This appendix gathers the questions that matter most into one place. You can use it in several ways. A new board member can read it before their first meeting to understand what good oversight looks like. A chair can pick three or four questions for each quarterly meeting, so that over a year the board covers the whole ground without overwhelming anyone. A chief executive can use it to prepare, asking these questions privately before the board asks them publicly. And a head of security can use it as a checklist of what the board is likely to want to know, and prepare honest answers in advance.
A few words on how to ask. These questions are meant to open a conversation, not to set a trap. Ask them in a calm, curious tone. The goal is to understand, not to catch someone out. If people feel that a difficult answer will be punished, they will learn to give easy answers instead, and the board will lose the very information it needs.
Listen for the shape of the answer as much as the content. A good answer is usually short, specific, and honest about uncertainty. It names a person, a date, a number, or a document. It says "we do not know yet, and here is how we will find out" when that is the truth. A weak answer tends to be long, general, and reassuring. It uses phrases like "we take security very seriously", "industry best practice", or "we have a number of controls in place". These are not lies, but they are not answers either. When you hear them, ask the question again in a narrower form: "Which controls? Who checked? When?"
Under each group below, you will find the questions themselves and a brief note on what a good answer tends to sound like. The notes are not scripts. Your company is different from Meridian Foods and from every other company, and a good answer for you will reflect your size, your sector, and your history.
Strategy and risk appetite
- What are the three cyber events that would hurt us most, and how badly? A good answer names specific scenarios, such as "our production systems are locked for a week" or "our customer database is stolen and published", and gives a rough sense of cost in money, time, and reputation. It should not be a list of every possible threat.
- How much cyber risk are we willing to accept, and who decided that? Many companies have never written this down. A good answer points to a statement of risk appetite that the board has seen, even if it is short. If none exists, the honest answer is "we have not set one", and the follow-up is to agree when it will be drafted.
- How does cyber risk appear in our enterprise risk register, and how does it compare with our other top risks? Cyber risk should sit alongside financial, operational, legal, and market risks, using the same scales. If it lives in a separate document that the board never sees next to the others, the board cannot weigh it properly.
- What has changed in our cyber risk since we last discussed it? This question keeps the conversation current. Good answers mention new systems, new suppliers, new regulations, new incidents in the sector, or changes in the business itself, such as an acquisition or a new product.
- Which of our strategic plans for the next two years will change our cyber risk, and has security been involved in planning them? New markets, new products, mergers, and digital projects all change the shape of the building. Security should be in the room early, not called in to approve things at the end.
People and accountability
- Who is the single named executive accountable for cyber risk, and to whom do they report? The answer should be a name, not a department. If the head of security reports through several layers to someone who rarely meets the board, bad news may travel slowly.
- When did our head of security last speak directly to this board or its risk committee, without management filtering the message? Boards benefit from hearing the security leader's own words at least once a year, and from having a private channel if something serious needs to be said.
- Do we have enough people, with the right skills, to do what we have asked of them? A good answer is candid about gaps, describes how they are covered (for example, by an outside provider), and names the risk that remains.
- How would an employee report a security worry or a mistake, and what happens to them when they do? You want to hear that reporting is easy, that people are thanked rather than blamed, and that reports actually arrive. A company with no reported mistakes is not a company with no mistakes.
- How do we know our security training changes behaviour, rather than just being completed? Completion rates measure attendance, not learning. Better answers mention trends in reported suspicious messages, results from realistic exercises, and changes in real-world incidents.
Suppliers and partners
- Which of our suppliers could stop our business, or expose our customers' data, if they were compromised? A good answer is a short, ranked list. Most companies have hundreds of suppliers, but only a handful are critical. The board should know who they are.
- What checks do we make before giving a supplier access to our systems or data, and do we repeat them? You are listening for a proportionate process: deeper checks for critical suppliers, lighter ones for the rest, and periodic review rather than a single questionnaire at signing.
- Do our key contracts require suppliers to tell us quickly if they suffer a breach, and allow us to verify their security? Notification clauses and audit rights are easy to include at signing and very hard to add later.
- Where are we dependent on a single provider, and what would we do if they were unavailable for a week? This is concentration risk. Good answers acknowledge the dependency and describe a fallback, even if the fallback is slow or manual.
Cloud and technology
- Which of our critical systems and data live with cloud providers, and do we understand which parts of security are theirs and which are ours? The shared-responsibility idea is simple to state and easy to forget. A good answer can name the dividing line for the most important services.
- Do we know all the software services our staff use, including those bought on a credit card without IT's involvement? Few companies can answer yes with confidence. A good answer describes how unknown services are discovered and brought under basic control.
- Is multi-factor authentication turned on for every account that could do serious harm: email, finance systems, administrator accounts, and remote access? This single control prevents a large share of common attacks. The answer should be yes, or a clear plan with a date.
- How quickly do we fix known weaknesses in our most important systems? You want a measure, such as "critical issues on internet-facing systems are fixed within fourteen days, and we meet that target ninety per cent of the time". Vague answers here deserve a follow-up.
