Manufacturing & Quality · Practical Guide
How to Use a 5×5 Risk Matrix
By Qhubio · Published
A 5×5 risk matrix places each risk at the intersection of a likelihood rating and an impact rating. Qhubio multiplies the two ratings to produce a score from 1 to 25, helping a team discuss priorities consistently.
Build a free 5×5 risk matrixDefine likelihood and impact before scoring
Likelihood describes how plausible the event is within your stated scope and time horizon. Impact describes the consequence if it occurs. Agree the rating anchors before comparing risks; ‘possible’ over one shift is different from ‘possible’ over a product’s lifetime.
Qhubio uses five likelihood labels: Rare, Unlikely, Possible, Likely and Almost certain. Impact runs from Insignificant through Minor, Moderate and Major to Severe. Relate these generic labels to your process evidence and objective; the labels are not calibrated probabilities or monetary values.
Interpret Qhubio’s score bands
Calculate Likelihood × Impact. For example, likelihood 3 and impact 4 produce a score of 12. In Qhubio’s generic model this is High. Use the result to open a discussion about controls, urgency and ownership rather than to automatically approve or reject the risk.
| Score | Band | Discussion prompt |
|---|---|---|
| 1–4 | Low | Confirm the controls and review the assumptions. |
| 5–9 | Moderate | Consider proportionate improvement and follow-up. |
| 10–16 | High | Review the exposure and prioritise treatment. |
| 17–25 | Critical | Escalate the exposure and response for prompt review. |
Worked example: machining cell risk
A team discusses bore diameter drift as a tool approaches its life limit. Its illustrative starting assessment is likelihood 4 and impact 4, score 16 High. The existing control is periodic dimensional sampling; the team records why that coverage may miss a change between samples.
The team proposes a tool-life limit and a defined reaction to measurement trends. Once those controls are implemented and checked, a likelihood rating of 2 with impact still 4 would produce 8 Moderate. This is an example of a reassessment, not a promised outcome of installing a tool counter. Evidence must support the lower likelihood.
Avoid false precision and colour-only decisions
Equal scores can describe very different exposure. Likelihood 1 × impact 5 and likelihood 5 × impact 1 both produce 5, yet their consequences and appropriate responses differ. The rating scales are ordinal: multiplying the numbers does not turn them into a quantitative loss estimate.
- Describe the event and consequence before assigning a score.
- Record the time horizon and evidence used for likelihood.
- Review severe consequences separately from the colour band.
- Do not lower ratings solely because an action is planned.
- Use consistent definitions when comparing rows or assessments.
Keep the matrix connected to the working record
The heatmap shows where risks sit; the register explains what each point means. Keep controls, mitigation, ownership and status alongside the risk description. Review the assessment after process changes, new complaints or evidence that challenges an assumption.
If you need to distinguish several failure causes, customer effects or prevention and detection controls within one process step, move to Process FMEA. There is no valid automatic conversion from a 1–25 matrix score to FMEA S/O/D ratings or Action Priority.
Continue with a structured Process FMEA
Need to explain how a process fails rather than only where its broad risks sit? Use Process FMEA to connect each failure mode to effects, causes and controls.
Related guides
Methodology references
The worked examples and generic score bands are Qhubio’s practical illustrations. Consult the applicable methodology and your organisation’s criteria for formal decisions.
