An incident response plan for cyber attack situations names who decides in the first hour, what systems to isolate, and how to communicate with customers and regulators. Of the 25 experts we asked, 9 emphasized naming a decision-maker, and the same number stressed keeping the plan on paper, offline, because your network may be compromised.

Key takeaways

  • Name one person with authority to make decisions and isolate systems without waiting for a meeting.
  • Print the plan and keep it offline with phone numbers for your IT provider, insurer, bank, and lawyer.
  • List every system holding customer or payment data, rank them by risk, and test your backups by actually restoring them.
  • Write down who tells customers, staff, and regulators what happened, and by when—many states have legal deadlines.
  • Run a tabletop exercise twice a year with a realistic scenario so your team knows what to do before an attack.
  • Map every third-party app and integration with access to your systems, and know who can revoke each one in the first hour.
What the experts said most often
Name a decision-maker with authority to act9
Keep the plan printed and offline9
Include contacts and first-hour steps on one page8
Isolate affected systems without destroying evidence5
Define who communicates with customers and regulators3
Run tabletop exercises to find gaps3
Map data locations and third-party access3
Revoke access and change passwords from clean device2

Number of experts (out of 25) who made each point. Many made more than one.

We asked experts what small businesses must include in an incident response plan for cyber attacks. Their answers cluster around five areas: who leads the response and makes decisions fast; what to protect and restore first; how to reach each other and outside help when systems are down; what customers and regulators must be told and by when; and how to test the plan before a real attack arrives.

1. Who decides and acts in the first hour

The costliest delays come from nobody knowing who can make decisions when systems are down. Experts agree on naming one person—with a backup—who has authority to isolate systems, spend money on response, and contact your bank and insurer without waiting for a meeting or committee approval. Research on how to prepare for incident response shows that organizations with named decision-makers and pre-agreed procedures reduce response time significantly. Protecting your premises matters too, and our earlier expert roundup on physical security services for small business covers that side of the picture.

Richard Nicholas, Technology & AI Lawyer, Browne Jacobson LLP

“One person's name, the person who will make the decisions when they are asked in the middle of the night or over the weekend and who will take those decisions quickly. Some won't have easy answers or all of the information (do we report X, do we assume Y is affected?, what do we tell customers and staff?) so that person will have authority to decide quickly. The name of the forensics company that can find the answers to "what has been affected" (or an insurer). You want this contract in place before the attack. A lawyer who is on your side and who you can speak to in confidence. You might also need to speak to a regulator who has the power to hand out fines.”

He also emphasizes keeping a contract with a forensics company and a lawyer in place before an attack occurs.

Richard Nicholas is a technology and AI lawyer at Browne Jacobson LLP, specializing in incident response and regulatory matters.

John R Dennison, Attorney & AI Infrastructure Analyst

“In the first hours of an attack, the costliest delays come from nobody knowing who is authorized to act. The plan should name, before anything happens: who can take systems offline and spend money on response; who contacts the attorney, the insurer, and law enforcement; and who communicates with customers. One concrete tip: write down, on a single page kept offline, the names and mobile numbers of your decision-makers, your insurer's breach hotline, and outside counsel.”

He stresses that the most costly delays in the first hours come from nobody knowing who is authorized to act.

John R Dennison is an attorney and AI infrastructure analyst who advises small businesses on incident response (JohnDennison.com).

Chandim Chatterjee, Engineering Manager / Tech Lead, Meta

“For a small business, an incident response plan is less a document than a rehearsed set of decisions: who declares an incident, who can pull the plug, and who talks to customers. Start with that — a named incident lead, a backup, and authority to isolate a system without waiting for a meeting.”

He frames the plan as a rehearsed set of decisions rather than a static document.

Chandim Chatterjee is an engineering manager and tech lead at Meta with experience in incident response at scale.