- What rules do we have for staff using artificial intelligence tools, and how do we know they are followed? A good answer describes a simple acceptable-use policy, approved tools, and some way of noticing when sensitive data is being shared where it should not be.
Incident readiness
- If we discovered a serious incident at 3 a.m. on a Sunday, who would be called first, and who would make the key decisions? The answer should be specific and should not depend on one person who might be on holiday.
- When did we last practise our incident response, and did the board or executive team take part? Plans that are never rehearsed tend to fail on first contact with reality. A tabletop exercise once a year, involving leadership, is a sensible baseline. Appendix C offers one.
- Have we tested restoring our most important systems from backup, and how long did it take? "We have backups" is not the same as "we can restore". A good answer gives a date and a duration from a real test.
- Have we decided in advance how we would approach a ransom demand? The decision does not have to be final, but the principles, the decision-makers, and the legal and ethical constraints should be agreed before a crisis, not during one.
- Do we know which regulators, customers, and partners we must notify after a breach, and how quickly? Deadlines can be short and vary by country and sector. A good answer points to a prepared list and named owners.
- Who would speak for the company publicly, and have they practised? Crisis communications are a leadership skill. Statements drafted in advance, even in rough form, save precious hours.
Insurance and financial resilience
- What does our cyber insurance actually cover, what does it exclude, and what are the limits? A good answer is plain and specific. Board members should know the headline limit, the main exclusions, and any conditions that could void a claim.
- Are the security controls we described to our insurer actually in place today? Insurers increasingly require specific controls. If a claim reveals that a promised control was missing, the claim may be reduced or refused.
- If we suffered a serious incident, how much would it cost us beyond what insurance would pay? This is the retained risk. It helps the board decide whether to spend more on prevention, more on insurance, or accept the gap knowingly.
Metrics and reporting
- What are the five or six measures that tell us whether our security is getting better or worse? You want a small set of meaningful indicators, tracked over time, not a long list of activity counts.
- What is the one thing in our security that you are most worried about, and what would it take to fix it? This question often produces the most useful answer of the meeting. Ask it every time.
- What would you tell us if you were not worried about how it would be received? A gentle version of the previous question, best asked in private. It signals that honesty is welcome.
- Has anyone independent checked what we are being told? Internal audit, an outside assessment, or a certification body can provide assurance. A good answer describes when the last independent review happened and what it found, including the uncomfortable parts.
Using the question bank over a year
A simple rhythm works for most boards. In the first quarter, focus on strategy and risk appetite. In the second, on people and suppliers. In the third, on cloud, technology, and incident readiness, ideally combined with a tabletop exercise. In the fourth, on insurance, metrics, and the plan for the year ahead. Questions 29 and 30 are worth asking every quarter.
Keep a short record of the answers. Over time, the record becomes valuable in its own right. It shows how the company's understanding has grown, where promises were made and kept, and where the same worry has been raised three times without being resolved. That last pattern, more than any single answer, is the one a board should notice.
KEY POINTS
Key points
- Choose three or four questions for each quarterly meeting instead of asking everything at once.
- Ask in a calm, curious tone; punishing difficult answers teaches people to give easy ones.
- When an answer is general and reassuring, narrow it: which controls, who checked, and when.
- Ask the head of security what worries them most at every meeting.
- Keep a short record of answers and look for concerns raised repeatedly without resolution.
QUESTIONS
Common questions
- What questions should a board ask about cybersecurity?
- The Boardroom Firewall groups the key board questions into seven themes: strategy and risk appetite, people and accountability, suppliers, cloud and technology, incident readiness, insurance, and metrics and reporting. Examples include which three cyber events would hurt the company most, who the single accountable executive is, and whether critical systems have been restored from backup in a real test.
- How can a board tell a good answer about cyber risk from a weak one?
- A good answer is short and specific: it names a person, a date, a number or a document, and admits what is not yet known. A weak answer is long, general and reassuring, relying on phrases such as industry best practice. When directors hear general reassurance, they should narrow the question by asking which controls, who checked them, and when.
- How often should a board discuss cybersecurity?
- Cyber risk should come up at every quarterly meeting, but a board does not need to cover everything each time. Choosing three or four questions per quarter lets the board cover strategy, people, suppliers, technology, incident readiness, insurance and metrics within a year. Questions about key security measures and the biggest current worry are worth asking every quarter.
- Who should be accountable for cyber risk in a company?
- There should be one named executive accountable for cyber risk, and the board should know who that person is by name rather than by department. Directors should also check who that executive reports to, because a head of security buried several layers below the board means bad news reaches leaders slowly.
- What is the most useful question to ask a head of security?
- Asking what worries them most about the company's security, and what it would take to fix it, often produces the most valuable answer in a board meeting. A gentler variant, best asked privately, is what they would say if they did not have to worry about how it would be received. Both signal that honesty is welcome.