Use cases

Fake account creation and free trial abuse: stop bot signups

Stop bots and signup farms creating accounts to claim free trials, credits, bonuses and referral rewards. Prosopo Protect checks every signup request, on the page and the API, before the account exists.

What fake account creation is

Fake account creation is opening accounts that no real customer will use, to collect whatever a new account is given. It is usually automated, and it is usually invisible as an attack, because each account on its own looks like an ordinary new customer who signed up once and never came back.

The value is in the signup offer, so the services that give the most away to a new account are the ones that get farmed: SaaS and AI products with free credits or a free tier, marketplaces and fintech apps with signup bonuses and referral rewards, retailers with first-order discounts, and iGaming sites with welcome bonuses.

What fake accounts are used for

AbuseWhat the attacker getsWho pays
Free trial and free tier abuseYour product, or its API, without paying, often resoldYou, in compute, storage and model credits
Signup bonus and promo code abuseThe bonus or discount, once per accountYou, in the value of the offer
Referral fraudReferral rewards for referring their own fake accountsYou, in rewards for customers who do not exist
SMS verification abuseTexts sent to number ranges a rogue carrier shares revenue onYou, per message; see SMS pumping
Spam and fake reviewsAccounts that can post, message or reviewYour real users, and your reputation
Accounts for resaleAged or verified accounts sold to other abusersYou, when those accounts are used later

How fake signups are made

The signup is usually scripted. A headless browser or an anti-detect browser fills in the form, or the script skips the form and calls the signup API directly. Each signup gets a new email address from a disposable-email service or an alias of one inbox, a new phone number from a temporary-number service, and a new network address from a residential proxy pool, so no single address, email domain or IP stands out.

Where the form has a CAPTCHA, commercial solver services answer it for a small fee per thousand. Where checks are strict enough to defeat scripts, some operations use people in signup farms, working through the same proxies and phone-number services, which is slower and more expensive for them, so it is used where each account is worth more.

Signs of fake account creation

At signup:

  • Bursts of new accounts from one network, one proxy provider or a country you do not usually sell to.
  • Many accounts created from one device or browser profile.
  • Email addresses from disposable-email domains, or the same pattern repeated with a number or a plus tag added.
  • Phone numbers from one range, or verification texts to countries your customers are not in.
  • CAPTCHAs answered by a solver service rather than by the visitor.
  • Signups through the API with no matching page visits.

After signup:

  • The free allowance used immediately and in full, then no further activity.
  • Referral rewards paid between accounts created in the same hour from the same network.
  • Many accounts with the same payment card, delivery address or device.

A CAPTCHA on the signup form stops only the cheapest bots

A CAPTCHA asks one question at one moment, and fake account creation has three ways around it. Solver services answer the challenge, signup farms are real people who pass it, and a script calling the signup API never loads the page the CAPTCHA is on. A CAPTCHA still removes the cheapest bots, which is worth having, but it cannot be the only check on a signup flow that gives away something of value.

The check that holds up runs on every signup request, page or API, and looks at the visitor rather than at whether a puzzle was solved: how it behaves, where the connection really comes from, and whether the browser's description of itself holds together.

How to stop fake account creation

  1. Check for automation before the account exists. Cleaning up fake accounts afterwards means the bonus, the credits or the text message has already been paid for.
  2. Protect the signup API as well as the page. Scripts move to the API as soon as the page is protected.
  3. Reject disposable email domains at signup, including domains that redirect to a disposable service through DNS.
  4. Send SMS verification only after the bot check passes, and restrict which countries you send to.
  5. Tie trial value to more than the email address. Limit free credits per device and network as well, so a new email address does not reset the allowance.
  6. Pay referral rewards after real activity, not at signup.
  7. Watch new accounts after signup for the patterns above, and keep the evidence for each block so you can explain it to a real customer who was stopped.

How Prosopo stops fake signups

Prosopo Protect checks every request to your signup page and signup API, and blocks automation before the account is created, while real people sign up without seeing a challenge.

