Why Your Signup Form Collects Fake Emails (and How to Stop Them)
HeroBounce Team
You open your signup list and the problems are easy to spot. Addresses at domains you have never heard of. Random strings before the @. A burst of forty signups inside two minutes. Then the welcome emails go out and a chunk of them bounce.
"Fake email" covers at least five different problems, and each one has a different fix. This guide shows how to tell them apart, which defences work against which problem, and what to do with the bad addresses already in your database.
Five kinds of bad email, and how to tell them apart
| Type | What it looks like | Where it comes from | Best fix |
|---|---|---|---|
| Typos | john@gmial.com, anna@gmail.con | A real person with a slip of the keyboard | Typo suggestion at the form |
| Disposable addresses | A throw-away domain you do not recognise | Someone who wants your free resource without a real inbox | Block at the form |
| Bot signups | Random local parts, bursts of signups seconds apart | Scripts that find your form and submit it at speed | Rate limiting, honeypot, CAPTCHA, validation |
| Deliberate fakes | Valid-looking address, often disposable | People who want a free trial or gated download and give you nothing real | Validation, and a decision about confirmation |
| Dead or shared mailboxes | An old work address, info@, sales@ | A real person who typed an address that no longer works or nobody owns | Mailbox check, role flag |
The first group is an accident. The rest are choices, and they respond to friction in different ways.
Diagnose before you fix
Four quick checks tell you which problem you have.
- Group signups by email domain. A long tail of unfamiliar domains points to disposable services. One unknown domain repeating many times points to a bot.
- Sort by time. Bots submit in bursts. Real signups are spread out.
- Read the local parts. Strings like asdf1234 or xkqjw83 are not names. Several addresses built on the same pattern come from the same script.
- Look at your confirmation rate and your bounces. Bounces with a "mailbox does not exist" reason come from typos and fakes. A low confirmation rate on a double opt-in form points to bots and fake data. On our own signup form, about 1 in 6 accounts never confirms its email.
Then compare your bounce rate with the limits mailbox providers use. BounceZero's 2026 benchmarks put the safe hard-bounce ceiling at about 2%, and Google's sender guidelines say to keep your reported spam rate below 0.10%. If you are above either, bad signups are already costing you reputation.
The layers that work, and what each one costs
Vendors who write about this agree on one point: no single control is enough. The practical approach is to stack cheap layers, from the ones real users never notice to the ones they do.
| Layer | What it stops | Cost to a real user |
|---|---|---|
| Format check and typo suggestion | Typos and malformed addresses | None |
| Honeypot field (hidden, bots fill it in) | Simple bots | None |
| Rate limiting per IP address | Bursts of signups | None, unless many people share one IP |
| Real-time email validation | Mailboxes that do not exist, disposable addresses, no mail server | A short wait while the check runs |
| CAPTCHA | Most scripted signups | A click, sometimes more |
| Double opt-in | Anything that cannot confirm | Real users who never click the link |
Double opt-in is the strongest filter, and it is also the one that costs you real people. We compare it with single opt-in in our double opt-in guide. Real-time validation covers most of the same ground without asking the user to leave the form and open their inbox.
What real-time validation should do at the form
A validator does more than a regex. It checks the format, looks up the domain's mail servers, asks the mail server whether the mailbox exists, and flags disposable and role-based addresses. Each result comes with a reason, which is what you need to decide what to tell the user.
| Result | Reason examples | What to do at signup |
|---|---|---|
| Valid | Mailbox accepted | Accept |
| Invalid | Mailbox not found, no mail server, invalid format, disposable | Ask for a different address. If a typo suggestion comes back, offer it |
| Risky | Accept-all domain, server deferred, server unreachable | Accept, and flag the account for a later check |
| Unknown or timed out | The mail server took too long | Accept, then re-check in the background |
Two rules keep this safe for real users:
- Never reject on "risky". Many company domains accept mail for any address, and a deferred answer says nothing about the mailbox. Rejecting these turns away real people.
- Fail open. If the check times out, let the signup through and re-check later. A slow mail server should not cost you a customer.
With HeroBounce, the REST API is available on paid plans, and each response includes the status, the sub_status reason and a typo suggestion when there is one. The free plan includes 300 validations a month in the dashboard and through AI assistants. The API documentation and our integration guide show the request and response in full.
Clean up what you already collected
Stopping new bad signups does nothing for the ones already in your database.
- Export your signups as a CSV.
- Validate the file. Upload it to bulk email verification.
- Download the lists you need. Valid, Invalid, Catch-all and Risky come as separate downloads, and the Overview Report shows the split.
- Suppress the Invalid list and keep it out of every send.
- Treat Catch-all and Risky separately instead of mixing them into your clean list.
- Re-validate on a schedule. Addresses go bad over time, so a list that was clean at signup is not clean a year later. See our email list hygiene guide.
Common mistakes
- Relying on client-side checks only. A bot posts straight to your endpoint and never runs your JavaScript. Validate on the server.
- Telling bots exactly why they were rejected. Detailed error messages help script writers. Keep the message short and human.
- Blocking every role-based address. On a B2B form, info@ or hello@ may be the only address a small company uses. Flag them, and decide per use case. Our role-based address guide covers when to keep them.
- Treating a timeout as a rejection. Slow servers are normal. Let the signup through and check again.
- Stopping at the form. The bad addresses that got through, and the ones that went stale, are still on your list.
Frequently asked questions
Why am I getting so many invalid emails on my signup form?
Usually it is a mix of typos, disposable addresses and bots. Group your signups by domain and by time to see which one dominates, then match the fix to the cause: typo suggestions for typos, blocking for disposable addresses, rate limiting and a honeypot for bots.
How do I stop fake emails without hurting real signups?
Stack layers that real users do not notice: a typo suggestion, a honeypot, rate limiting and real-time validation that fails open. Add a CAPTCHA or double opt-in only where the abuse justifies the friction.
Can email validation stop all fake signups?
No. Validation removes addresses that do not exist and disposable ones. A bot that uses a real, working mailbox will pass, which is why you combine it with rate limiting and a honeypot.
Should I use double opt-in or real-time validation?
They solve overlapping problems. Double opt-in is the strongest filter but loses real users who never confirm. Real-time validation catches most bad addresses at the form with no extra step, and you can still use double opt-in where the risk is high.
How fast does validation need to be on a signup form?
Fast enough that users do not wait, and tolerant enough that a slow mail server does not block a signup. Run the quick checks inline, set a timeout, and re-check in the background if it expires.
Clean the list before you send
HeroBounce checks syntax, DNS, MX and the mailbox itself over SMTP, then tells you why each address is valid, invalid or risky. Download the results as valid, invalid, catch-all and risky lists.
300 free validations every month | Paid plans from $17/month