Roland Jakob, Founder, Praxica

“And the biggest weakness in a small business's incident response plan often isn't the technology. I'd build the plan around five essentials: who has authority to declare an incident and disconnect compromised systems, how to contain the damage without destroying evidence, who contacts IT support, legal counsel and insurers, which systems get restored first, and how customers and regulators are notified when required. If your team needs a meeting to decide who can disconnect a compromised system, you've already given the attacker a head start. Your incident response plan should eliminate ambiguity, not create more of it.”

He warns that the biggest weakness in small business plans is often decision-making, not technology.

Roland Jakob is founder of Praxica, a consultancy that helps small businesses build incident response capability.

Rishabh Pande, Independent Researcher

“The mistake small businesses make is rehearsing the breach call but not the freeze decision: who can halt payments, revoke access, and notify the bank in the first hour, by name, with authority already granted.”

In fraud incidents, the first hour is paperwork and permission, not technology.

Rishabh Pande is an independent researcher focused on fraud detection and incident response in small organizations.

Michael Olson, Senior Strategic Buyer; Founder, Guardrails

“A small business does not need an enterprise plan. It needs a one-page plan that names who does what in the first hour. One: a call list — owner, IT help, bank, insurer, and the person who can approve a shutdown. Three: the first-hour steps — disconnect the affected machine, preserve logs and messages, change passwords from a clean device.”

He recommends a simple one-page plan that names who does what in the first hour with concrete steps to disconnect and preserve evidence.

Michael Olson is founder of Guardrails, a consultancy for small business incident response and recovery.

Tim Lytle, Owner, Lytle Labs

Tim Lytle, Owner, Lytle Labs

“The most important part of a small business's incident response plan is naming one person who has the authority to make decisions without waiting for a meeting. Give that person a backup in case they're unavailable. Everything else in the plan should help them do that job.”

Everything else in the plan should support that one person's ability to act without waiting.

Tim Lytle is owner of Lytle Labs, a software development company that serves small and mid-size clients.

2. Keep the plan offline and on paper

If your incident response plan lives only on your network or in the cloud, it will be unreachable the moment your systems are compromised. Experts consistently recommend printing the plan and storing copies offline—in a truck, a desk drawer, or at a business owner's home—with phone numbers for all key contacts.

Ben Walker, Founder and CEO, Ditto Transcripts

“Names and phone numbers. Write down who leads the response and how to reach your IT help, your insurer, and your lawyer. Keep a printed copy in case your email goes offline.”

He handles sensitive court, police, and patient recordings and updates his plan every six months to meet CJIS requirements.

Ben Walker is founder and CEO of Ditto Transcripts, which manages sensitive legal and law-enforcement recordings and must comply with strict security standards.

Jack Butler, Owner, Utah Asphalt Sealcoating

“My one concrete tip: print the plan and keep a copy off the computer. If your email is locked, you can't open the plan that's sitting in your email. A paper copy in the truck and one at the shop means you can still act in the first hour, when it matters most.”

He keeps copies in multiple physical locations so the team can act even if their primary office systems are offline.

Jack Butler is owner of Utah Asphalt Sealcoating, which manages operations across multiple job sites.

Stacy Derrick, Founder & Business Coach, Stacy Derrick Creative, LLC

“My one tip you can act on this week: print the plan. Keep a paper copy, with phone numbers, somewhere you can reach without a computer. If an attack locks you out of email and your files, a plan saved in the cloud won't help you.”

She emphasizes that small businesses without IT departments must be able to act without computer access.

Stacy Derrick is founder and business coach at Stacy Derrick Creative, LLC, working with small business owners on operations and resilience.

Santhosh Konduru, Independent AI researcher; Senior Member, IEEE

“One practical tip: keep a printed copy of the plan and key contacts somewhere off your network. If ransomware locks your systems, you won't be able to access a plan stored on those same systems.”

If ransomware locks your systems, a plan stored on those systems cannot be accessed.

