1. Red
  2. Learn
  3. Incident response

Email clients about a security breach: template and timing

Copy-paste email templates for telling customers your website was breached: what to say first, what to leave out, and which deadlines start when you find out.

By Timothy Lindblom, FounderPublished 2 November 202610 min readHow we check facts
On this page
  1. What do I do before I write the email?
  2. Who has to be told, and by when?
  3. What must a breach email contain?
  4. Which template should I send?
  5. Template 1: first notice, scope still open
  6. Template 2: full notification, scope confirmed
  7. Template 3: incident notice, no customer data involved
  8. Template 4: the follow-up
  9. What phrases should I leave out?
  10. What if my agency or host had the breach, not me?
  11. How should I send it?
  12. Where does a scan fit?

Send customers a short, factual email as soon as you know their data may be involved. Say what happened, which data, what you have done, what they should do and who to contact. State what you do not yet know and give a date for the next update. Do not write that data was unaffected until you hold evidence.

FactDetailSource
Who it is forSite owners, shops and agencies who run a site that holds customer datathis page
Authority deadline (EU and UK)72 hours after becoming aware, where feasibleGDPR Article 33, ICO
Individuals (EU and UK)Without undue delay, when the breach is likely to be high riskICO
Individuals (New York)Within thirty days of discoveryNY GBS 899-aa
Individuals (Florida)No later than 30 days after determination of a breachFla. Stat. 501.171
Content the FTC expectsHow it happened, what was taken, what you did, what people should do, contact detailsFTC guide
Checked24 October 2026this page

This page gives you wording and an order of work. Laws differ by country and by US state, and the one that applies is set by where your customers live. If your contract or your customers' data makes you a processor, read the processor section below before you send anything.

What do I do before I write the email?

Work out four facts first. An email written before you know them tends to promise things you later retract.

  1. What the attacker could reach. A defaced page, a planted script or an unfamiliar admin account can each sit next to a customer table. List what that account, file or access could read.
  2. When it started and when you knew. Several deadlines run from the day you became aware, so write that date and time down now.
  3. Which customers and which data. Names and emails, order history, addresses, passwords, payment details and uploaded documents carry different advice for the reader.
  4. What you have stopped. Credentials changed, the hole closed, logs kept.

If you are still working out whether the site was hit at all, the signs of a hacked website come first, and the first 24 hours checklist covers the rest of the clean-up. Do not wipe evidence while you do this. The FTC says not to destroy forensic evidence during investigation and fix, and CISA says to take system images and keep logs before rebuilding. Credentials come next: the FTC notes a system stays exposed until the credentials are changed, even after the attacker's tools are gone. Our walkthrough for rotating keys and passwords gives the order.

Order of work before and after the customer email Preserve logs and images Scope data and dates Decide who must be told Send first factual note Update on the date given
Preserve, scope and decide come before the first email. The email promises an update date, and the update keeps that promise.

Who has to be told, and by when?

Three groups can have a claim on you, and each has its own clock. The email to customers is only one of them.

WhoWhenSource
Data protection authority (EU, UK, Sweden)Within 72 hours of becoming aware, where feasible, unless a risk to people is unlikely. Late notice needs reasonsGDPR Art. 33, IMY
Individuals (EU and UK)Without undue delay when the breach is likely to be high risk, in clear and plain languageGDPR Art. 34, ICO
Residents of New YorkWithin thirty days of discovery, plus state agenciesNY GBS 899-aa
Residents of FloridaNo later than 30 days after determination; the state department when 500 or more are affectedFla. Stat. 501.171
California Attorney GeneralA sample notice when more than 500 California residents are notifiedCalifornia AG
Police, FBI or CISA (US)Report the incident; CISA lists its report page and the FBI field officeCISA

Checked 24 October 2026. Statutes get amended, and these rows are examples of the pattern and leave out many laws. Two points hold across them. Several clocks run from the day you became aware or determined a breach, and the ICO says to report early and update later instead of waiting for a full picture.

The FTC adds one more rule for US readers: ask your law enforcement contact about the timing of notification so it does not impede the investigation.

What must a breach email contain?

Put the following in this order. Customers read the opening lines and forward the rest.

  1. What happened, in one plain sentence.
  2. Which kinds of data were involved, and which were not, as far as you know.
  3. When it happened and when you found out.
  4. What you have done.
  5. What the reader should do, specific to the data type.
  6. A named contact, a real address or phone line, and the date of the next update.

The FTC guide lists the same elements, and says not to make misleading statements and not to withhold key details. The ICO asks for a contact point, the likely consequences, the measures taken or proposed, and practical advice such as password resets. The GDPR calls for clear and plain language in the message to individuals.

Which template should I send?

Pick by what you know today, not by how the news cycle feels.

What you knowStateTemplate
Personal data was accessed or may have been, and you do not yet know how muchDetected, scope open1: first notice
Your review confirms which data and which customersConfirmed by owner evidence2: full notification
An attacker was on the site, but the access could not reach customer data, shown by evidenceDetected, no data path found3: incident notice, no data involved
Time has passed since an earlier emailAny4: follow-up

Template 1: first notice, scope still open

Use this when you have reason to think personal data may be involved and the review is running. It tells the customer early and does not claim more than you know.

