Back to Blog

How to stop one company starting trial after trial

When the same company takes trial after trial, your conversion number is wrong. What email-domain and workspace rules catch, and where they break.

$24,480 a year went on trials for companies that had already had one. You don't know that yet, because it's the morning after your biggest webinar of the year and you're still pleased with yourself. The campaign cost $6,400 and brought in 160 new trials. That's $40 a trial, the best number you've ever posted in the team channel, and three people reacted with a rocket. You open the list to pick the ten companies you'll call first.

Every figure here is made up, so try it with yours. The third name on the list feels familiar, and you search your database for its domain. That 12-person agency has started 7 trials since March. Seven 14-day trials is 98 days of your product, and it has never paid you anything.

If you've ever typed same company multiple trials B2B into a search box, this is the morning that made you do it. You check the next domain, then the one after that. By 9:30 you've compared all 160 with your old trials:

  • 34 of them came from just 12 companies that had already had a trial and never bought.
  • 61 more used Gmail or Outlook addresses, so you can't say who they are.
  • That leaves 65 you can vouch for (160 - 34 - 61).

Your $40 trial was a $51 trial if every Gmail address is someone new ($6,400 ÷ 126, which is 160 - 34), and a $98 trial if you only count what you can vouch for ($6,400 ÷ 65). So you pull the whole year: 3,000 trials, 150 customers, 5%. That's the number in your board deck. It turns out 180 companies started 900 of those trials, five each, and none of them paid. Those 180 should count once each:

Real conversion=1503,000−900+180=1502,280=6.6%\text{Real conversion} = \frac{150}{3{,}000 - 900 + 180} = \frac{150}{2{,}280} = 6.6\%

You reported 5.0% all year. The real number is at least 6.6%, and you rewrote onboarding twice to fix one that was wrong.

The money is worse. Of those 900 trials, 720 were repeats (900 - 180). At $34 a trial for compute, AI features and onboarding email, that's $24,480 a year, or 1.2 months of runway if you spend $20,000 a month. A company running trials back to back needs only 26 a year to cover 364 days (26 × 14). On your 8-seat plan at $49 a seat, it should be paying you $392 a month.

At 10:05 a.m. a customer on that exact plan replies to a thread from Friday: "Still waiting on the export fix. We pay for this." Your one success manager hasn't reached it, because they've got 34 onboarding calls booked this week for those 12 companies, 17 hours at 30 minutes each.

Before the board call you've got to decide: cut the trial to 7 days, demand a work email or put a sales call in front of every trial. Each of those lands hardest on the other 2,100 trials, the only ones that have ever turned into customers.

Why the same company gets multiple trials

Your trial is attached to an account or a workspace, and both are free to create. A company has many people with many email addresses, so nothing in your app says the second workspace has anything to do with the first.

Plenty of this starts innocently. A marketing lead tries you in March, and a developer at the same company signs up in May without hearing about it. Then someone notices that a new workspace means a new 14 days, and an accident becomes a habit. Across enough companies that's trial farming, counted in your reports as growth.

Company startsa trial14 days offull accessTrial endsno purchaseNew workspacenew emailReport countsa new trial
Company startsa trial14 days offull accessTrial endsno purchaseNew workspacenew emailReport countsa new trial

The obvious fix is one trial per email domain. Build it, because it's cheap and it answers the innocent cases well. It just stops at the edge of your list of free email providers.

How to allow one trial per company

Decide what a company gets. Write it down, for example one trial per company every 12 months.

Find out how big this is. Run this against your own data, leaving out free email providers:

repeat_trials_by_domain.sql
select email_domain,
       count(*)        as trials,
       min(started_at) as first_trial
from trials
where email_domain not in (select domain from free_email_domains)
group by email_domain
having count(*) > 1
order by trials desc;

Every row is a company your reports counted more than once. From there, correct your conversion rate.

Send colleagues to the workspace that already exists. When someone with a verified address signs up at a domain that has a workspace, offer to let them ask to join it. That takes care of the developer who never heard about the March trial.

Record the trial against the domain, and keep the record. If deleting a workspace clears its history, deleting is a free reset. Keep who had a trial and when after the workspace is gone, allow one trial workspace per account, and run the rule when a trial starts, because a new workspace never goes through your signup form.

Give people a way to ask. A no with no way forward loses real buyers, and some second trials are good news.

A message you can use

Your company had a trial with us in March, so we can't start a new one automatically. Want another look? Reply here and we'll sort it out with you.

Where email-domain rules break

A domain rule only works on addresses that identify a company, and many addresses don't.

Free providers. Apply the rule to gmail.com and one person's trial uses up the allowance for everyone on Gmail, so you exclude every free provider, on a list you now maintain. After that, anyone using a personal address looks like a stranger. In the scenario, those are the 61 you couldn't place.

