In 2025, only 4% of surveyed organizations reached the Mature stage of cybersecurity readiness in Cisco's Cybersecurity Readiness Index, which drew on responses from 8,000 private-sector business leaders across 30 markets. That gap isn't a software shortage. It's a shortage of written decisions.
Ask an owner what happens if email goes down for two days and the answer is usually a shrug and a call to the IT provider. Ask instead who has authority to shut the email system off, how long billing can run on paper, and how much recent invoice data the business can afford to lose, and the quality of the answer tells you whether a plan exists at all.
A cyber security plan earns its place by two measures. How fast the business gets back to work, and how fast the right person makes a decision. Tools, policies, and training slides exist to serve those two numbers. A usable plan is small enough to review in one leadership meeting and specific enough to open before lunch on the worst day of the year.
Table of Contents
- The Five Parts Every Cyber Security Plan Needs
- How to Inventory the Assets That Keep You Open for Business
- How to Rank Risks by Business Impact
- Choosing Safeguards and Measuring Whether They Work
- Assigning Owners and Decision Rights Before an Incident
- Writing a Response Procedure for the First 24 Hours
- Scheduling Your First Tabletop Exercise This Month
The Five Parts Every Cyber Security Plan Needs
A cyber security plan is a short written document with five parts, and each part answers one question a business leader can act on without a translator.
| Component | Question it answers | What to write down | Suggested owner |
|---|---|---|---|
| Assets | What do we run that the business cannot operate without? | Systems, data, accounts, and vendors ranked by revenue impact | Operations lead |
| Risks | What could stop or expose us? | Likely threats ranked by business impact | Executive sponsor |
| Safeguards | Which controls reduce each risk? | MFA, tested backups, patching, staff training, vendor access rules | IT lead or outside advisor |
| Responsibilities | Who decides and who acts? | Named roles, a backup for each role, and decision rights | CEO or managing partner |
| Response procedures | What happens in the first hours? | Escalation steps, contact lists, communication templates, recovery order | Named incident lead |
None of this is original. It maps closely to how the National Institute of Standards and Technology's Cybersecurity Framework 2.0 organizes the same work. NIST published the update on February 26, 2024, the first major revision since the framework's creation in 2014, and it widened the framework beyond critical infrastructure such as hospitals and power plants to organizations of every size, sector, and maturity.
The structure now runs across six functions: Govern, Identify, Protect, Detect, Respond, and Recover. The added Govern function matters most to owners because it connects security decisions to leadership accountability, legal obligations, supply-chain exposure, finance, and reputation. NIST describes the framework as "a taxonomy of high-level outcomes" rather than a prescriptive list of products or controls, which is the permission a growing business needs. Pick the safeguards that fit your risks and your budget, and write down why you picked them.
The distinction that matters: a plan is a management document, not a product list. The document names what the business protects, who decides, and how fast it recovers. The product list is an appendix that changes every year.
Length beats completeness. A plan nobody opens protects nothing. The target is a one-page outline with an owner named against each of the five parts, plus a short page behind each one.
How to Inventory the Assets That Keep You Open for Business
Start with processes, not hardware. A server inventory tells you what you own. A process inventory tells you what a client pays for, and therefore what has to come back first.
The steps below run in order, and each one exists to answer a question you'll be asked during an incident.

- List every process that brings in revenue or delivers client work. Quoting, billing, project delivery, patient scheduling, production orders. Processes are what clients notice when they stop, so they set the order of everything after this step.
- Name the systems and data each process depends on. Email, file storage, accounting, scheduling software, cloud tools. If a process touches three systems, list all three, because the weakest one sets your real recovery time.
- Mark sensitive or regulated data. Client files, patient records, payment details, employee records. The marking tells you which systems create notification duties, legal review, and insurer conversations, not just downtime cost.
- Estimate maximum tolerable downtime for each process. Write the number of hours or days before clients call, deadlines slip, or revenue stops. This is a business judgment, and it's the single most useful paragraph in the plan.
- Set a recovery time objective and a recovery point objective for each system. Recovery time is how fast a system must return. Recovery point is how much recent data you can afford to lose. Leaders set both, because they encode cash flow, client commitments, and contract terms rather than technical preference.
- Record the vendors and suppliers with access to your systems or data. The payroll provider, the billing clearinghouse, the print vendor with a remote login. Access you've forgotten is access you can't revoke during an incident.
- Give every asset a business owner by name. A person, not a department. When the owner is a department, three people assume someone else reported the outage.
The mistake that appears in nearly every inventory is a list of servers and laptops with no mention of cloud accounts, software subscriptions, and vendor logins. Those are usually where the data lives and where an attacker finds the least-watched door.
Finish with a ranked asset list the leadership team can read in one meeting, ordered by revenue impact, with downtime tolerance beside each line.
How to Rank Risks by Business Impact
Verizon's 2025 breach summary analyzed 22,052 real-world security incidents, including 12,195 confirmed data breaches, across 139 countries. The same summary attributes 22% of breaches to credential abuse and 20% to vulnerability exploitation, puts third-party involvement at 30%, double the earlier figure, and reports exploitation of vulnerabilities up 34%.
