company | The product is spreading inside a team | Join them to one trial |
| Repeat claimer | Back-to-back trials, no reply to anything | The trial is the product | Stop the repeat claim |
Only the last row is trial farming. The second row explains itself once you do the arithmetic. Someone who needs you for one week a quarter would pay $49 × 12 = $588 a year for 4 weeks of use, so of course they restart. Sell them a $29 one-month pass and four passes a year is $116 you currently get none of.
So a blanket rule of "one trial, ever" fails in both directions. It turns away the former customer and the person with a seasonal job, and it only stops the repeat claimer if you can tell they're the same person, which is the hard part.
How to work out which kind you have
Find the repeaters. Go through past trials and list the people who've started more than once.
Ask them. Send one short email from a real person, with no accusation in it: "You've tried us a few times. What would have made the first trial enough?"
Look at the last trial. If they never finished setup, your trial is the thing to fix.
Redo your conversion rate with repeat users taken out, so your next decision rests on people instead of accounts.
Write the rule down, such as one trial per person per year plus one extension on request, so you can say no kindly when you need to.
The basic checks for repeat trials, and how they break
The last row of that first table needs a way of saying no. These are the basics most teams try, and each one does catch something.
Checks on the email address
Email verification proves someone can open an inbox. It catches typos and made-up addresses, misses anyone with a second inbox, sometimes stops a real person whose code landed in spam, and needs ongoing deliverability work. Unique emails stop the same address twice and cost almost nothing, but they miss every different address, and reports-test used twelve.
Email has quirks too. Google's help pages say dots don't matter in a personal Gmail address, and that mail to janedoe+anything@gmail.com arrives at janedoe@gmail.com. Disposable-domain blocks catch known throwaway services, miss new domains and ordinary mailboxes, turn away people who use privacy forwarding, and need someone updating the list forever.
Checks that add a step
CAPTCHA slows simple scripts. A person starting a trial by hand passes it the seventh time as easily as the first, while people using screen readers struggle and you lose some signups. A phone requirement stops casual second accounts, misses anyone with another number, shuts out people without a reliable mobile and bills you for every text. Requiring a card stops the casual restart too. It misses anyone with another card, turns away people who want to look first, and fills your inbox with "I forgot to cancel" refunds.
Limits on networks and usage
One trial per IP address catches repeats from one connection, misses every other network and lands squarely on offices, campuses and households, so exceptions become tickets. VPN blocks cover the ranges on your list, hit privacy-minded people and staff on company VPNs, and give you one more list to maintain. Rate limits catch bursts, but someone starting one trial a week never comes close, and tight ones slow real users in a busy hour. Usage caps limit what one trial can cost you, but each new trial brings a fresh cap, and a real buyer hits it at the worst moment.
Changing the trial itself
Shortening the trial is the most tempting fix here, and it hurts the wrong people most. Going from 7 days to 3 cuts your cost per trial, but 84 days of free use now takes 28 trials instead of 12, and the people who ran out of time lose 4 more days. Manual review catches what a person can spot. At one minute each, 1,200 trials is 20 hours a month, it stops on weekends, and a tired reviewer guesses.
Now turn it all on and count the honest side. Say the extra steps cost you 10% of those 700 people, as an illustration. That's 70 fewer trials a month and, at 10% conversion, 7 fewer customers. After a year you're short 84 customers and $4,116 a month (84 × $49), and the new-email problem is still there.
Deciding who gets the next trial
Every check above looks at one thing, and you've just watched each one run out. In AbuseZero, creating a fresh account isn't enough to erase previous activity. It recognizes repeat devices across new accounts, even when email addresses keep changing, and detects repeated payment fingerprints across users, even when new accounts and emails hide the connection. Where a domain list runs out, it checks mailbox aliases, disposable domains and mail routing. Where an IP limit runs out, it sees beyond the address with VPN, hosting, network reputation, Tor and proxy intelligence. AbuseZero's multi-layer AI risk analysis examines 140,000+ signals, detects linked claims and flags abuse that single checks miss, and those signals are read together, never one at a time. There's also a proprietary algorithm, with AI that checks the claim again. You don't have to build your own abuse prevention system.
You still decide who's welcome. One API call before the trial starts returns a clear allow or deny, under rules you control. Eligibility can be once ever or once every set number of days, so a former customer can come back after a year. Observe mode shows you what would be denied before anyone is blocked, and the quickstart protects your first benefit. Fixing a trial that's too short or a price that doesn't fit stays with you.
Frequently asked questions
Why do users keep restarting a free trial?
Usually the trial ended before they got what they came for, or the price doesn't fit how often they need you. For a smaller group, the trial itself is all they want.
Does a shorter trial stop people restarting it?
It lowers the cost of each trial, but anyone determined just needs more of them, and the people who needed more time get hurt most.
How can I limit repeat trials without hurting real users?
Keep signup open and decide when the trial is claimed. AbuseZero evaluates each claim, and setting eligibility per benefit leaves room for people you want back.
Keep your trial for the people it can win
Start with Builder
Pay as you go. Evaluate your first trial claim today.
See pricing
Builder, Pro, Scale and Enterprise.
Talk to us about Enterprise
High volumes and custom terms.