Disclosure: I work on SendBridge, an email verification service. Most of the controls below work with any tooling, or none at all.
When risk teams talk about onboarding fraud, the conversation usually jumps straight to document checks, device fingerprinting and sanctions screening. The humble email address field gets far less attention, even though it is the first piece of identity data almost every customer gives you. In my experience it is also one of the cheapest places to catch problems early.
Why the email field matters for risk
An email address is used for account recovery, transaction notices, legal communications and, increasingly, as a key that links a customer record across systems. If that address is fake, disposable or mistyped, several things go wrong at once:
- Synthetic identities get a foothold. Fraud rings create thousands of accounts with throwaway inboxes to abuse sign-up bonuses, free trials or referral programmes.
- Required notices never arrive. If you are obliged to send a statement, a terms change or a security alert, a dead address undermines your ability to show the customer was informed.
- Account takeover gets easier. A typo such as
gmial.comcan mean recovery emails land in a domain someone else controls. - Your own sending reputation suffers. High bounce rates can push legitimate mail, including fraud alerts, into spam folders for everyone.
A layered approach that does not hurt conversion
The goal is not to interrogate every customer. It is to remove obvious garbage quickly and route the uncertain cases to the controls you already have.
1. Syntax and typo checks in the form
Basic format validation plus a suggestion for common domain typos ("Did you mean gmail.com?") fixes a surprising share of bad addresses before they are ever submitted. It costs nothing and customers generally appreciate it.
2. Domain and mailbox checks at submission
Before creating the account, confirm that the domain has valid MX records and, where possible, that the mailbox exists. This is the step that catches invented addresses. It can be done with open-source libraries, your own SMTP checks (carefully, to avoid being rate limited) or a verification API.
3. Disposable and role-based address policies
Decide explicitly how you treat disposable inbox providers and role accounts like info@ or admin@. For a newsletter, they may be fine. For a regulated financial product, a disposable address is a strong signal that should trigger step-up verification rather than an outright block.
4. Confirmation as a control, not a formality
Double opt-in or a confirmation link proves the person controls the inbox at that moment. Track how long confirmations take and how many never happen; sudden changes in those numbers are often the first sign of a scripted attack.
5. Re-verification over time
Addresses decay. People change jobs, domains expire, inboxes get abandoned. Periodically re-checking the customer base, especially before important mailings, keeps your records accurate and gives you an early warning when a cluster of accounts suddenly goes dead.
Signals worth feeding into your risk engine
- Domain age and whether the domain has working mail servers at all
- Disposable or free provider versus a company domain
- Catch-all domains, where any address appears valid
- Many sign-ups sharing a pattern (e.g.
name+1,name+2) or the same new domain - Mismatch between email country signals and other identity data
None of these should decide an application on its own, but combined with device and behavioural data they make scoring noticeably sharper.
Practical tips for implementation
- Fail open on timeouts: if a verification check is slow, let the user continue and re-check asynchronously.
- Log the verification result with the account so investigators can see it later.
- Review your thresholds quarterly together with fraud and marketing teams; blocking too aggressively has a real cost.
- Document the control in your onboarding procedures so auditors understand why it exists.
For teams that prefer not to build the mailbox checks themselves, tools such as SendBridge provide an API and bulk list cleaning, but the policy decisions above matter more than the vendor you choose.
Email verification will never replace proper KYC. It is, however, a low-friction first line of defence that reduces noise for every control that comes after it.
Peter Braun (SendBridge)
Comments