Below 1,000 employees the pattern is sharper. The small and midsized business snapshot counts 3,049 incidents, of which 2,842 involved confirmed data disclosure. In that slice, ransomware appeared in 88% of breaches against 39% at larger organizations.
Alarm is a bad sort order. Rank by hours the business stays down, and by how fast someone with authority can decide to restore, shut a process off, or call the insurer.
| Threat | Verizon 2025 figure | Business impact | First safeguard |
|---|---|---|---|
| Ransomware | 88% of SMB breaches vs. 39% at larger organizations | Operations stop until systems are restored | Isolated backups and a timed restore test |
| Credential abuse | 22% of all breaches | Account takeover, fraudulent payments, data exposure | MFA on email, cloud, and remote access |
| Vulnerability exploitation | 20% of breaches; exploitation up 34% | Intrusion through unpatched systems | Fast patching of internet-facing systems |
| Third-party involvement | 30% of breaches, doubled from prior reporting | Vendor access exposes your data or systems | Vendor access review and contract terms |
Turn the table into your own ranking
Score each threat twice, on a three-point scale. Likelihood is low, medium, or high for a business like yours, in your industry, with your current controls. Impact is low, medium, or high against the asset list, measured in downtime hours, cash out the door, or clients you lose. Compare the two scores. Do not average them. A medium-likelihood threat that stops billing for two days outranks a high-likelihood nuisance that costs an afternoon.
Tie each threat to a named asset. A phishing email in the accounting inbox can freeze payments the same day. The same email in a design archive nobody outside the studio can open is a cleanup task, not a stop-the-business event.
Keep the final list to five to seven risks. Longer lists get read once and abandoned. A leader can hold seven items during a bad week. Put that short list on one page for a single leadership meeting: threat, hours you can tolerate down, who decides, and the response.
For each risk, write one response. Reduce it with a safeguard, transfer part of the cost through insurance, accept it with a stated reason, or avoid it by changing the process. The written reason is what you can show a client, an insurer, or a regulator later. I would rather see "we accept this because restore takes four hours and the revenue at stake is small" than a blank cell.
Two pitfalls sink the exercise. Treating risk as a purely technical question hands the ranking to whoever is most comfortable with firewalls, and downtime cost never gets a vote. Ranking by fear puts a rare nation-state intrusion above credential theft, which shows up in about one breach in five and usually moves faster than any exotic attack.
Choosing Safeguards and Measuring Whether They Work
Buy safeguards in the order losses arrive, not the order vendors call. Four layers cover most of the exposure at a small or midsized business, and each layer needs a number that proves it's working.

- Multifactor authentication on the accounts that matter. MFA requires a password plus a second proof, such as an authenticator app or a security key. CISA's guidance for small and midsized businesses treats MFA as a baseline control and recommends prioritizing privileged accounts, administrative access, and remote access. Write down which systems require it, who owns enforcement, how exceptions get approved, and how access is removed on a departure.
- Same-day removal of departed accounts. The offboarding clock is a measurable security control. A shared password that outlives an employee by three weeks is an open door with a name on it, and the fix costs nothing but a checklist and an owner.
- Patching, email filtering, and short phishing training. Internet-facing systems get the shortest patch window, because that's where automated scanning finds you. Training works when it runs fifteen minutes a quarter with real examples from your own inbox, not once a year in a compliance video.
- Backups built to your numbers. CISA ties backup design to the recovery point objective and recovery time objective on your asset list, and recommends scheduled recovery tests that confirm full and partial restores work. A copy that lives only on the same network an attacker controls isn't a backup.
On isolation, CISA's cross-sector performance goals recommend backups that automatically and continuously protect critical data and system configurations, with a copy kept in a location isolated from the business network. That separation is what stops one compromised administrator account from encrypting every recovery copy at once. Document the installation media, license keys, configuration details, and retention periods too, because a restore stalls when the only person who knew the license key is on vacation.
The order matters more than the tools. Buying endpoint protection before agreeing on how much downtime a project deadline tolerates is spending money on a question nobody asked. Recovery time is the yardstick: every safeguard above either shortens the hours between incident and reopening, or it doesn't earn its place.
Whether any of it works is a separate question. Mastercard's 2025 survey of more than 5,000 SMB owners across four continents found that 86% had conducted a risk assessment and created a prevention plan, yet only 23% were very satisfied with that plan and only 23% were very confident in their ability to identify threats. Getting employees to take cybersecurity seriously was challenging for 73%, and only about one quarter were very confident in their ability to educate staff. Buying the controls and believing the plan are clearly two different achievements.
Five measures worth a quarterly review
- MFA coverage, as the percentage of critical accounts using phishing-resistant or app-based MFA rather than SMS codes.
- Time to disable a departed employee's access, measured in hours, with a target you'd defend to a client.
- Restore time, from the last real test, not from the vendor's marketing page.
- Incident exercises completed, as a count against the number on the calendar.
- High-risk findings closed by their due date, which is the only number that shows whether anything changes.
