How to protect a Product Hunt launch from fake signups
Fake signups can eat your runway on launch night while you sleep. A Product Hunt launch runbook with budgets, off switches, alerts and the checks that break.
Say the balance in your bank app is $7,000 short the morning after launch night: $19,400 when you went to bed, $12,400 now.
You spend $6,200 a month more than you earn, so yesterday that was just over three months of runway ($19,400 ÷ $6,200 = 3.1). This morning it's exactly two. You lost more than a month of your company in one night, and you slept through it.
Your launch checklist covered the tagline and the first comment. Nothing on it mentioned Product Hunt launch fake signups, so you start tracing the money backwards.
The bank app shows 14 charges of $500 from your model provider, the first at 1:08 a.m. and then one about every 26 minutes. That's $7,000. A 15th was declined at 7:08 a.m., when your card hit its daily limit.
Behind the charges is your signup log, and for a moment it's the best thing you've ever seen: 3,650 new accounts. Then you sort by usage. Your launch post promised 250 free generations at $0.008 each, so a fully used offer costs you $2. From 1:08 a.m., nine new accounts a minute used all 250 and never came back.
- 9 accounts × $2 = $18 a minute
- $18 × 60 minutes = $1,080 an hour
- $1,080 × 6 hours = $6,480, across 3,240 accounts
Every minute you slept cost $18. The other $520 went to the people you launched for, all 410 of them. You were about to screenshot 3,650 for your investors, and about 89 in every 100 of those accounts were never going to buy anything.
The declined charge did the rest. At 7:08 a.m. your provider balance ran out and every model call started failing, for everybody. The first ticket from your 140 paying customers lands at 7:14: "Every generation is failing. Are you down?" You get a new card on the account at 7:53, after 45 minutes of errors, and nine customers ask for their $35 back ($315). Real signups had been arriving at about 58 an hour (410 ÷ 7.1 hours), so about 43 people got an error message as their first look at your product. One of them says so in your Product Hunt comments.
The $18 a minute starts again the moment your balance comes back. So at 7:54 a.m., on the most watched day your product may ever have, you get to choose. You can pull the free offer from your launch post with everyone looking, or keep paying $1,080 an hour out of the two months you've got left.
Here's that night on a clock, using the scenario's illustrative numbers.
You can redo that scenario with your own numbers. This next part really happened:
From our early beta
After a launch went viral on X, one team gained 6,400 new users in a month. 2,700 of them were duplicate accounts. About 32 became paying customers.
That's about 42% duplicates (2,700 ÷ 6,400) and 0.5% conversion (32 ÷ 6,400). That launch was on X. A Product Hunt launch gives you a public offer and a day everyone can see coming.
Why Product Hunt launches attract fake signups
A launch post is a public notice. Your offer and your link sit on one page anyone can read, including people who collect free credits. It's also the day your own numbers help you least. On a normal Tuesday, 3,650 signups would set off every alarm you have. On launch day you've been hoping for exactly that, so nothing looks wrong until the money does.
The obvious fix is to add a check to the signup form first. Some of those are worth having. But launch day is when you most want signup to be easy, so each check costs you real people at peak attention, and most stop working as soon as one small thing changes.
Basic signup checks on launch day, and how each one breaks
- Email verification. Catches typos and dead addresses, and accounts that never confirm stop costing you. It misses anyone with a second inbox, and a code stuck in spam loses a real visitor. More on what it proves.
- Unique emails. Refuses the same address twice, nearly free. But Gmail ignores dots, and mail to name+anything@gmail.com reaches name@gmail.com, so one inbox has many addresses. Any rule built on that needs upkeep.
- Disposable-domain blocks. Stop the throwaway services on your list. New domains aren't on it, people using privacy forwarding get refused, and someone has to update it forever.
- CAPTCHA. Filters simple scripts. Some automation gets through, it does nothing about a person signing up by hand, and it trips up screen reader users at your busiest hour.
- Phone or card requirements. Both slow casual second accounts. Neither stops someone with another number or card. People who came to try before buying leave, and every text is a bill. See whether to require a card.
- One signup per IP, and VPN blocks. Stop repeats on one connection and known VPN ranges. They miss every other network and refuse offices, coworking spaces and privacy-minded people. You'll be handling exceptions by lunch.