Santhosh Konduru is an independent AI researcher and senior member of IEEE focused on incident response and business continuity (santhoshk6724.netlify.app).

Kristopher Clark, Solutions Architect, Independent practice

Kristopher Clark, Solutions Architect, Independent practice

“My concrete tip is to keep a one-page call tree and account shutdown list outside the company network. Print it and keep a copy with the owner. If email or single sign-on is compromised, the team can still reach each other and start cutting off access.”

A one-page call tree and account shutdown list kept outside the network allows the team to reach each other and cut off access even when email and single sign-on are compromised.

Kristopher Clark is a solutions architect with infrastructure experience across risk-averse organizations (Independent practice).

Teila Garraway, Legal Operations & AI Technology Implementation Consultant, SourcePro Consultants

“The most useful thing a small business can add to an incident response plan is a one-page decision sheet that names a single owner for each step and lists who has authority to call the insurer, outside counsel, IT provider and affected customers. Keep a printed copy offline, because during an attack your email and shared drives may be exactly what you cannot reach.”

During an attack, email and shared drives may be exactly what you cannot reach.

Teila Garraway is a legal operations and AI technology consultant at SourcePro Consultants, working with law firms on incident response.

Boris Kreiman, Founder and CEO, Linkbaits

“Two things separate plans that work from plans that sit in a drawer: a printed copy (the plan that lives only on the server is unreachable in the one moment you need it), and an annual 30-minute walk-through so everyone has actually read it. One concrete tip: write down your bank's fraud hotline and your cyber insurer's breach line today and tape them next to the plan. In a ransomware event, minutes matter, and nobody should be searching for phone numbers while files encrypt.”

A printed copy and an annual walk-through separate plans that work from plans that sit in a drawer.

Boris Kreiman is founder and CEO of Linkbaits, a digital marketing consultancy that serves small and mid-market businesses.

Brad Margist, Co-founder, CyberHoot

“They need one page everyone can find at 2 a.m. Put five things on it: who is in charge and who they call first (your IT provider, your cyber insurer's claims line, a lawyer); how you will reach each other if email is down, with personal phone numbers printed; the first moves, which are to unplug the affected computer from the network without turning it off or wiping it, and to change passwords from a clean device; where your offline backups are and who has tested a restore; and who tells customers, and by when, since many states set notification deadlines.”

Put five things on one page that everyone can find at 2 a.m., with personal phone numbers printed.

Brad Margist is co-founder of CyberHoot, a small business cybersecurity training and incident response company.

3. List critical systems, map data, and test backups

Before an attack, you must know what you have, where your customer data lives, and whether your backups actually work. Experts recommend a simple inventory of critical systems ranked by impact, a map of all data locations and third-party access, and testing restores—not just trusting that backups exist. anchor text:Learn how third-party access and physical security contribute to your overall business security posture.

Chris Cobb, Founder, Sannidhi

“The first thing a small business needs is a current map of where its customer data lives. Most owners cannot say within an hour which laptops, cloud accounts, shared logins and vendor portals hold client information, and an attacker counts on that gap. Write down every system, its owner, and how to cut its access in one place. They knew where everything was before the emergency started.”

Most owners cannot say within an hour which laptops, cloud accounts, and vendor portals hold client information.

Chris Cobb is founder of Sannidhi, with 30 years building technology for large organizations.

Jurgis Plikaitis, CEO, ClearNym

“The last thing, and this is crucial: make a list of the data you have and where it is stored. Since the rules for breach notification differ in various states, time will pass quickly while you try simply to find out what was stored on that lost disk. Just one piece of paper, that's enough. List every place where you have client and employee information and write down who has access to this data.”

Create a data map on one piece of paper listing every place where you have client and employee information.

Jurgis Plikaitis is CEO of ClearNym, a consultancy for small business data security and incident response.

