Fraud ProtectionConversion Recovery

Stop Fraud Without Turning Away Real Customers.

Blocking every risky signup also turns away honest people who mistyped their email or happen to be on a VPN. Score each submission across email, IP, and risk signals, then guide legitimate users to fix the fixable and reserve hard action for real abuse. You stay in control of every decision.

3
outcomes: allow, guide, or step up
200–1000
risk score with explainable reason codes
1 call
field-level signals in one response

14-day free trial. No credit card required.

Create your account
Did you mean jane.doe@gmail.com? Please confirm your email address.
Full name
Email address

Did you mean jane.doe@gmail.com? Please confirm your email address.

Password
Behind the scenes
The hidden cost of blocking

“Block Everything Risky” Costs You Real Customers

If you reject every high-risk score, you stop a lot of fraud. You also send away legitimate people who made an honest mistake, and they rarely come back. The most valuable part of a risk assessment is not the final number, it is the reason behind it.

Honest typos get rejected

A customer types gnail.com instead of gmail.com or drops a character in their address. A blanket block rule treats that mistake the same as an attack, and the signup disappears.

Privacy-conscious users get caught

Plenty of real people sign up from a VPN, a corporate proxy, or a shared cloud network. Reject on the connection alone and you turn away good customers whose only unusual signal was their network.

You never learn why they left

A hard block shows a dead end. The user cannot tell you what went wrong, cannot correct it, and usually leaves without a word. The abandonment looks like normal drop-off in your funnel.

Deliverability quietly erodes

Undeliverable and mistyped addresses that slip through pollute your database, bounce your emails, and drag down your sender reputation over time.

14-day free trial. No credit card required. Includes 1,000 free analyses.

Guide, don't just block

Sort Every Submission Into Three Outcomes

Instead of reacting to a single number, read the individual signals and route each submission to one of three outcomes. The “guide” bucket is where you recover legitimate users a naive rule would have rejected.

Allow

No blocking signal and acceptable risk.

The submission goes through normally.

Guide

The problem is something the user can fix.

A clear, friendly message on the specific field, and a chance to correct it.

Step up or hold

Unrecoverable or clearly abusive.

A firm message, or an extra check such as email confirmation, per your policy.

Fraud Protection surfaces the signals and the risk score. The thresholds, the wording, and the final decision are yours. Everything here is a recommended default you can tune to your own risk appetite.

The signal-to-message map

Turn Each Signal Into a Message People Can Act On

Every row is one situation you can detect, the exact attribute that describes it, and a message that tells the user what to fix. Point email problems at the email field and network problems at the page.

Email situations

Surface these whenever the signal is present, regardless of the overall score. Fixing a mistyped or undeliverable address protects your deliverability and your data.

SituationSuggested message to the userSuggested action
Auto-correction available
emailCorrection is a non-empty string
Did you mean {emailCorrection}? Please confirm your email address.Guide (offer the fix)
Invalid format
isFormatValid is false
The email address provided does not look valid. Please review it.Guide (let them retry)
Mailbox does not exist
isDeliverable is the string "no"
We could not confirm this email address with the provider. Could you double-check it?Guide (let them retry)
Disposable or temporary email
emailType is "disposable"
Temporary or disposable email addresses are not accepted. Please use a permanent email address.Guide (require a different address)
No-reply address
addressSignals.isNoReply is true
This address cannot receive replies. Please use an email address you can check.Guide (require a different address)
Role address (info@, support@)
addressSignals.isRoleAddress is true
This looks like a shared inbox. For your account, please use a personal email address.Guide (soft, tune to your policy)

IP and network situations

Network signals are context, not a verdict. Only act on them when the overall level is high or highest, so you never turn away a privacy-conscious real user.

SituationSuggested message to the userSuggested action
Tor network
connectionType is "tor"
The connection you are using was flagged as high risk. Please switch to your regular internet connection to continue.Guide (they can switch networks)
Open proxy
connectionType is "open-proxy"
We could not verify your network connection. Please switch to a standard connection and try again.Guide or step up
VPN
connectionType is "vpn"
You appear to be on a VPN. If registration fails, try turning it off and submitting again.Allow with a soft hint, or step up
Datacenter or cloud IP
connectionType is "cloud-provider"
We could not verify your connection. Please try again from a standard home or office network.Step up (avoid a hard block)
Blocklisted IP
blocklisted.isBlockListed is true
Your connection was flagged for suspicious activity. Please try a different network or contact support.Step up or hold

Note that isDeliverable is a string with the values “yes”, “no”, or “unknown”, not a boolean. Compare it as text. When it comes back “unknown”, show no message and rely on the overall score.

14-day free trial. No credit card required. Includes 1,000 free analyses.

Putting it together

The Decision Flow, and How It Plays Out

The order matters. Check the specific, fixable signals first so the user gets a precise message, and fall back to the score only for everything else.

  1. 1

    Handle a failed call first

    If the scoring call did not succeed, apply your own failure policy: a short-timeout retry, a queue for manual review, a fallback to your existing validation, or a step-up check. Decide this deliberately rather than letting a failure silently allow or block.

  2. 2

    Offer an email correction

    If a confident correction is available, offer it before anything else. It is the friendliest possible outcome and recovers the user with a single tap.

  3. 3

    Guide fixable email problems

    Invalid format, undeliverable, disposable, no-reply, or role address: show the matching field message and let the user correct it. Email guidance runs regardless of the overall score.

  4. 4

    Check the network only when risk is elevated

    Only if the overall level is high or highest, look at the connection type. Skip this entirely when the score is medium or lower, even if the user is on a VPN.

  5. 5

    Fall back to the overall level

    For everything you did not explicitly check, use the top-level level as the safety net: allow low risk, review medium, step up high, and hold or restrict the highest per your policy.