Each one converts into money and speed. MFA coverage predicts fraudulent payment risk, restore time predicts lost billable hours, and closure rate predicts whether next year's insurance renewal gets easier or harder. Put all five on a single page and a leadership team can read them, argue about them, and assign fixes inside one meeting. That page is what separates a plan that gets reviewed from a binder that gets dusted off after something breaks.
Assigning Owners and Decision Rights Before an Incident
Responsibilities fail in the gap between assumed and written. Everyone assumes the IT provider handles "the technical part," and nobody knows who authorizes taking email offline at 6 a.m. on a Saturday.
The Cybersecurity and Infrastructure Security Agency's small business guidance recommends a written incident response plan that assigns responsibilities before, during, and after an incident, covers realistic scenarios such as ransomware, includes business leadership and legal counsel rather than technical staff alone, and is reviewed and exercised at least annually. CISA describes it as an action plan for the CEO, which is the correct frame. This is business continuity, not an IT document.
| Role | Decides | Must know | Backup |
|---|---|---|---|
| CEO or managing partner | Taking systems offline; approving recovery spending | Business priorities and cash position | Operations lead |
| Operations lead | Order of system recovery | Client deadlines and production schedules | Office manager |
| IT lead or outside provider | Containment, account lockouts, evidence preservation | Admin accounts, backup locations, network map | Second technical contact |
| Legal counsel | Notices to customers, regulators, and insurers | Contracts and regulated-data duties | Outside firm on retainer |
| Finance lead | Unplanned spending and insurer reporting | Policy terms and payment approvals | CEO |
| Communications owner | Messages to staff, clients, and the public | Approved templates and key facts | Operations lead |
Write each decision right as one line a stranger could follow. Who can shut down the email system without asking. Who can spend up to a set limit on recovery without a second signature. Who calls the insurer. Ambiguity here is expensive, because the first hour of an incident is spent determining roles while operations are already down.
Practical rule: name a backup for every role, then keep the contact list offline, printed, and out of the systems that might be unavailable. A plan that lives only in the file server you can't reach isn't a plan.
Writing a Response Procedure for the First 24 Hours
The first 24 hours decide whether an incident is a bad week or a lost quarter. Recovery time is set by how fast people can make three calls: isolate, inform, restore. Write the procedure as six steps a tired manager can run at 2 a.m., with the reason each step exists.

- Declare the incident and name the lead. One person coordinates, and everyone else reports to them. This prevents two people chasing the same server while the insurer call waits.
- Isolate affected devices and disable compromised accounts. Pull the network cable, disable the mailbox, revoke active sessions. Rules written in advance turn containment into a two-minute action instead of a debate.
- Preserve logs and evidence before rebuilding anything. Capture error screens, mailbox audit logs, and the ransom note itself. Rebuilding first destroys the record your insurer, counsel, and forensics team will ask for.
- Confirm what is down against the recovery order. Compare reality to the ranked asset list. Restore in the order clients are waiting on, not in the order complaints arrive.
- Make the calls that carry deadlines. Counsel checks disclosure duties by state and data type, the insurer gets notice inside the policy window, and staff receive one approved paragraph so personal social accounts do not become your news channel.
- Start recovery from clean backups and time every milestone. Detection, containment, restore, and first client notice are the numbers that prove control and become the next plan revision.
The UK government's 2025 cyber security breaches survey found that smaller organizations struggle to build incident-response plans because of limited expertise and capacity, often test them only after a breach, and remain uncertain about the exact steps to take when an incident occurs. Plans fail less from missing controls than from slow handoffs: no named lead, no spending approval, no one sure whether counsel or the insurer comes first.
A short plan that has been walked through once beats a long manual nobody has read. Length feels like rigor. Rehearsal is rigor.
Keep the six steps to one page beside the offline contact sheet, and leadership can review the response path in a single meeting before anyone signs off.
Scheduling Your First Tabletop Exercise This Month
A plan is a hypothesis until someone tests it. Put a 90-minute tabletop on the calendar this month. A tabletop is a discussion-based walkthrough of one realistic scenario, no technology involved, and it exposes more gaps per hour than any tool purchase.
Invite the CEO or managing partner, the operations lead, the technical contact, the finance lead, and legal counsel. Run a single scenario: ransomware on the file server that holds project files. Walk the group from first alert through client communication, and make every participant answer in their actual role, not in theory.
Time the decisions, not just the steps. How long until someone declares an incident. How long until a person with spending authority approves containment work. How long until the group agrees who calls the first client. A tabletop measures decision speed, and decision speed is what recovery time is made of. When the exercise stalls, write down exactly where: no offline contact list, no named backup approver, a backup nobody has ever restored.
Then update the one-page outline with named owners, backup decision-makers, and a dated list of fixes. Review the full plan once a year and after any material change, such as a new software platform, a new office, an acquisition, or a vendor that suddenly has administrator access.
Where outside support would help, Nextorum provides vCISO leadership and cybersecurity planning for businesses across Dallas, Fort Worth, Carrollton, and McKinney. If your leadership team can't answer who shuts down email, how long billing can run on paper, and how much recent data you can afford to lose, Nextorum can schedule a working session that puts those answers on one page, with owners, decision rights, and recovery targets named.