Luke Sandelands, Founder, Stack Architect

“The plans that I come across assume the attack arrives at the company itself, when in my experience the more likely way in is a third party somebody connected two years ago that nobody has thought about since. So the element I would put first is a live inventory of every integration holding a token to your systems, with a named person who can revoke each one, because during an incident the question that burns the most time is who actually has the permissions to switch something off. Third, test the restore rather than trusting the backup, since plenty of the businesses I work with have backups running that nobody has ever attempted to restore from. The step I'd advise to your readers is to list every app with API access to their email, their storefront and their payment processor, then remove anything they cannot give a current reason for.”

Many breaches come through third-party integrations; maintain a live inventory of every integration holding a token to your systems with a named person who can revoke each one.

Luke Sandelands is founder of Stack Architect, which helps small to mid-size businesses inventory and secure third-party access.

4. Define communication and notification responsibilities

Who tells customers, staff, and regulators what happened, and by when? Experts stress that unclear authority here creates delays and legal exposure. Many states have notification deadlines, and contracts may impose shorter ones. Write down who is authorized to say what to whom before an attack happens. Professional organizations provide guidance on testing incident response plans through tabletop exercises and simulations.

Kira Klein, Salesforce Software Engineer, ANZ Bank New Zealand Limited

“Teams should also define how they will communicate with customers, vendors, regulators, and employees if services are disrupted.”

Teams should define how they will communicate with customers, vendors, regulators, and employees if services are disrupted.

Kira Klein is a Salesforce software engineer at ANZ Bank New Zealand Limited, which manages large-scale transaction systems requiring incident response discipline.

Carlos Paredes, Founder, NosisTech LLC

Carlos Paredes, Founder, NosisTech LLC

“Write down who decides before you need to decide. Many small businesses have backups and antivirus, but nobody named to make the calls during an attack. It names an incident lead and a backup person. It lists who to call: your IT provider, cyber insurer, bank and lawyer.”

Name an incident lead and list who to call, then say who tells staff how to report strange emails or locked screens without fear of blame.

Carlos Paredes is founder of NosisTech LLC, a small business security consultancy.

5. Run tabletop exercises to find gaps before an attack

Paper plans often hide critical gaps. Experts recommend running a tabletop exercise twice a year with a realistic scenario—ransomware on a Friday afternoon—so the team actually finds missing phone numbers, unclear authority, and untested procedures. These exercises usually uncover issues difficult to identify in writing alone.

Oliver Bodemer, Principal Systems & AI Architect, Founder, e-Softworks, Inc.

“Concrete tip: run a 30-minute tabletop exercise twice a year and fix whatever stalls, usually unclear authority or missing contact details.”

Run a 30-minute tabletop exercise twice a year and fix whatever stalls, usually unclear authority or missing contact details.

Oliver Bodemer is principal systems and AI architect and founder of e-Softworks, Inc., a consulting firm for small business security.

More expert takes

Sarfraz Bilal, Founder, ToolFindLab

“One: stop and verify in the primary source before acting — the live site, the network dashboard, never the agent's word or an email alone. In eight days my AI agent published a guide that duplicated a post from three weeks earlier, reported an ad-network approval that the dashboard showed was a rejection, and carried a "pending" status for a month that was actually a ban. Each time, checking the primary source first stopped a bad decision from compounding. Two: one named owner per incident — for us that's me — so nothing waits for a committee. The concrete tip: write down, in one line, which primary source proves each claim — dashboard, live site, or inbox — and check that source before you believe it.”

Verify decisions against primary sources—the live site or network dashboard—before acting, then document each decision.

Sarfraz Bilal is founder of ToolFindLab, an AI and automation consultancy for small business operations.

Christian Ashby, Founder, WPops.land

“A small business incident response plan needs five things written down before anything goes wrong: who decides, what gets saved first, how to cut off only the affected part, how customers keep being served, and how you verify you are clean before reopening. My one concrete tip: save the evidence before you touch anything. When two serious PHP flaws became public in October, our first step was to archive the logs for every website we host, 17 minutes after we read the advisory. When one site was targeted that afternoon, we made just that site read-only within 15 minutes while every other site kept taking customer forms, blocked the attacker when they came back, and checked by hand that nothing had been compromised before reopening it.”

