Back to Blog

How to stop repeat free trials started on the web for a mobile app

Web checkouts sit outside the app stores' trial rules, so a new email can mean a new free trial. What that costs, which fixes break and what to change.

People are taking a second free trial of your mobile app on your website with a new email, and the app store can't stop them, because the store never saw the first one.

That's web to app free trial abuse, and it hurts most when one trial can give someone everything they came for. Here's a scenario, with numbers you should replace with your own.

You run a driving theory test app. (In the US, think permit test.) It has the full question bank, the hazard clips and 20 mock exams. Your best ad says "Pass your theory test in 7 days. Free." It sends people to a web quiz, then a web checkout that starts a 7-day free trial with only an email, and a link opens the app already signed in. You built it this way so an ad could go straight to a trial with no install in the middle. After the trial it's $29 a month, and the average customer stays 2 months, because nobody needs you once they've passed. That's $58 a customer.

It's 10:47 p.m. on the last Sunday in October, and you're reading new reviews, because trial numbers have been great and you wanted to enjoy that. The top one has five stars: "Passed first time and paid nothing. When your free week ends, go back to the website and use another email."

You put October next to May. The $5 ad cost per trial and the $0.60 to serve one are assumptions.

May
2,400 trials x $5        = $12,000 in ads
2,400 trials x $0.60     = $1,440 to serve them
336 paying (14%) x $58   = $19,488 in revenue
Left over                = $6,048

October
2,800 trials x $5        = $14,000 in ads
2,800 trials x $0.60     = $1,680 to serve them
168 paying (6%) x $58    = $9,744 in revenue
Left over                = -$5,936

Trial starts are up 17%, which is why you raised the budget. Paying customers have halved. That's an $11,984 swing every month ($6,048 + $5,936).

Ad cost per paying customer, illustrative020406080100MayJunJulAugSepOctUS dollars
Ad cost per paying customer,illustrative050100MayJunJulAugSepOctUS dollars

The flat line is the $58 a customer brings in. The other line is what you spend on ads to get one: $12,000 ÷ 336 = $35.71 in May, $14,000 ÷ 168 = $83.33 in October. Since September, every new customer has cost more than they'll ever pay you.

Now say 1,100 of October's trials went to people who'd already had one. At $5.60 each, that's $6,160 to give someone a week they've already had. Your own retargeting brought most of them back, since that campaign's audience is everyone who started a trial and didn't pay. And if you've told your ad platform to find more trial starts, it just got 1,100 more examples of what a good one looks like.

Then the paying customers notice. On Monday, someone forwards you that review and asks for their $29 back, because a friend got through the same mock exams for free. It's the ninth request this month, so 9 × $29 = $261. With $70,000 in the bank, losing $5,936 a month gives you about 11.8 months ($70,000 ÷ $5,936).

The decision waiting for you is the worst part. That ad brings in nearly every customer you have. Turn it off and the loss goes away, along with the $9,744. Keep it and you go on paying to hand out the product. Most learners need you for about three weeks, and 3 emails × 7 days = 21 days. The ad is honest. Someone really can pass for free, and if 7 days isn't enough, 14 costs one more email.

Why app store trial rules don't reach your website

Apple's help pages say each person can redeem one introductory offer per subscription group. Google Play lets you limit a new-customer offer to people who've never had that subscription, or never had any subscription in your app. Those rules are useful. They apply to purchases made through the store.

Your web checkout isn't one of those. The trial starts in your database, under an account you created, and the app only hears about it when the link signs someone in. The store has no record of that trial, so its rule has nothing to count. A new email gives you a new account, and your app opens it like any other.

In the appNoYesOn your websiteNew emailSame emailLearner wants a free weekWhere do they start?App store checkoutStore: had this offer?7 days freeFull priceYour web checkoutHave you seen this email?
In the appNoYesOn your websiteNew emailSame emailLearner wants a free weekWhere do they start?App store checkoutStore: had this offer?7 days freeFull priceYour web checkoutHave you seen this email?

The obvious fix is to move every trial back inside the app. The store's rule would then cover everyone, and that's a real option. You'd also give up the reason the web funnel exists, and you'd put an install between your best ad and the trial. The other obvious fix, one trial per email, is the rule you already have and the one a new email gets past.

The basic fixes, and how each one breaks

Email comes first, because it's the only thing your checkout asks for. Verification proves someone can open an inbox. It says nothing about whether that inbox is new to you. Gmail ignores dots, and plus-tagged mail reaches the same inbox, so one mailbox can look like many different addresses. And a second real address takes a minute to create.

