Fractional Executive Bench, Advisory & LeadershipPlaybook3 min readUpdated September 2026

The First 72 Hours After You Discover a Data Breach

The decisions that matter most after a data breach happen in the first few hours, before anyone outside the company knows anything happened, and long before the full scope is understood.

Move too fast on communication and you commit to a scope you don't actually know yet. Move too slow on containment and the incident gets worse while everyone debates what to say about it.

Vendors Covered in this Article

Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.

Hour zero to hour four: contain before you communicate

Isolate the affected systems and preserve logs before anyone touches a keyboard to reimage or patch. A well-intentioned engineer wiping a compromised server to stop the bleeding can also destroy the evidence that forensics needs to figure out what actually happened.

Resist any pressure to make a public statement in this window. Nobody outside the incident response group has enough information yet to say anything accurate, and an early statement that later needs correcting does more damage than a short delay.

Hour four to hour twenty-four: figuring out what you still don't know

Scope the incident before you scope the message: what systems were touched, what data lived on them, and how many people are potentially affected. If the internal team hasn't handled an incident like this before, bring in an outside forensics firm now rather than after a few more days of internal guesswork.

Write down what's confirmed versus what's suspected, and keep the two separate in every internal update. Suspicion that gets treated as fact in an internal memo has a way of becoming the version that leaks.

Once containment is under way, the next stretch of work runs in this order:

  1. Scope the incident: which systems were touched, what data lived on them, and how many people may be affected.
  2. Bring in an outside forensics firm if the internal team hasn't handled an incident like this before.
  3. Record what is confirmed separately from what is only suspected, and update both lists as facts arrive.
  4. Engage breach counsel and let them own the notification timeline once the scope is confirmed.

Why the notification clock is a question for counsel, not a calendar

Legal notification obligations vary by state, by the type of data exposed, and by whether other regulations apply, so don't guess at a deadline internally and work backward from it. Bring in breach counsel immediately, ideally counsel your outside general counsel already has a relationship with, and let them own the notification timeline once scope is confirmed.

Who talks to whom while this is unfolding

Name a single point of contact, usually the CEO or general counsel, before the incident happens, not during it. Every executive independently texting customers or fielding press calls with their own version of events is how a contained incident turns into a confused, contradictory public story.

Prepare a short holding statement in advance that says the company is aware, is investigating, and will share more once confirmed. It buys the time the investigation actually needs without looking evasive.

After the 72 hours: the training gap most post-mortems find

Almost every post-mortem finds the same thing: the plan existed somewhere, but the people who needed it hadn't actually read it. Turn what went wrong into an actual, numbered SOP in a tool like Trainual so the next incident runs off a documented playbook instead of institutional memory.

If your company already tracks policy and incident history in Vanta, log the response there too. Customers and auditors who ask about this incident later should find a documented process, not a gap.

Testing the plan before you actually need it

A written plan that's never been rehearsed usually falls apart on the first real detail: nobody knows which vendor has the forensics contract, or who actually has the authority to take a system offline at 2 a.m. Run a short tabletop exercise once or twice a year, a couple of hours where the incident commander and the rest of the response group walk through a made-up scenario out loud.

The exercise doesn't need to be elaborate to be useful. The goal is just to find the gaps, the contact who left the company eight months ago and was never replaced on the list, the step that assumes access to a system that would itself be down in a real incident, before a real breach finds them for you.

What the executive team specifically owes the response, not the engineers

The technical containment work belongs with whoever runs infrastructure and security, not the CEO. What the executive team owes the response is faster: deciding on the notification timeline once counsel has advised on it, approving the customer-facing statement quickly instead of workshopping the wording for a full day, and making the call on anything that needs board involvement.

The most common executive misstep isn't being too slow on any one of those, it's trying to also run the technical investigation personally, which pulls attention away from the decisions only an executive can actually make and slows down the parts of the response that were moving fine without them.

Executive Capability Standard

What Good Looks Like

A real incident response plan tells someone exactly what to do in the first hour without improvising, and it's been read by the people who'd actually use it.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Read through whatever incident response plan already exists, or admit there isn't one, and identify who's supposed to be the point of contact.
2. Do Manually:Draft a one-page response checklist yourself: who to call first, what to preserve, who talks to customers, kept somewhere accessible outside the systems that might be down.
3. Delegate:Assign a single incident commander role in advance, not during the incident, so no one is negotiating over who's in charge while systems are still compromised.
4. Automate:Set up log preservation and alerting ahead of time so evidence doesn't get overwritten automatically before anyone notices the incident.
5. Buy:Retain breach counsel and a forensics firm on retainer before an incident happens, not while the notification clock is already running.

How to Get Started

Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.

Frequently Asked Questions

Should we notify customers before we know the full scope?

Generally no. Wait for enough clarity to be accurate, but don't wait so long it looks like a cover-up. Breach counsel should guide the exact timing since it interacts directly with legal notification deadlines that vary by jurisdiction.

Who should be the single point of contact during a breach?

Usually the CEO or general counsel, decided before an incident ever happens, not whichever executive happens to be most available that day. Consistency in who's speaking matters more under pressure than it does in normal operations.

Do we need to hire an outside forensics firm?

For anything beyond a small, fully contained single-system incident, yes. Outside forensics protects the integrity of the investigation and often the legal privilege around it in a way an internal team, however capable, usually can't replicate on its own.

What's the biggest mistake companies make in the first day?

Talking to press or customers before containment is confirmed, or reimaging affected systems before evidence is preserved. Both feel like decisive action in the moment and both can make the actual investigation significantly harder afterward.

About the numbers

This guide doesn't quote a sourced benchmark. Figures in it are estimates or general guidance, so check them against your own numbers.

Related Guides