When a security team asks for funding, the person approving it needs to understand what could happen to the business without the change. Could staff stop taking orders? Could deliveries be delayed or customer records stolen? Explain which losses the proposal could reduce, what it costs to run and how much confidence there is in that comparison.
First separate spending needed to meet legal, contractual or essential operating requirements from optional improvements. A calculation showing a small financial benefit does not remove an obligation. Nor does an estimate of losses avoided create something the business can automatically record as an asset on its balance sheet. Finance and its advisers must assess that accounting question separately.
Describe what would stop and what it would cost
Choose a specific interruption and a period to examine. For example, ask the operations team what would happen if staff could not use the order system: how long work might stop, which deliveries would be missed and which customers would be affected. The technology team can then explain how an attacker could cause that interruption and whether existing safeguards and recovery arrangements would limit it. A software weakness alone does not tell you the size of a business loss.
Distinguish a delayed sale from a sale lost for good. If an order is fulfilled later, the business may still receive the payment. If it is cancelled, some costs, such as materials or delivery, may also be avoided. Lost contribution means the lost sales income minus those avoided costs: the money that sale would have left to cover other costs and profit. Add the costs of investigating the incident, restoring systems, extra staff and any contractual payments. Check for overlap so the same cost is not counted twice.
Show what the average-loss calculation means
Expected annual loss is an estimate of the average loss per year across many possible years, including years with no incident. In a simple calculation, multiply the estimated number of incidents per year by the average loss from each one. Both estimates need supporting evidence and a plausible range. A score predicting whether attackers will exploit a software weakness cannot tell you how often your own business will stop working. That also depends on how its systems are connected and how well its safeguards work.
The difficult part is establishing whether the safeguard could produce that reduction. Repeat the comparison with more frequent incidents, larger losses and a smaller improvement. Show when the preferred option changes. Installing a safeguard does not prove it prevents the interruption: test whether it blocks the relevant attack or helps staff get the affected work running again.
Ask whether the business could pay for a bad year
An average does not show the bill the business may face during one serious incident. Liquidity means cash, or funds the business can access in time to pay its bills. An interruption could require urgent spending on recovery while customer payments are delayed. Check whether the business could fund that gap. Consider serious but plausible incidents alongside the average-loss calculation, including safety consequences that a financial estimate cannot adequately describe.
A computer model can explore many possible outcomes, but its results depend on the assumptions supplied. Even a result showing that most modelled losses fall below a particular amount is not a guarantee that real losses will stay below it. Several teams may rely on the same supplier or the same recovery staff. If they all need help at once, treating their interruptions as unrelated will understate the pressure on recovery.
Identify what insurance would leave the business to pay
Ask the insurance adviser to explain which events and costs the policy covers, what it excludes, how long cover takes to begin and when a claim could be paid. Check the amount the business must pay itself before insurance contributes, as well as the maximum payout. You cannot simply subtract that maximum from the expected annual loss. A policy pays for eligible costs of particular incidents, not for a calculated annual average.
For each serious incident considered, show what the insurer might reimburse and what would remain with the business. Those remaining losses are sometimes called retained exposure. They include uncovered costs and amounts above the policy limit. The business may also need to pay for recovery before the insurer settles the claim. Insurance can reimburse eligible losses; it cannot get orders moving again. Preventing interruptions, recovering from them and paying their costs are related but different decisions.
Approve real spending against a testable assumption
Show the estimated reduction in losses beside the money that must actually be approved. Use the same period for both and explain the interruption the proposal addresses. Name the person responsible for running the safeguard and involve the manager whose team's work it protects. They need to agree how to check whether it is helping.
In the hypothetical calculation, approval commits AED 164,400 upfront and AED 41,100 a year because the safeguard is assumed to halve the frequency of interruptions. Review whether it works against the cause of that interruption and whether changes to systems or suppliers have made the original estimate less credible. If the assumption weakens, reconsider the spending case. Do not report AED 123,300 a year as a saving the business has actually achieved.