A mistyped email that can be corrected

Guide

The user types jane.doe@gmial.com. Because emailCorrection is non-empty, you show: Did you mean jane.doe@gmail.com? One tap and the legitimate user is through. A blanket score rule might have flagged this for review; the correction path recovers it instantly.

An address the provider does not recognize

Guide

isDeliverable is the string "no", so you show the review prompt on the email field instead of rejecting the signup. If the mailbox truly does not exist, the user fixes it. If they abandon, you avoided a guaranteed bounce.

A disposable email

Guide

emailType is "disposable", so you ask for a permanent address. This is a clear, honest message, not a silent block, and genuine users who reached for a throwaway inbox will switch.

A high-risk network

Step up

connectionType is "tor" and the overall level is highest, so both conditions are met and you show the connection notice asking the user to switch networks. If that same Tor connection had come back at medium or lower, you would let it through with no network message at all.

Field-level guidance is not a substitute for a firm policy on the genuinely abusive end of the range. Keep a real threshold for the highest submissions, and let the friendly guidance handle the large middle ground where honest mistakes live.

Where Guided Fraud Protection Pays Off

Recover the legitimate users a blunt block rule would lose, without lowering your guard against real abuse.

SaaS Sign-Up Flows

  • Recover legitimate registrations that a blanket block would reject by guiding users to fix a mistyped or undeliverable email.

  • Reserve step-up verification for genuinely high-risk signups instead of adding friction for everyone.

  • Cut false rejections at the top of your funnel while still flagging synthetic identities and disposable addresses.

Checkout and Lead Forms

  • Catch the email typo before an order confirmation or a lead handoff bounces.

  • Show a calm, on-field correction prompt so buyers complete the purchase instead of abandoning at the last step.

  • Keep privacy-conscious shoppers on a VPN moving through checkout when their only unusual signal was the network.

Deliverability and List Quality

  • Prompt users to correct undeliverable and mistyped addresses at capture time, before they pollute your database.

  • Protect sender reputation by keeping guaranteed bounces out of your marketing automations.

  • Turn a data-quality problem into a one-tap fix instead of a silent rejection.

Risk-Based Onboarding

  • Route low-risk users straight through, guide the fixable middle, and hold only the highest-risk submissions for review.

  • Apply your own thresholds and wording on top of explainable reason codes. The product scores; your team decides.

  • Reduce unnecessary manual review by resolving honest mistakes automatically at the field level.

Fraud Protection · No Credit Card Required

Recover the customers a block rule would lose

Guide legitimate users to fix the fixable and reserve firm action for genuine abuse. Score each submission across email, IP, and risk signals. 14-day free trial.

  • Field-level guidance for fixable email problems
  • Explainable reason codes on every score
View Pricing

Read the API implementation guide to ship the decision flow in hours.

Frequently Asked Questions

What does 'guide, don't just block' mean?

It means using the individual signals in a risk assessment, not just the final score, to sort each submission into one of three outcomes: allow, guide, or step up. When a problem is something a real person can fix, a mistyped or undeliverable email for example, you show a clear message on that field and let them correct it instead of silently rejecting the signup. You reserve firm action for the genuinely unrecoverable or abusive cases.

Does Fraud Protection block users for me?

No. All outputs are advisory risk signals. Fraud Protection scores each submission and returns explainable reason codes; your team and your backend logic decide whether to allow, guide, step up, or hold. You own every decision and set your own thresholds and wording.

Won't guiding instead of blocking let more fraud through?

No. Guidance handles the large middle ground where honest mistakes live, while you keep a firm threshold for the genuinely abusive end of the range. In a typical policy your team might refuse a disposable address and hold a blocklisted network, while a customer who simply mistyped their email gets a chance to fix it rather than being turned away alongside the attackers. The signals are advisory; you decide what each outcome does.

Which signals should I show as a field message?

Email situations are worth a message whenever they appear, regardless of the overall score: an available correction, an invalid format, an undeliverable mailbox, a disposable address, a no-reply address, or a role address. These are all things a real person can correct in seconds. Network situations (VPN, Tor, proxy, datacenter, blocklisted IP) should only surface when the overall level is high or highest.

Why treat email and network signals differently?

A mistyped or undeliverable email is a data-quality problem worth fixing even on a low-risk submission, because it protects your deliverability. A network signal is context, not a verdict. Plenty of legitimate users are on a VPN, a corporate proxy, or a shared cloud connection, so acting on the network alone turns away privacy-conscious real customers. Only weigh the connection when the overall risk is already elevated.

How do I detect an email typo I can auto-correct?

Read sources.email.emailCorrection. When it is a non-empty string, the API is confident about the intended address (for example jane.doe@gmail.com for a typed gmial.com). Offer that correction first; it is the friendliest possible outcome.

Is isDeliverable a true or false value?

Neither. isDeliverable is a string with the values “yes”, “no”, or “unknown”. Compare it as text, not as a boolean. When it is “no”, prompt the user to review the email field. When it is “unknown”, show no message and rely on the overall score.

What should I do when the scoring call fails?

Decide your failure policy up front rather than letting a failure silently allow or block. Common approaches are a short-timeout retry, queuing the submission for manual review, falling back to your existing validation, or requiring a step-up check. Choose the path that matches your risk tolerance and the sensitivity of the action.

How do I integrate this?

Call the Fraud Protection API from your backend with at least an email or an IP address, then apply the decision flow to the response. The full request and response schema, the signal-to-message map, and copy-paste-ready decision logic are in the API implementation guide.

What does it cost, and is there a free trial?

Fraud Protection uses analysis-based monthly subscription pricing, and billing is triggered only by completed analyses. A 14-day free trial is available with no credit card required. View current plans and pricing.