Most small-business risk assessments fail for a mundane reason: they are questionnaires completed once, not documents anyone uses. The answers are given under time pressure, the file is saved, and the same gaps are discovered again a year later. A checklist becomes useful only when every line has an owner, a piece of evidence and a review date.
This one is designed for an SME in Bangladesh with a handful of IT staff, a mixture of cloud and office systems, and one manager who can accept risk. Copy it into a spreadsheet, delete what does not apply, and keep the rest under version control.
Set the scope before you tick anything
An assessment with no boundary produces a list of everything and a decision about nothing. Write down four things first:
- Critical services: the five to ten things the business cannot run without for a day — order taking, payroll, invoicing, production scheduling, customer support.
- Supporting assets: the applications, servers, cloud accounts and data stores behind each service, including the ones managed by a vendor.
- Obligations: contractual security clauses, payment-card rules, data retention duties and any regulatory reporting that applies to your sector. Confirm the legal position with your adviser.
- Assessment team: a facilitator, an IT owner and one business owner per critical service. A risk register written only by IT misses the impact that matters.
State the review period and risk appetite together. "We accept low and medium risks; high risks need a funded treatment plan; critical risks stop the project" is more useful than a statement about culture.
Use a simple scoring rule
You do not need a statistical model. Score likelihood and impact from 1 to 4, multiply them, and record the result before and after treatment. Define the levels in words so two people score the same event similarly.
- Likelihood 1: no credible path in the next year. 2: possible if nothing changes. 3: likely given current controls. 4: already happening or reported in the sector.
- Impact 1: inconvenience. 2: a day of lost work. 3: lost revenue, customer harm or a contractual breach. 4: inability to operate, regulatory exposure or irreversible data loss.
A score of 12 or above deserves a named treatment plan with a date. A score of 6 or below can be accepted, but write down who accepted it. For each finding, choose one of four treatments: reduce, avoid, transfer or accept. "Transfer" usually means insurance or a warranty, not a vendor's assurance.
The checklist
Governance and scope
- Named risk owner: one person accountable for the register. Owner: managing director. Evidence: signed cover page. Review: annually.
- Critical service register: services, owners and tolerable downtime. Owner: operations lead. Evidence: versioned list. Review: twice a year.
- Policy set: access, backup, email, acceptable use and incident response, each with an owner and a review date. Owner: IT lead. Evidence: document index. Review: annually.
- Insurance and contracts: cyber cover, vendor liability and notification duties checked against the register. Owner: finance. Evidence: policy schedule. Review: annually.
Assets, data and dependencies
- Hardware and software inventory: devices, operating systems, licence status and end-of-support dates. Owner: IT lead. Evidence: exported asset report. Review: quarterly.
- Cloud and SaaS register: account owner, business purpose, data held and admin count. Owner: IT lead. Evidence: register export. Review: quarterly.
- Data map: where customer, employee and financial data lives, who can access it and how long it is kept. Owner: department heads. Evidence: data map. Review: annually.
- Dependency map: internet, power, payment channels and courier or logistics providers. Owner: operations. Evidence: one-page diagram. Review: annually.
Identity and access
- Named accounts: no shared logins for email, finance or administration. Owner: IT lead. Evidence: account export. Review: quarterly.
- Multi-factor authentication: enforced on email, remote access, cloud admin and banking. Owner: IT lead. Evidence: configuration screenshot. Review: quarterly.
- Joiners, movers and leavers: a checklist that removes access on the last working day. Owner: HR. Evidence: signed offboarding form. Review: each departure.
- Privileged access: separate admin accounts, limited to named staff, with activity logging. Owner: IT lead. Evidence: admin group list. Review: quarterly.
Endpoints, email and browsing
- Supported devices: operating system and browser versions still receiving updates. Owner: IT lead. Evidence: patch report. Review: monthly.
- Endpoint protection: anti-malware or endpoint detection on every laptop and server, with alert ownership. Owner: IT lead. Evidence: console coverage report. Review: monthly.
- Email authentication: SPF, DKIM and DMARC published for every domain, with a monitored reporting address. Owner: IT lead. Evidence: DNS records and aggregate reports. Review: quarterly.
- Phishing route: a reporting button or mailbox, plus a written first-hour procedure. Owner: IT lead. Evidence: test report. Review: twice a year.
Backup, recovery and continuity
- Backup scope: critical data, configurations and cloud workloads covered by a schedule. Owner: IT lead. Evidence: backup job report. Review: monthly.
- Restore test: a real restore into an isolated environment, with the time recorded. Owner: IT lead. Evidence: test log. Review: quarterly.
- Offline copy: one copy that ransomware cannot encrypt through the same credentials. Owner: IT lead. Evidence: storage configuration. Review: quarterly.
- Continuity plan: manual workaround, contact tree and recovery order for each critical service. Owner: operations. Evidence: tested plan. Review: annually.
Vendors, monitoring and response
- Vendor register: who holds your data, what they can access and which contract covers it. Owner: procurement. Evidence: register. Review: twice a year.
- Vendor assurance: a security questionnaire or certification evidence proportionate to the data. Owner: procurement. Evidence: completed questionnaire. Review: annually.
- Monitoring: uptime, certificate expiry, failed backups and admin logins watched by a named person. Owner: IT lead. Evidence: alert configuration. Review: quarterly.
- Incident exercise: a tabletop scenario with decisions, contacts and evidence retention. Owner: risk owner. Evidence: exercise notes. Review: annually.
Sign-off that means something
Close with a one-page record: scope, date, participants, findings by score, treatments approved, risks accepted and the next review date. The business owner signs it, not the IT lead. Attach the evidence file.
Mistakes that waste the exercise
- Treating a vendor questionnaire as an assessment. It records what a supplier says, not what your business depends on.
- Scoring everything high. If every risk is critical, nothing is prioritised and the budget goes to the loudest item.
- Leaving actions without dates or owners. An unassigned action is a wish.
- Repeating the assessment annually with no quarterly check of access, backups and patches — the areas that change fastest.
A four-week rhythm
Week one: scope, team and evidence. Week two: score findings with service owners. Week three: agree treatments and costs. Week four: sign-off and schedule quarterly checks. Keep the register small enough to review in one meeting; evidence can live elsewhere, but decisions belong in one place.
The assessment sits inside Managed Technology & Cybersecurity, alongside hardening and monitoring work. The pricing page explains how one-off assessment work differs from a retainer. For the budget side of the same decision, read the cost guide for SME security assessments.
Sources
- NIST, Cybersecurity Framework 2.0, accessed 17 November 2026.
- NIST SP 800-30 Rev. 1, Guide for Conducting Risk Assessments, accessed 17 November 2026.
- OWASP Application Security Verification Standard, accessed 17 November 2026.
- CISA, Cyber Essentials, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.