Risk Matrix – What It Is, How to Build It and How to Read It

A risk matrix (also called a risk map) is a table that combines the likelihood and impact of a risk into a single level and marks it with a color. It is used for prioritization – it shows which risks to address first – and for communication with the board. It most often has a 5×5 scale, the risk level is the product of likelihood and impact, and the thresholds and colors are set by the organization in line with its risk appetite. A matrix is a decision-support tool, not a measurement: its result has to be read together with the justification of the assessment.

A risk matrix at a glance: one axis is likelihood, the other is impact; risk level = likelihood × impact; each color corresponds to a different response (accept, treat, escalate); the matrix organizes decisions but does not replace them.

In this article we explain what a risk matrix is and what it consists of, how to define likelihood and impact scales, how a 3×3 matrix differs from a 5×5 one, how to calculate the risk level and interpret the colors, and which mistakes to avoid. At the end we show a risk heat map with example risks, before and after controls. The broader context – the process, the register and the owners – is covered in our guide to risk management.

What Is a Risk Matrix?

A risk matrix (in IEC 31010, a consequence/likelihood matrix) is a technique for visualizing risk. Each risk is placed in the cell at the intersection of its likelihood and impact, and the cell has an assigned level and color. In everyday language, the terms risk map and risk heat map are also used: they usually mean a matrix colored by level, on which specific risks or their number per cell are additionally shown.

A matrix is used at two stages of risk management. In risk analysis it helps establish the level of each risk, and in evaluation it helps compare that level with the acceptability criteria. The results go into the risk register, and the matrix is its graphical summary for the board.

What Does a Risk Matrix Consist Of?

A matrix consists of several elements, each requiring a decision from the organization:

Element What it is Example
Likelihood axis A scale of the chance that the event will occur, with described levels Five levels from “rare” to “almost certain”
Impact axis A scale of consequences, usually across several dimensions Five levels from “negligible” to “critical”
Matrix cells Each intersection of the axes has a risk level value Likelihood 3 × impact 4 = 12
Thresholds The boundaries between risk zones 1-4 low, 5-9 medium, 10-16 high, 17-25 critical
Colors The marking of risk zones Green, yellow, orange, red
Response rules The response and decision-maker assigned to each zone High risk is accepted by the board

The last element is the one most often skipped, and without it a matrix is just a colored table: colors make sense only when each of them means a different decision. In ISO 31000 this corresponds to the concept of risk criteria, i.e. the level the organization considers acceptable. An example 5×5 matrix with cell values (risk levels):

Likelihood ↓ / Impact → 1 – negligible 2 – minor 3 – moderate 4 – major 5 – critical
5 – almost certain 5 10 15 20 25
4 – likely 4 8 12 16 20
3 – possible 3 6 9 12 15
2 – unlikely 2 4 6 8 10
1 – rare 1 2 3 4 5

How to Define Likelihood

Every level of the scale should have a description that lets two people rate the same event in the same way. The simplest approach is to write the scale as a frequency over a fixed time horizon. An example (an illustrative scale; each organization adapts it):

Level Description Frequency
5 Almost certain once a year or more often
4 Likely once every 1-2 years
3 Possible once every 2-5 years
2 Unlikely once every 5-10 years
1 Rare less often than once every 10 years

Three rules help keep ratings consistent. First, the event should be defined precisely (not “an incident” but “a successful ransomware attack within 12 months”). Second, the assessment should rest on data, such as incident history and test and audit results, rather than on intuition alone. Third, the levels of the scale must be mutually exclusive and cover every possible value, with no gaps.

How to Define Impact

Impact is best assessed across several dimensions at once, with the highest of them as the final result – conservatively, because one serious consequence is enough to make a risk significant. An illustrative scale:

Level Operational (critical process downtime) Financial (loss) Regulatory Reputational
5. Critical over 24 h over PLN 500k sanctions or restriction of activity lasting loss of trust of key customers
4. Major 4-24 h PLN 200-500k notification to an authority, proceedings negative coverage in industry media
3. Moderate 1-4 h PLN 50-200k explanations for an auditor or authority complaints from several customers
2. Minor up to 1 h PLN 10-50k a shortcoming with no external consequences isolated complaints
1. Negligible no downtime under PLN 10k none none

Financial thresholds have to be matched to the size of the organization: a loss of PLN 200k means something different for a small services company and for a large operator. One practice is to express thresholds as a percentage of annual revenue or profit. Data for impact assessment is best taken from business impact analysis (BIA): the maximum tolerable downtime of the process, the cost of an hour of downtime, and dependencies on systems and vendors.

