A conventional cyber resilience assessment shows what an organization has: plans, controls, owners, backups, and a rating that summarizes them. A document review alone cannot show whether people can use those resources at 2 a.m. with the building systems down. A real incident tests that second claim first.
The quick answer: a good assessment covers critical services, disruption tolerances, decision rights, and recovery evidence across people, process, and technology. A document-based assessment can prove that a plan exists, a control is configured, or an owner is named. Execution under pressure is a separate claim that requires observation.
Reflex Security publishes this article and builds live crisis simulations, one of the instruments discussed below. That commercial interest should be considered when evaluating the argument.
What Is a Cyber Resilience Assessment?
A cyber resilience assessment is a structured review of whether your organization can anticipate, withstand, recover from, and adapt to cyber disruption, judged against a framework's stated outcomes. The output is a rating per outcome plus a list of gaps with owners.
Those four verbs come from NIST SP 800-160 Vol. 2 Rev. 1. Cyber Resilience: What It Is and Why Security Alone Isn’t Enough explains what cyber resilience means in full. The assessment question here is narrower: how should an organization measure it?
The NCSC Cyber Assessment Framework is one widely used cyber resilience assessment framework, and the NCSC released version 4.0 in August 2025. Its 4 objectives and 14 principles are written as outcomes: "what needs to be achieved rather than a checklist of what needs to be done." Assessors judge 41 contributing outcomes as achieved or not achieved, with partially achieved available for some, guided by indicators of good practice.
One caveat. The CAF is primarily designed for UK critical national infrastructure and essential services, and the duties behind it come from regulation such as the NIS Regulations, not from the framework itself. The hypothetical commercial property operator used throughout this article is not assumed to fall within that regulated scope. The article borrows the CAF structure, not its regulatory authority.
Input reviewed | Typical evidence | What it tells you |
Plans and playbooks | Incident response plan, BMS runbook, tenant-notification template | An intended response exists and has an owner on paper |
Controls and recovery | Backup schedules, access-control configuration, restore test logs | The technical means to withstand and recover are in place |
Governance and people | RACI chart, training records, contact lists | Somebody is accountable, and somebody attended training |
Each row shows that a capability is documented, but none demonstrates how it performs during an incident.
What CISOs Need to Know About Their Organization’s Resilience
Strip away the framework vocabulary and boards, insurers, and major clients want answers to six questions. A useful assessment attaches evidence to each answer and states where that evidence ends.
Which services would hurt most to lose? Name services first and systems second. For a property operator that means building access, climate and life-safety systems, tenant communications, and rent collection.
How long can each service be down before harm sets in? A tolerance written in hours forces a different plan from one written as "as soon as possible."
Who decides, and who takes over when that person cannot be reached? Name a deputy with the authority to make the same decisions.
Which third parties sit inside the response? The building systems vendor, the managed security provider, and the company hosting your ERP each hold part of the recovery.
How fast does recovery run in practice? Use the duration of the last real restore or exercise, and set the plan's target beside it.
What will the team do in the first hour, and what evidence proves it?
An assessment answers the first four well, because they are documentation questions. Questions 5 and 6 are about behavior. A document can assert those two answers. Only observation can supply them.
Assessment reports often support the first four answers clearly but leave the last two supported only by implication. The board may nevertheless read the whole report as a single verdict.
Cyber Resilience Assessment Across People, Process & Technology
Take a hypothetical operator: a commercial real estate company managing 30 office and mixed-use buildings. Its building management systems (BMS) run heating, ventilation, cooling, and lighting. A tenant access-control platform issues badges and visitor passes. A property-management ERP holds leases, the rent roll, and service charges. On-site facilities staff, many of them contractors, are the first people to notice when a building starts behaving oddly.
A cyber resilience readiness assessment of that estate covers three layers, and each produces a different kind of evidence.
People | Process | Technology | |
What is assessed | Named incident roles, deputies, training, and who facilities staff call after hours | Incident plan, escalation paths, tenant-notification procedure, third-party contacts | BMS segmentation, access-control resilience, ERP backups, and restore order |
Typical evidence | RACI chart, on-call rota, training completion records | Approved plan, templates, contact tree, last review date | Network diagrams, configuration exports, backup job logs |
Where documents mislead | A named role says nothing about whether the person answers at 2 a.m. or knows the call is theirs | An approved template has a sender on paper and none in practice; contact trees go stale when contractors rotate | A completed backup job shows the data was copied, and says nothing about the order the business needs it back in |
The bottom row shows the difference. Documents describe technology relatively well because an assessor can check configuration against a diagram. They describe people less reliably because judgment under pressure leaves no artifact until it is exercised.