Subject: Security incident at [company]: what we know so far

Hello [first name], on [date] we found that someone gained unauthorised access to [the website / our online store]. We are writing now because information about you may have been involved: [kinds of data, for example names, email addresses and order history]. We do not yet know whether it was copied.

What we have done: we closed the access we found on [date], changed the passwords and keys that could have been used, and kept the logs for review. What we do not know yet: how many customers are affected and whether payment details were reachable. We will write again by [date], whatever we have found.

What you can do now: be careful with emails that mention your order or account, and do not click links in a message that asks you to log in. If you reused your [company] password elsewhere, change it there. Questions go to [named person, address, phone].

Template 2: full notification, scope confirmed

Send this when the review has named the data and the customers. Open with the facts, not a reassurance.

Subject: Notice of data breach affecting your [company] account

Hello [first name], we are writing to tell you that between [start date] and [end date], an unauthorised person had access to [system]. We found it on [date] and stopped it on [date]. The information involved for your account was: [exact list]. The information that was not involved: [list, only what the review confirmed].

We have [fixed the hole, replaced credentials, engaged a specialist, told the authority on date]. We do not know of any misuse so far, from the review of [logs and dates reviewed]. That is a statement about what we examined, and we will tell you if it changes.

What to do: [step for each data type: change your password; watch your card statement and tell your bank; treat messages about your order with care]. Contact: [named person, address, phone, hours]. We will write again by [date].

Template 3: incident notice, no customer data involved

Use this for a defaced page, a planted script or a hidden admin account where you hold evidence that customer data was out of reach. Write down how you decided.

Subject: [company] website: an incident we found and fixed

Hello [first name], on [date] we found that someone had changed [what: pages, a script, an admin account] on our website. We removed it on [date]. We reviewed [the database access logs and account permissions] for [date range] and found that the access could not reach customer records. We will tell you if that conclusion changes. No action is needed from you. Contact: [named person, address].

The test for this email is the evidence in the sentence. If you cannot name what you reviewed, send Template 1 instead. The ICO's worked cases include credential theft that exposed customer data, where both the regulator and the clients were told, and a compromised account used for customer communications was part of the reasoning.

Template 4: the follow-up

Send it on the date you promised, even when nothing has changed. A missed date reads as a second incident.

Subject: Update on the incident at [company], [date]

Hello [first name], an update to our email of [date]. Since then we have [what was reviewed, what was found]. This changes the picture in one way: [what changed, or "nothing has changed"]. Still open: [list]. Next update: [date]. Contact: [named person].

What phrases should I leave out?

Customers have read many breach emails. Certain phrases signal that a company is managing perception.

AvoidWhyWrite instead
"We take security seriously"The reader has just been told otherwiseThe steps you took, with dates
"A sophisticated attack"Unknown until investigated; reads as an excuse"Someone gained access using [credential / outdated component], as far as we know"
"No evidence of misuse"True only if you looked"We reviewed [logs] for [dates] and found [result]"
"Your data was not affected"Cannot be known before scope is confirmedThe list of what was and was not involved, per your review
"Out of an abundance of caution"Hides that the risk is real"Your information may have been involved"
"Resolved" with no dateA hole closed and a password changed are different facts"Access closed on [date]; credentials changed on [date]"

Do not name a cause you have not confirmed. The FTC tells businesses to describe what they know about the compromise, not to speculate. For a third-party cause, such as a plugin or an agency login, say only what the evidence shows. If a service provider says the fault is fixed, the FTC says to verify that they really fixed it. Our page on patching duties split between host, agency and owner explains who owns which layer.

Am I affected? Which email can I send right now

A scan from outside cannot see your database, your logs or which accounts were used. The state of your email depends on evidence you hold.

  1. Detected You found unauthorised access, a planted file or a rogue admin, and personal data may have been reachable. Send Template 1 now and keep the logs.
  2. Unknown Something looks wrong but you cannot yet say what the access could read. Do not send a no-impact email. Scope first, then choose Template 1 or 3.
  3. Confirmed by owner evidence Your review names the data, the customers and the dates. Send Template 2, and tell the authority if the law that covers you requires it.

No scan, ours included, moves you from "unknown" to "confirmed".

What if my agency or host had the breach, not me?

Then your first duty is to find out what they hold about your customers. The GDPR says a processor tells the controller without undue delay after becoming aware. Sweden's IMY says the controller stays responsible for reporting even when a processor had the breach. Under Florida law a third-party agent tells the covered entity within 10 days of determining a breach.

Ask the provider these questions in writing:

On what date did you become aware of the incident affecting [our site / our customers' data]? Which of our customers' data categories could the access reach, and for which dates? What logs do you hold, and have they been preserved? What did you change, and at what time? Who is your named contact for the next update?