Basic fixWhat it catchesWhat it missesWho it wrongly blocksWhat it costs to keep working
Email verificationTypos and dead addressesAny real second inboxLearners whose code lands in spamDeliverability work
Unique emailsThe same address twiceEvery different addressAlmost nobodyVery little
Disposable-domain blocksKnown throwaway servicesNew domains and ordinary mailboxesPeople using privacy forwardingA list someone updates forever
Gmail dots and plus addressingOne Gmail inbox written several waysA second real mailboxFamily members sharing one inboxUpkeep as mail services change
CAPTCHA on the checkoutSimple scriptsA learner filling in the form by handPeople using screen readersFees and lost trial starts
Phone requirementCasual second accountsAnyone with another numberLearners without their own reliable numberA charge for every text
One per IP addressRepeat trials on one connectionMobile data and any other networkColleges and householdsExceptions and tickets
VPN blocksKnown VPN rangesEverything off your listPrivacy-minded learnersAnother list to maintain
Rate limitsBursts from one sourceOne extra trial a weekA class signing up togetherTuning per route
Caps on mock examsOne account doing everythingEach new account's fresh capAnyone cramming the night beforeDeciding where your product stops

Requiring a card at the web checkout is what most teams try next, and it does stop casual repeats. It misses anyone with another card, and in this product it lands on the wrong people, because plenty of learner drivers are 17 and don't have a card of their own. Say trial starts fall by 40%, to 1,680. The same $14,000 of ads now costs $8.33 per trial instead of $5.

Shortening the trial from 7 days to 3 cuts what each account gets. It also means an honest learner gets 3 days to see their mock scores move, and getting 9 free days now takes 3 emails. Manual review doesn't rescue any of this: 2,800 trials at one minute each is about 46.7 hours a month.

What to change in your web funnel

None of those checks is enough alone, but a few changes to the funnel will stop you paying for the easiest repeats.

Put every trial in one table, web and in-app, with where it started and when it ended. Your table can say when an account comes back. A new email makes a new account, so it can't say when a learner does. Auditing what you already have shows you how big this is.

Write the limit down. One free week per learner is a start. Decide what happens when someone fails and rebooks, because that's a real customer with a real need.

Start the trial on your server, after your checks, and have the app unlock only what the server says.

Take past trial starters out of retargeting wherever you have their email, and send "welcome back" ads to the paid plan. Each click you don't buy saves up to $5.

Optimize ads for paying customers, not trial starts, if your ad platform allows it. Otherwise your repeat trials count as its best results.

Hold something back for paying customers. A trial with 3 of the 20 mock exams proves the app works without finishing the job.

Decide what a no looks like. Show the paid plan and a way to reach you, and say it kindly.

All of that helps, and you still own the domain list, the review queue and the reply to the 17-year-old whose card got refused.

Get an answer before the free week starts

You don't have to build your own abuse prevention system. AbuseZero is one API call from your web checkout, made before the free week begins, and it returns a clear allow or deny under rules you control. A new email got past most of that table, and AbuseZero recognizes repeat devices across new accounts, even when email addresses keep changing. A card requirement doesn't compare cards across accounts, and AbuseZero detects repeated payment fingerprints across users, even when new accounts and emails hide the connection. Nobody on your team has to look after a domain list, because it checks mailbox aliases and disposable domains for you, and it sees beyond an IP address with VPN, hosting, network reputation, Tor and proxy intelligence. You're getting a proprietary algorithm, with AI that checks the claim again. AbuseZero flags abuse that single checks miss and detects linked claims, with multi-layer AI risk analysis examining 140,000+ signals. Creating a fresh account isn't enough to erase previous activity.

Set eligibility to match your limit, and import past trials as claim history so they count from the first evaluation. Observe mode shows you what would be denied before any learner is turned away, and the quickstart gets your first benefit protected.

Frequently asked questions

Don't Apple and Google already stop repeat free trials?

For trials bought through their stores, they have eligibility rules. A trial that starts on your own website is your checkout and your record, so those rules don't apply to it.

Should I move all my trials back into the app?

If the web funnel isn't earning its keep, yes, and you'll get the store's rule with it. If your best ads depend on it, keep it and put your own limit at the web checkout.

Is every second trial abuse?

No. Someone who fails and rebooks months later, or a sibling on the same family connection, isn't what you're after. Keep a way for people to ask, and read what to do when a customer asks for a second free trial.

Keep your free week for new learners

Keep reading

Copying your prompt…

Copied the to your clipboard. Give it to your coding agent & it will handle the rest.