You could refuse them with a work-email-only rule, but look at the trade. At the corrected 6.6%, 61 trials are about 4 customers. On your $392 plan that's $18,816 a year turned away (4 × $392 × 12) to save at most $2,074 in trial costs (61 × $34). Plenty of founders, freelancers and small teams run their whole business from a Gmail address.

Shared domains. One university domain can hold dozens of labs that have never met, and a large company's domain can hold divisions with separate budgets. Your rule gives the first one a trial and refuses the deal you wanted most.

More than one domain. Regional sites and acquired brands look like new companies.

Agencies and teams. An agency has one domain and many clients, so a strict rule can refuse someone who brings you new customers, and a loose one funds the agency in your opening scene. Teams blur it further: an invited colleague with a personal address can start their own workspace, and your per-workspace limits start over with it.

Other basic checks for repeat company trials

Basic checkWhat it catchesWhat it missesWho it wrongly blocksWhat it costs to keep working
Email verificationTypos, and claims to a domain with no mailbox thereA colleague's real inboxPeople whose code lands in spamDeliverability work
Unique emailsThe same address twiceA coworker's addressAlmost nobodyVery little
Disposable-domain blocksKnown throwaway servicesOrdinary mailPeople using privacy forwardingA list updated forever
Gmail dots and plus addressingOne inbox written several waysA second real mailboxPeople who sort their mail with plus addressesUpkeep as mail services change
CAPTCHASimple scriptsReal employeesPeople using screen readersFees and lost signups
Phone requirementA casual second tryAnyone else on the teamPeople with only a desk lineA charge per text
Card requirementCasual repeat trialsA second company cardStaff who need approval to use oneFewer trial starts
One per IP addressAn office on one connectionRemote staffUnrelated companies in one coworking spaceExceptions and tickets
VPN blocksKnown VPN rangesEverything off your listEmployees on a company VPNAnother list
Rate limitsBursts from one account7 trials over 8 monthsBusy real teamsTuning per route
Usage capsCost per trialThe number of trialsEvaluators testing at real volumeDeciding what stops
Shorter trialsSome cost per trialRestarts, which come soonerBuyers waiting on approvalFewer people reach paying
Manual reviewWhat a person recognizesAddresses nobody can placeWhoever looks odd to a tired reviewerHours every month

Cut the trial from 14 days to 7 and a company running back to back needs 52 trials a year instead of 26, while the buyer waiting on legal loses half their time.

Plot the company-level rules and a pattern shows up.

Illustrative rule strictness vs false blocksStrict and harshLoose and harshLoose and gentleStrict and gentleLooserStricterFew false blocksMany false blocksEmail verificationUnique emailsDisposable blockOne per IPCard requiredOne trial per domainWork email only
Illustrative rule strictness vs falseblocksStrict and harshLoose andharshLoose andgentleStrict andgentleLooserStricterFew falseblocksMany falseblocksEmail verificationUnique emailsDisposable blockOne per IPCard requiredOne trial per domainWork email only

The bottom-right corner is where you want to be, and no basic rule sits there. Meanwhile the free-provider list, the shared-domain exceptions and the second-trial inbox are all yours to keep running. Every rule on this page looks at an address, when what you need to know is whether these people have already had this trial.

When the email domain can't tell you

We built AbuseZero to evaluate what a domain list can't, whether this claim looks like a repeat, and we think it's very good at it. The 61 addresses you couldn't place are where a list gives up, and where 140,000+ signals go to work, examined by multi-layer AI risk analysis that detects linked claims and flags abuse that single checks miss. Those signals are read together, never one at a time. There's also a proprietary algorithm, with AI that checks the claim again.

A new email and a new workspace got past the domain rule, while AbuseZero recognizes repeat devices across new accounts, even when email addresses keep changing. A second card got past your card requirement, and AbuseZero detects repeated payment fingerprints across users, even when new accounts and emails hide the connection. Your one-per-IP limit saw an address, and AbuseZero sees beyond it with VPN, hosting, network reputation, Tor and proxy intelligence. Creating a fresh account isn't enough to erase previous activity.

You don't have to build your own abuse prevention system. One API call before the trial starts returns a clear allow or deny, under rules you control, and eligibility can be once ever or once every set number of days. Observe mode shows you what would be denied before anyone is blocked. The quickstart walks you through your first benefit.

Frequently asked questions

How do I stop the same company starting multiple trials?

Decide what one company gets, send colleagues to the existing workspace, and keep a record of each trial by domain, even after a workspace is deleted. Expect that to stop at free and shared domains.

Should a B2B free trial only accept work emails?

It depends who buys from you. If founders, freelancers and small teams are your customers, you'll turn away people who'd have paid.

Is a second trial from the same company always abuse?

No. A new team, a major release or a year's gap can all be a real second look, so give those people a way to ask.

Keep your trial for companies that are new to you

Keep reading

Copying your prompt…

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