Cyber risk

The quantification method

Make the assumptions visible.

A protection decision rests on evidence and judgement. See where each enters the analysis before relying on the estimate.

Follow a worked reasoning path
A red solid built from three substantial strata, separated to reveal the surfaces beneath the finished form.
Look beneath the finished answer.
Separate the layersReturn to the first view

An abstract study, not an analysis result.

Illustration, not a customer result or model output

Should the warehouse fund more recovery protection?

Suppose an interruption stops orders leaving. Follow the activity through recovery, rather than treating a restarted system as the end of the loss.

  1. Inputs

    What can we establish?

    Identify the systems needed to process orders, the controls in place and available recovery records. Ask operations what staff must do before dispatch resumes.

  2. Assumptions

    What still depends on judgement?

    Can staff work manually? How long might checking those orders take? Record evidence gaps and use ranges where a single value would imply false confidence.

  3. Decision

    What would change the choice?

    Compare the losses that could remain with the cost of protection. If the choice depends on an untested recovery assumption, seek better evidence before committing.

The output is a basis for choosing, accepting exposure or investigating further. An estimated reduction in risk is not a saving already realised.

FAIR: how often a loss event may occur and what it may cost

Our cyber risk work uses FAIR, an approach that relates how often a loss event may occur to how much it may cost. Threat conditions, controls and the organisation’s circumstances inform the estimates. Financial consequences can include interruption, response, recovery and other losses relevant to the scenario.

How investigation evidence is assembled
Expected loss: an estimate, not a prediction or a maximum bill

For the defined period, review the range of possible outcomes alongside expected loss. This modelled estimate is neither a prediction of the next incident nor the maximum possible cost.

Compare current exposure, what proposed controls could change and the losses that could remain with implementation and operating costs. Record gaps and disagreements. Revisit the analysis when meaningful changes to assets, controls or operations affect its assumptions.

The analysis system we deliver and its IP belong to you. You choose any AI model it uses and who operates it, and you can run it without us.