Archive logs for evidence before you touch anything; isolate only the affected part while other systems keep serving customers.

Christian Ashby is founder of WPops.land, which manages hosted websites and incident response for many small businesses.

Alan Hutson, Director, isosecurity.co.uk

“A short Incident Response Plan (IRP) guides the organisation before, during and after a suspected or confirmed incident and establishes responsibilities. Ideally, the plan should be aligned to ISO 27001, Annex A controls A.5.24 through A.5.28, covering incident management preparation, assessment, response, learning and collection of evidence. 1. Purpose, Scope and What Constitutes a Cyber Incident - This prevents people from wasting time debating whether something is "serious enough" to invoke the plan. Testing and maintenance - Importantly, the plan must be tested annually – a tabletop exercise is fine.”

An incident response plan should align with ISO 27001 standards and include purpose, escalation paths, roles, containment steps, communication, recovery, evidence logging, and post-incident review.

Alan Hutson is director of isosecurity.co.uk, a small business security consultancy.

Alex Motor, Founder, Obsidian Ridge Labs

“One concrete tip: freeze the output before you let anyone fix it. A small business incident plan should name one person with veto power before anything ships or changes, and require the failure to be traced to its assumption, not its code. Every automated test passed, because the wrong mapping was internally consistent. The fix was a spec file that is the source of truth, and tests that encode the real-world meaning of each code, not just internal consistency.”

Require a person with veto power to approve all changes before they go live, and test against real-world meaning, not just internal consistency.

Alex Motor is founder of Obsidian Ridge Labs, which integrates gaming and payment systems for small businesses.

Frequently asked questions

What is the most important part of an incident response plan?

Naming one person with authority to make decisions and act without waiting for a meeting or committee approval. That person should have a backup and clear authority to isolate systems, spend money on response, and contact your bank, insurer, and lawyer.

Where should I keep my incident response plan so I can access it during an attack?

Print the plan and keep copies offline in multiple locations—at your desk, in a truck, or at home. If you keep it only on your network or in the cloud, it will be unreachable when your systems are compromised. Include phone numbers for all key contacts on the printed copy.

What should I test before an attack happens?

Test your backups by actually restoring a file or database. A backup that has never been restored is a hope, not a plan. Also run a tabletop exercise twice a year where your team walks through a realistic ransomware scenario and identifies gaps in authority, contacts, and procedures.

How do I know who must be notified after a breach?

Many states have legal notification deadlines that apply, and your contracts with customers may require faster notification. Write down in your plan who is authorized to tell customers, regulators, and staff, and by when. Include contact deadlines for your insurer and lawyer.

Should I include third-party apps and integrations in my plan?

Yes. Many breaches come through third-party connections that were set up months or years ago and forgotten. Create an inventory of every app and integration with access to your email, payment systems, or customer data, and name someone who can revoke each one in the first hour.

The bottom line

An effective incident response plan for a cyber attack is simpler than most small businesses think. It needs one person with authority to decide and act, a printed offline copy with phone numbers for your IT provider, insurer, bank, and lawyer, a list of what systems hold customer data and where backups live, clear rules for who tells customers and regulators what, and regular practice—a tabletop exercise twice a year with a realistic scenario. The difference between recovery in hours and chaos for weeks often comes down to whether the team can find the plan and phone numbers when their usual systems are down, and whether they know who is in charge and what to do first. Test your backups, run your drills, and update the plan after each exercise. The plan that survives that hour of practice is the one that will hold when it counts.

Responses were collected through a media query in November 2026 and have been lightly edited for length. This article is general information on incident response planning, not legal, insurance, or cybersecurity advice specific to your business.