  • How the visitor behaves. Mouse movement, scroll rhythm and typing cadence on the signup form. Automation repeats itself in ways a person does not.
  • Where the traffic really comes from. Residential proxy detection checks the TCP connection for signs of a relay and matches exit addresses against the proxy networks we rent and probe ourselves, so a new home broadband address for every signup does not hide the operator.
  • Browsers that contradict themselves. Anti-detect browsers report one device and behave like another, and named detectors such as Inconsistent hardware readings for device catch the contradiction.
  • Solver services. A CAPTCHA answered by a paid solver is a block reason of its own: Solver service detected.
  • Disposable email and abusive networks. The Spam Filter rejects throwaway email domains, including those hidden behind DNS redirects, and can block VPN, proxy, Tor and datacentre traffic per site.
  • Your own rules. Access Rules match on IP range, ASN, country, user agent, JA4 TLS fingerprint and IP category, so a network that farms your signups can be blocked or challenged as soon as you save the rule.

Protect is a script tag on your pages, with edge protection on Cloudflare Workers or AWS Lambda@Edge, so signup calls to your API are checked before they reach your servers. API calls that are blocked get the right HTTP status and an X-Prosopo-Decision header, so your app can handle a block cleanly. Every block names the reason it was made.

Free-tier abuse at a subscription search product

A subscription search product had its free tier consumed at scale by a competitor reselling the results through the API. Over thirty days, Prosopo scored 7,824,429 sessions and stopped 315,766 as automated, and 79% of real visitors cleared without seeing a puzzle. The automated client ran slowly from home broadband addresses, so its rate did not stand out, and what caught it was the client contradicting itself. The full figures are in the scraped SaaS case study.

Choosing fake account protection

Check these before you compare prices:

  • The signup API is covered, not only the signup page.
  • It runs before the account exists, so no bonus, credit or text message is paid out to a bot.
  • Residential proxies are caught on the connection, because IP block-lists do not catch a new home broadband address for each signup.
  • Every block has a reason, so your team can see why a signup was stopped and answer a real customer.
  • The price does not rise with the attack. Protect is priced by number of sites rather than per request, within fair use, from $1,990 a month, and starts with 14 days in preview mode so you can see what it would block before it blocks anything.

Form spam, where the goal is the message rather than the account, is covered in spam bot protection. Logins to real customers' accounts are covered in account takeover prevention. To compare vendors on price and deployment, see bot management solutions compared.

See what's reaching your site

A Bot Exposure Assessment is a 30-minute call with an engineer. We look at your traffic, show where your current setup leaves gaps, and agree how a trial would be judged. Nothing to install first.

Questions

Something else? Ask an engineer.

What is fake account creation?

Fake account creation is opening accounts that no real customer will use, usually with automation, to claim what a new account is given: a free trial, free credits or API quota, a signup bonus, a referral reward or a promo code. The accounts are also used to send spam, post fake reviews or sell on. The attacker repeats the signup with a new email address, phone number and network address each time, so the service sees many new customers instead of one abuser.

What is free trial abuse?

Free trial abuse is one person or group opening many accounts to keep using a free trial or free tier without paying, or to resell what the free tier provides. For a SaaS or AI product the cost is real, because each fake account consumes compute, storage, SMS or model credits that a paying customer would otherwise fund.

How do you stop fake account creation?

Check the signup request for automation before the account is created, rather than cleaning up accounts afterwards. That means checking how the visitor behaves, whether the connection comes through a residential proxy, and whether the browser's claims hold together, on the signup page and on the signup API. Then reject disposable email domains, send SMS verification only after the bot check has passed, and tie trial value to the device and network as well as to the email address.

Can a CAPTCHA stop fake signups?

Only some of them. A CAPTCHA on the signup form stops simple bots, but commercial solver services answer CAPTCHAs for a small fee, signup farms use real people, and many services also accept signups through an API the form calls, which a widget never sees. Checking every signup request for automation, including API calls, and naming the reason for each block, holds up better than a challenge alone.

How do you detect fake accounts?

Look at how the accounts were created as well as what they do. Signs include bursts of signups from one network or proxy pool, many accounts from one device, disposable or patterned email addresses, phone numbers from the same range, a CAPTCHA solved by a paid solver service, and accounts that use the free allowance immediately and then go quiet.