A cyber resilience readiness assessment that stops at documents rates the people layer from the org chart. In a property company, the incident lands on that layer first: the night supervisor who sees the BMS console misbehave, the leasing manager whose tenants call before security does. Both appear in the binder as boxes. Neither has made the first decision with a clock running.
Cyber Resilience Maturity vs. Proven Incident Readiness
A cyber resilience maturity assessment describes how documented, owned, and repeatable the capability appears at the time of review. That has real value: it shows where the program is thin and gives the board a vocabulary for progress. Incident Response Readiness Assessment: How Ready Is Your Team? covers maturity levels and scoring; this article does not repeat them.
Proven readiness describes behavior. Someone watched your team work a realistic incident and recorded what happened: when the first decision came, who made it, and what the team missed. It is a narrower claim than a maturity rating. It is also much harder to inflate.
Take one control in the hypothetical property operator, recorded both ways:
Rated: the BMS alert escalation path is documented, owned by the facilities director, and reviewed every year.
Observed: in an exercise, the alert sat with a contractor for 25 minutes, because nobody had told the contractor the escalation path applied to them.
Both statements can be true, but only the observed result predicts how the response works in practice. Exercises can fail similarly. Much of why most tabletop exercises fail comes down to generic, low-pressure scenarios that never force anyone to make a real decision. Cyber Resilience Metrics: How to Measure What Actually Matters covers how to track either form of evidence across quarters.
The Blind Spots Traditional Resilience Assessments Don’t Expose
Document-based assessments grade artifacts and self-attestation. That method sees what someone wrote down and what people say they do. It misses five failures that only appear once people start moving. The examples below are hypothetical, set in the same property operator.
Decision latency. The binder named the facilities director as incident lead for building systems. In the room, nobody reached the director for 40 minutes, and the BMS vendor's remote session stayed open the whole time.
Escalation hesitation. The binder said any suspected compromise of access control goes up the chain at once. The room watched three badge anomalies arrive and waited for a fourth, because nobody wanted to wake the CISO for a false alarm.
Unclear ownership. The binder held an approved tenant-notification template. Legal and the leasing team then argued for an hour over who sends it and whose name goes at the bottom.
Cross-functional friction. The binder showed IT and facilities as one response team. The room ran two bridges, and neither knew the other had isolated the same network segment.
Dependency surprises. The binder said the ERP backup is tested quarterly, and a cyber resilience evaluation built on self-attestation scored that control as achieved. Nobody in the room knew whether to restore the rent roll or the access-control data first, or who could decide.
These failures become observable only when participants must act under time pressure, so a document review cannot capture them. Reviews and exercises that never apply pressure share the problem, which is why exercises with no failure condition pass by default.
An assessment grades the binder. A stress test grades the people holding it.
Can Your Team Perform Under Pressure? Moving From Assessment to Stress Test
Keep the assessment because it identifies the services, disruption tolerances, and named owners that the stress test should examine.
The CAF makes the same move. Under principle D1, its Testing and Exercising outcome (D1.c) expects exercises that "test all parts of your response cycle." Among the signs of not achieved, it lists exercises that "test only a discrete part of the process (e.g. that backups are working)." The framework that grades your documents also asks you to test your people.
So end every cyber resilience assessment with a short list of claims to test: the 40-minute escalation, the notification argument, the restore order. A stress test checks each claim against a team working a live scenario under time pressure. In a Reflex simulation, for example, AI agents play the adversary, the reporter, and the other stakeholders, and react to each call the team makes.
At Hacker Halted in October 2026, two teams facing the same incident both found the attack. They diverged on what no binder records: who called the bank, when the incident was declared, and how evidence was handled.
Cyber Resilience Testing: How CISOs Can Prove Resilience covers how to build the test, from scope to pass conditions. The tabletop exercise has grown up from discussion toward execution-focused simulation. The assessment identifies the claims to test; the stress test shows whether the team can meet them.
The People Pen Test: Giving CISOs Proof Their Teams Can Respond
A people pen test applies a penetration tester’s logic to the response rather than the infrastructure. Cyber Crisis Simulation: The CISO’s Guide to the Next Generation of Cyber Exercises explains the analogy in full. For an assessment program, the useful question is what evidence the exercise produces and who receives it.