3×3 or 5×5 Risk Matrix?

The number of levels is a trade-off between simplicity and resolution. In practice, 3×3, 4×4 and 5×5 matrices are all used (for example, AS/NZS 4360:2004 used 4×4 and MIL-STD-882B used 5×5). The most common choice is 3×3 or 5×5:

Feature 3×3 4×4 5×5
Number of cells 9 16 25
Resolution Low: large groups of risks in one cell Medium Higher: risks are easier to tell apart
Simplicity of ratings High: three levels are easy to define Medium Lower: requires precise level descriptions
The “middle” option There is a middle level that ratings easily drift toward No middle, so a rating forces a decision There is a middle level
Typical use Small organizations, first analyses, quick reviews Organizations that want to avoid the “middle” Medium and large organizations, registers with many risks

An example 3×3 matrix with cell values (thresholds: 1-2 low, 3-4 medium, 6-9 high):

Likelihood ↓ / Impact → 1 – low 2 – medium 3 – high
3 – high 3 6 9
2 – medium 2 4 6
1 – low 1 2 3

BCMLogic expert commentary: a 5×5 matrix with poorly defined levels is worse than a 3×3 with good definitions. The number of levels should match how precisely the organization can tell ratings apart and what data it has. Starting with 3×3 makes it easier to build a shared language, and the scale can be expanded once data and the need appear.

How to Calculate the Risk Level

The risk level is most often calculated as the product of likelihood and impact, and the result is assigned to a zone according to thresholds. The product is not the only option:

Method How it works Strengths and limitations
Product (L × I) Level = likelihood × impact Simple and widespread; different pairs can give the same result (e.g. 2 × 6 and 3 × 4)
Sum (L + I) Level = likelihood + impact Results of 2-10; the same limitation: different pairs give the same result (e.g. 1 + 5 and 3 + 3)
Manual table The organization assigns a zone to each cell directly, without a formula Full control (e.g. impact 5 is always “high”); each cell needs justification
Override rules A formula plus exceptions, e.g. critical impact escalated regardless of the result Protects against hiding rare but devastating events; must be written into the methodology

On a 5×5 scale the product gives only 14 distinct results (1, 2, 3, 4, 5, 6, 8, 9, 10, 12, 15, 16, 20, 25), so many risks get the same result even though they describe different things. That isn’t a flaw in the formula but the nature of ordinal scales: levels 1-5 are categories, not measurable quantities, so multiplying them gives a convenient label for prioritization, not a measurement. In GRC-class systems the level is calculated automatically from the matrix; taking BCMLogic as an example, according to the description of its risk management module, the matrix takes likelihood and multiple types of impact into account.

How to Interpret Risk Matrix Colors

A color stands for a decision, not decoration. With typical 5×5 thresholds the zones are not equal: of 25 cells, 8 are low, 7 medium, 7 high and only 3 critical. That is why thresholds should be treated as the organization’s own settings, not objective truths: a cell scoring 9 and a cell scoring 10 differ by one point, but they sit on opposite sides of a boundary and call for a different decision. Example rules for each zone:

Zone Level Typical response Decision-maker Review
Low (green) 1-4 Acceptance, monitoring within cyclical reviews Risk owner Once a year
Medium (yellow) 5-9 Treatment within a reasonable time or justified acceptance Risk owner Quarterly or half-yearly
High (orange) 10-16 A mandatory treatment plan with a deadline Board or risk committee At least quarterly
Critical (red) 17-25 Immediate escalation and action; continuous oversight Board Ongoing

A few rules help avoid misreading a matrix. First, it is worth always putting a number or the name of the zone in the cells rather than relying on color alone, because some viewers can’t tell red from green and a black-and-white printout erases the differences. Second, inherent risk (before controls) and residual risk (after controls) should not be mixed on one map without clear marking. Third, colors are interpreted together with the organization’s risk appetite: the same zone in two companies can mean a different response.

Common Mistakes in Risk Matrices

  • Scales without definitions – “medium impact” means something different to each participant, so ratings aren’t comparable.
  • Colors without consequences – zones have no assigned decisions, so red and yellow make no practical difference.
  • Treating the result as a measurement – research on matrices (Cox, 2008) shows they have poor resolution and can assign a higher rating to a smaller risk than to a larger one.
  • Ignoring critical impact – a rare but devastating event gets a low score because likelihood “eats” the impact.
  • No split between inherent and residual risk – it isn’t visible what the controls are “holding.”
  • Too many levels with too little data – false precision instead of a useful assessment.
  • A matrix without justifications – just dots on a map, with no recorded reasons for the ratings.
  • The matrix as the only tool – for key risks it is worth supplementing it with scenario analysis or quantitative analysis.

