What Should You Do in the First 24 Hours After Finding a Phishing Domain?
Published 2026-11-14 · Reviewed 2026-11-14 · 7 min read · By the Ingenious AI team
Is someone impersonating your domain right now?
Run the free DomainGuard scan: lookalike registrations, MX-armed variants, certificate-transparency and CAA checks for your domain, in one report.
Run my free DomainGuard scanThe short answer
In the first 24 hours after finding a phishing domain impersonating you: preserve evidence before anything changes, file browser blocklists and the hosting provider in the first hour, file the registrar in parallel, warn customers through channels the attacker cannot imitate, and check whether any accounts or payments were actually compromised. Do not contact the attacker, do not pay for the domain, and do not wait for one channel to answer before opening the others. The incidents that end cheaply are the ones where parallel filing started in hour one; the expensive ones spent day one deciding who owned the problem.
Hour zero: preserve evidence and name an owner
Before any filing, capture what exists, because takedowns destroy the scene. Screenshot the phishing pages with the address bar visible, record the full URLs, the domain's registration details, its DNS answers, and its certificate information, all timestamped. Save any phishing message that delivered the link, with full headers. Store it in one incident folder with one owner named — incidents without a named owner drift, and drift is the attacker's best ally.
In the same first minutes, make the three triage calls. Is the site live and harvesting — credentials, payment details, downloads? Is it being distributed — mail, texts, ads, search results? And is there any sign of actual victims yet: support reports, payment disputes, account anomalies? The answers set your tempo, but they never change the first-hour filings. A live harvesting site gets the full playbook at full speed; a dormant lookalike gets the filings without the customer warning.
Hour one: blocklists and host, in parallel
File Google Safe Browsing and Microsoft SmartScreen first, because warnings protect users within hours, while removals take longer. File the hosting provider second — precisely addressed, with the exact URLs and your evidence pack, because hosts suspending content is usually the fastest true removal. If the site sits behind a CDN, file the CDN's abuse form too and start the origin hunt through historical DNS rather than waiting.
Nothing in that paragraph is sequential permission-giving. Three filings, one evidence pack, one hour. DomainGuard findings arrive with the infrastructure identified — host, CDN status, registrar, abuse contacts — and the takedown packet assembled, which is why monitored brands start this playbook at filing speed instead of research speed. If you discovered the site through a customer report instead, expect the research to eat your first hour, and note that cost for the post-incident review.
Hours two to six: registrar, warnings, and the compromise check
File the registrar abuse report with the full evidence pack and the policy citation, asking plainly for suspension for phishing. Registrar action is the definitive end — the domain dies everywhere at once — and its timeline is the least predictable, which is precisely why it must never be the only filing. Chase it on evidence, not on anxiety: new URLs, new victim reports, and non-responses escalated to ICANN compliance after a reasonable interval.
Warn through unforgeable channels: a notice on your real website, your real social accounts, and — for SaaS and account businesses — an in-app message. Name the fake domain exactly, say what you will never ask for, and tell people what to do if they used the site. In parallel, run the compromise check with your support and engineering teams: password-reset anomalies, new-account spikes, payment disputes, and any customer who reports interacting with the clone gets a forced credential reset and a session review.
Hours six to twenty-four: verify, expand, and watch for the sequel
Verify removal yourself rather than trusting confirmations: resolve the domain, load the URLs, and check the blocklist warnings in a real browser. Attackers move kits between hosts and URLs inside the same domain, so a single dead page is not the end — re-check the domain's other paths and any sibling lookalikes in your monitoring queue, because campaigns travel in packs and the second domain is often registered the same day as the first.
Use the remaining hours for the unglamorous completions. Log the full timeline while memories are fresh. Notify your insurer or payment provider if their thresholds were met. Brief support with the approved answers for victim calls. And file the law-enforcement report for cases involving financial loss — late in the sequence not because it is unimportant, but because its value does not decay by the hour the way warnings and recalls do.
After day one: the review that buys down the next incident
Within the week, hold a thirty-minute review with one question per phase: how did we find it, and how could that have been sooner? How fast were the filings, and what slowed them? What did victims experience, and what did our warning actually reach? Every answer becomes a concrete change — a monitoring alert routed properly, a template pre-written, a contract clause for takedown authority, a defensive registration purchased.
Track your numbers across incidents: time to detect, time to warn, time to remove. Those three figures are your brand-protection program in miniature, and they almost always improve dramatically between the first incident and the second — if the review happens. The organisations that stay expensive to attack are not the ones that never get impersonated. They are the ones whose second response is visibly faster than their first.
Five actions to take today
- Preserve the evidence pack and name a single incident owner before filing anything.
- File Google Safe Browsing, Microsoft SmartScreen, and the hosting provider within the first hour, in parallel.
- File the registrar abuse report the same day and warn customers through your real site, socials, or in-app channels, naming the fake domain.
- Run the compromise check: customers who used the clone get forced resets and session reviews; check payment disputes.
- Verify removal yourself, re-check sibling lookalikes, and hold a 30-minute post-incident review within the week.
Worked example: Northbeam Outfitters runs the 24-hour clock
Northbeam Outfitters, a fictional retailer monitoring 5 domains, learned of a live clone from a CT alert at 08:10. Evidence was captured by 08:25; blocklist and host filings went out by 09:05; the registrar filing and a customer notice naming the fake followed by 11:00. Browser warnings were live by midday, the host suspended the content at 15:30, and 7 customers who had visited the clone were identified from support contacts and given forced resets — none had completed a payment on it. The registrar suspended the domain the next day. Total measured harm: zero payments, zero confirmed credential losses, one long day.
FAQ
Should we contact the site operator to ask them to stop?
No. Contact tips off the attacker, accelerates the fraud or the move to a new domain, and occasionally invites extortion. All communication goes through abuse desks, blocklists, and law enforcement.
When do we warn customers — immediately or after confirmation?
As soon as a live harvesting site is confirmed, through channels the attacker cannot copy. A short factual notice naming the domain protects people mid-campaign; waiting for perfect information protects nobody.
What if the phishing site is found on a weekend?
The playbook is the same and the filings are all web forms — this is why the evidence template and takedown authority are prepared in advance. Blocklist and host clocks run on weekends too.
How do we know the incident is over?
When the domain is suspended or the content is gone and verified by your own checks, sibling lookalikes are reviewed, and compromise checks come back quiet. Keep the monitoring alert open a few extra days for sequels.
Who should own the incident?
One named person, typically whoever owns security or operations, with authority to file, warn, and force resets. Incidents fail on diffusion of responsibility more often than on any technical factor.