At Reflex Security, the answer is the After Action Report (AAR), generated in minutes once a simulation ends. Three parts of the Reflex report carry the proof:
A minute-by-minute timeline built from timestamped exercise data. "Nobody reached the facilities director for 40 minutes" becomes a recorded interval instead of a recollection.
Findings backed by direct quotes from the exercise transcript, so each gap carries the words the team used at the time.
Action items with assigned owners and deadlines, which Reflex integrates with ticketing systems.
No single reader needs the entire report, so divide the evidence by audience:
The auditor or assessor receives the timeline and findings, set beside the outcomes the assessment rated.
The board receives the decision timeline and the findings that affect spending.
The insurer receives dated evidence that the response was exercised, with participants named.
Landlord-clients receive evidence that the tenant-facing parts of the response hold up.
Marcus J. Ranum, security entrepreneur emeritus, described the effect after a Reflex simulation: "This one made the group actually self-organize and decide under time pressure, and the gaps showed up on their own."
The assessment provides the rating, while Reflex produces evidence that can support or challenge it. Start with three claims your last assessment left untested, and see how a Reflex simulation runs.
See what your team does under real pressure.
Put your readiness to the test. Sixty minutes, your real environment, no slides.
Book A demo
Frequently Asked Questions
How often should an organization run a cyber resilience assessment?
An annual cycle suits the slow-moving parts: architecture, contracts, and governance. Re-run the affected sections after a major change, such as an acquisition, a new building systems vendor, or a reorganized facilities team. If a regulator or assurance scheme asks for the assessment, its cycle takes precedence. Exercises belong on a faster rhythm than the assessment itself.
Who should own a cyber resilience assessment: the CISO, risk or internal audit?
Split the roles. The CISO owns the content and the remediation, because the findings land on security and IT. Risk owns the link to business impact and tolerances. Internal audit or an external assessor supplies independence, and the CAF allows for both routes: self-assessment or an independent external entity. Assigning those responsibilities to three named owners is clearer than assigning them collectively to one committee.
Can a small security team run a cyber resilience assessment without a consultant?
Yes, for a first pass. The CAF is free to read, its indicators of good practice give you a structure, and the NCSC treats self-assessment as a legitimate use of it. What a small team gives up is independence: you are grading your own work. Bring in an outside assessor when the result has to persuade someone else.
Is the NCSC Cyber Assessment Framework only for UK organizations?
It was designed primarily for UK essential services and critical national infrastructure, and UK cyber regulators use it for oversight. Organizations elsewhere can adopt it as a rubric, and its outcome-based structure travels well. The CAF itself is guidance: duties come from regulation, and the NCSC has no regulatory role. Elsewhere it is a tool you choose, and your regulator may expect a different framework.
What evidence should a cyber resilience assessment produce for a board or insurer?
At minimum, a dated rating against named outcomes, the gaps with owners and deadlines, and the evidence behind each rating: plans, test logs, and contracts. Be ready for the follow-up question: has the plan been run? An exercise record with participants, timestamps, and actions answers it faster than a description of intent.
How is a cyber resilience assessment different from a penetration test?
An assessment reviews whether the organization meets resilience outcomes, mostly through documents, interviews, and configuration evidence. A penetration test attacks specific systems to find exploitable weaknesses in technical controls. The two complement each other. The assessment shows where you look exposed on paper, the pen test shows whether particular controls hold, and neither is designed to observe how the response team decides under pressure.
References
National Cyber Security Centre. (2025a, August 6). CAF objective D: Minimising the impact of cyber security incidents. Principle D1: Response and recovery planning (Version 4.0). https://www.ncsc.gov.uk/collection/cyber-assessment-framework/caf-objective-d/principle-d1-response-and-recovery-planning
National Cyber Security Centre. (2025b, August 6). Cyber Assessment Framework (Version 4.0). https://www.ncsc.gov.uk/collection/cyber-assessment-framework
National Cyber Security Centre. (2025c, August 6). Cyber Assessment Framework v4.0 released in response to growing threat [Blog post]. https://www.ncsc.gov.uk/blog-post/caf-v4-0-released-in-response-to-growing-threat
National Cyber Security Centre. (2025d, August 6). Introduction to the Cyber Assessment Framework (Version 4.0). https://www.ncsc.gov.uk/collection/cyber-assessment-framework/introduction-to-caf
Ross, R., Pillitteri, V., Graubart, R., Bodeau, D., & McQuaid, R. (2021, December 9). Developing cyber-resilient systems: A systems security engineering approach (NIST Special Publication 800-160, Vol. 2, Rev. 1). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.SP.800-160v2r1