If you are the agency, and the client is the controller (see who carries the liability when a client site is hacked), send them the facts first and let them decide the customer wording. Our ready-made client notification emails cover the portfolio side. For the log review itself, the first-hours response checklist lists what to preserve and rotate.

How should I send it?

  • Send one email per customer from a named person at a real address that can receive replies, not a no-reply address.
  • Do not put the notice in a newsletter, a footer or a link to a blog post. Send the facts in the body.
  • Do not ask the customer to click a link to log in. Breach emails are a favourite cover for phishing, so give the instruction in words.
  • Put a plain version of the notice on your website for customers with no email on file.
  • Keep a copy of every email, the send time and the list it went to.
  • Write down the date you became aware, the date you decided, and the reason you did or did not notify each group.

The GDPR requires controllers to document every breach, including the facts, effects and the action taken, and the ICO says to document and justify a decision not to report. Keep that record even if you choose not to notify. More guides like this sit with the other incident response pages. The NCSC says a well-prepared response helps limit public and media fallout, and its small business guide gives a 24-hour number, 0300 123 2040, for an active attack in the UK.

Where does a scan fit?

After the incident, an outside view helps you see what your site shows to everyone. The Red free scan reads the stack your site serves and matches each version to known vulnerabilities. One scan per domain, no card. It is a lookup: it can say a version is affected, not whether your setup is exposed, and it cannot see your logs or what an attacker reached. See the limits of a lookup-only scan. Domain ownership is proved with one DNS TXT record before anything runs. The full scan costs $100 per scan and is ordered from your dashboard once the free scan is done. Neither gives you a "confirmed" for the email. That comes from your logs and your review.

Questions people ask

How soon do I have to tell customers my website was breached?

It depends on the law that covers your customers. Under the GDPR the authority is told within 72 hours where feasible, and individuals without undue delay when the risk to them is high. Some US states set a fixed limit, such as 30 days in New York and Florida. Ask a lawyer which laws apply, and send a first factual note early.

What should a breach notification email say?

What happened, which kinds of data were involved, when, what you have done, what the customer should do, and who to contact. The FTC also says to describe what you know and avoid speculation. Add the date of your next update so the email does not read as the last word.

Do I have to email customers if I am not sure their data was taken?

If the evidence says personal data may have been involved, send a short note saying so and say what you do not yet know. Silence while customers read the news is the worse option. If you hold no sign that customer data was reachable, you may not owe a notice. Write down why you decided that.

Can I just say we take security seriously and no data was misused?

Avoid it. 'No evidence of misuse' is only true if you looked, and customers read it as a dodge. Say which logs and data you reviewed, for which dates, and what you found. If the review is not finished, say that.

My agency or host had the breach, not me. Who emails the customers?

Usually the business that owns the customer relationship, which the GDPR calls the controller. A processor must tell the controller without undue delay. In Sweden's IMY guidance the controller stays responsible for reporting even when a processor had the breach. Your contract and a lawyer settle who sends what.

Do I need to email customers about a breach that only defaced my site?

Only if personal data was reachable from what the attacker got to. A defaced page, a planted script or a hidden admin account can all sit next to a customer database. Find out what that account or file could read before you decide the notice is not needed.

Sources

  1. FTC: Data Breach Response, A Guide for BusinessWhat a notice should contain, describing what you know instead of speculating, evidence preservation, verifying a service provider's fix. Read 24 October 2026.
  2. Regulation (EU) 2016/679 (GDPR), Articles 33 and 34 (EUR-Lex)72 hours to the authority where feasible, processor to controller without undue delay, clear and plain language for individuals. Article text read via a reproduction on 24 October 2026.
  3. ICO: personal data breaches, a guideReport within 72 hours where feasible; tell individuals when high risk is likely; what the message must contain. Read 24 October 2026.
  4. ICO: telling the ICO and individuals about a breachReport early and update later; individuals must be told without undue delay when risk is high. Read 24 October 2026.
  5. ICO: personal data breach examplesA worked case where credential theft exposed customer data and both the ICO and clients were told. Read 24 October 2026.
  6. IMY (Sweden): personal data breaches72-hour rule; the controller stays responsible for reporting even when a processor had the breach. Read 24 October 2026.
  7. New York General Business Law section 899-aaNotice to residents within thirty days of discovery; required contents; notice to state agencies. Read 24 October 2026.
  8. Florida Statutes section 501.171Notice no later than 30 days after determination; a third-party agent tells the covered entity within 10 days. Read 24 October 2026.
  9. California Attorney General: data security breach reportingA sample notice goes to the Attorney General when more than 500 California residents are notified. Read 24 October 2026.
  10. CISA: StopRansomware guide, response checklistWho to engage and report to; take system images and preserve logs before rebuilding. Read 24 October 2026.
  11. UK NCSC: small business guidance, response and recoveryDefines an incident and points to the response and recovery guide; reporting number for active attacks. Read 24 October 2026.
  12. UK NCSC: incident managementA prepared response limits public and media fallout. Read 24 October 2026.

Facts on this page are dated and point at the primary source next to them. If a vendor status changes after the date shown, the vendor's page wins; tell us at support@run.cloud and we will update it.