- Rate limits. Stop bursts from one source. Nine signups a minute isn't a burst, so a limit of 30 never fires here. Set it lower and real launch traffic trips it. They answer a different question.
- Usage caps and shorter trials. A per-account cap is why the damage is $2 an account and not $200, so keep it. Each new account brings a new cap, though, and these used everything in minutes, so a shorter trial changes nothing.
- Manual review. A person can spot what a rule can't. Nine accounts a minute is one every 6.7 seconds, all night, and a tired reviewer will refuse someone real.
Stack them all and signup asks for a verified email, a puzzle, a phone number and a card on the day you most want people to just try it. So don't plan around the checks holding. Plan so a bad night has a fixed price.
Your Product Hunt launch runbook
Two weeks out: price the offer. Work out what a fully used offer costs, then set a budget.
For 800 expected signups at $2, that's $2,400. If your runway can't take it, change the offer now, while nobody's read it.
One week out: separate free money from paid money. Run free usage on its own provider key if you can, with a hard limit at your launch budget, or keep a counter in your app that stops free grants there. Some budget settings only send an alert, so read yours. Turn off auto-reload for the week and give paying customers' jobs their own queue.
Three days out: build the off switches. One flag for the offer and one for costly onboarding work, both reachable from your phone. Hold expensive work until an account is verified. Then flip each flag in production from your phone and confirm it worked.
The day before: set alerts and write your messages. Alert on spend, because launch signups are supposed to spike. Page yourself at 25% of the free budget, loud enough to get through Do Not Disturb. For $2,400 that's $600, about 33 minutes at $18 a minute.
A comment you can post
We've paused the launch offer for a little while because of a wave of fake signups. The product's still up. We'll post here as soon as the offer's back.
Launch night: watch in shifts. Put a name on every hour, even if the only other name is a friend who can flip a flag. When the alert fires, pause the offer, sign in to a paid account and try it, then post your comment. If it's out of hand, the first-hour runbook takes over.
The morning after: keep the evidence and count honestly. Export signup logs and provider usage before they expire. Flag accounts instead of mass-deleting, because real people signed up too, and clean the rest out of your metrics before anyone decides anything from them.
Run the scenario again. The alert fires at about 1:42 a.m. Say it takes nine minutes to wake up and flip the flag. By 1:51 the fake accounts have cost $774 (43 minutes × $18), not $6,480, and your paying customers never see an error.
Look at what you did at 1:51, though. You switched the offer off for everyone, so the real visitors arriving at breakfast find the headline of your launch paused. The runbook limits the bill. It can't say whether this particular claim should get the offer.
Deciding who gets the launch offer
AbuseZero answers that at the moment someone claims your offer, and it goes well past everything on the list above. It puts 140,000+ signals through multi-layer AI risk analysis, which detects linked claims and flags abuse that single checks miss. Where your disposable-domain list needs someone to keep it alive, AbuseZero checks mailbox aliases, disposable domains and mail routing. Where a VPN block refuses people for being on a list, it sees beyond an IP address with VPN, hosting, network reputation, Tor and proxy intelligence. It recognizes repeat devices across new accounts, even when email addresses keep changing, and creating a fresh account isn't enough to erase previous activity. The part you couldn't build in launch week is the proprietary algorithm, with AI that checks the claim again.
Your server makes one API call and gets a clear allow or deny, under rules you control, so launch night has another option besides switching the offer off for everyone. Run it in Observe mode during your beta to see what would be denied before anyone is blocked, then work through the go-live checklist. You don't have to build your own abuse prevention system in the two weeks before a launch. Keep the budgets and flags, since your application still runs signup and usage limits.
Frequently asked questions
How early should I prepare for fake signups on a Product Hunt launch?
Two weeks is comfortable. The minimum is a hard limit on free spend and one off switch that works from your phone, which takes an afternoon.
Should I offer free credits in my launch post?
Yes, if you've priced the worst case. Then evaluate each claim before you grant it. The quickstart covers that in AbuseZero.
How can I tell real launch traffic from fake signups?
Look at what the accounts do. Real visitors try a few things and wander off. If nearly every new account uses the whole offer and vanishes, pause it while you look.
Get your launch ready before it goes live
Start with Builder
Pay as you go. Protect your launch offer today.
See pricing
Builder, Pro, Scale and Enterprise.
Talk to us about Enterprise
High volumes and custom terms.