Risk Matrix Example

The heat maps below show three risks from a sample risk register of a fictional services company (illustrative data): R-021 (invoicing system unavailable), R-024 (customer data sent to the wrong recipient) and R-027 (active accounts of former employees). The value in a cell is the risk level, with the risk identifiers beneath it. First, the inherent risk, i.e. before controls are taken into account:

Likelihood ↓ / Impact → 1 – negligible 2 – minor 3 – moderate 4 – major 5 – critical
5 – almost certain 5 10 15 20 25
4 – likely 4 8 12
R-024
16 20
3 – possible 3 6 9 12
R-021
15
R-027
2 – unlikely 2 4 6 8 10
1 – rare 1 2 3 4 5

After existing controls are taken into account (residual risk):

Likelihood ↓ / Impact → 1 – negligible 2 – minor 3 – moderate 4 – major 5 – critical
5 – almost certain 5 10 15 20 25
4 – likely 4 8 12 16 20
3 – possible 3 6 9
R-021, R-024
12 15
2 – unlikely 2 4 6 8 10
R-027
1 – rare 1 2 3 4 5
Risk Before controls After controls What it means
R-021 3 × 4 = 12 (high) 3 × 3 = 9 (medium) The owner can accept it; planned actions will reduce it further
R-024 4 × 3 = 12 (high) 3 × 3 = 9 (medium) As above; both risks land in the same cell of the matrix
R-027 3 × 5 = 15 (high) 2 × 5 = 10 (high) Still high: needs a board decision or more urgent action

How to read this map. On the first map, all three risks sit in the high zone. On the second, two of them move to the medium zone, while R-027 stays in the high zone because the controls reduced only the likelihood and the impact remained critical. That risk should be at the top of the priority list and needs a board decision. R-021 and R-024 occupy the same cell (9) even though they concern entirely different matters: a matrix says how urgent a risk is, not what it is.

FAQ

What is the difference between a risk matrix, a risk map and a heat map?

In practice they are synonyms. A risk matrix is a likelihood × impact table. A risk map or heat map usually means a matrix colored by risk level, on which specific risks or their number per cell are marked.

Is a 5×5 matrix better than a 3×3 one?

Not always. A 5×5 matrix gives higher resolution but requires precisely described levels and data that let them be told apart. A well-defined 3×3 can be more useful than a 5×5 with levels rated by feel.

Is a risk matrix required by regulations?

No. Standards and regulations such as ISO/IEC 27001, NIS2 or DORA usually require documented risk analysis, not a specific technique. A risk matrix is one of the techniques described in IEC 31010 and the most common form of visualization.

How should the thresholds and colors of a matrix be set?

Thresholds follow from risk appetite and acceptance rules: the board decides from which level a risk requires its decision. A practical test is to ask whether each zone has a different response. If two colors mean the same thing, one of them is redundant.

Can likelihood and impact be multiplied?

Yes, but it should be remembered that this is a convention for prioritization. The scale levels are ordinal, so the product is not a measure of the size of a risk, and different pairs of values can give the same result. For high-impact risks it is worth looking at the impact value itself as well.

Is a matrix enough to assess risk?

For initial prioritization, yes. For the most important risks, it is worth supplementing it with a scenario description, justification of the ratings and, where possible, quantitative analysis, because a matrix on its own has limited resolution.

Summary

A risk matrix is only as good as the definitions it stands on: described likelihood and impact levels, thresholds linked to risk appetite, and colors that each mean a different decision. Used as a tool for prioritization and for conversations with the board – not as a precise measurement – it brings order to work with the risk register. Used without definitions, it only gives an impression of precision.

Read Also

See What This Could Look Like for Your Organization

If your organization keeps its risk matrix in a spreadsheet and is looking for consistent scales, automatic calculation of the risk level and board reports, it’s worth seeing what this looks like on a GRC platform. BCMLogic Solutions helps financial and technology companies manage operational risk, business continuity and regulatory compliance – in line with ISO 22301, ISO 27001 and requirements such as DORA and NIS2/KSC.

Book a Demo →