What is a RAG status and how do you decide the colour?

RAG stands for red, amber, green — a three-state summary of project health. Green means delivery is on plan. Amber means a variance exists and is being managed inside the project. Red means the variance cannot be resolved by the project and needs a decision from governance. The colour only carries information if the thresholds are written down first.

What the three colours mean

ColourPositionWhat it asks of the reader
GreenDelivering to plan within tolerance. No help needed.Note it and move on
AmberA variance exists. There is a recovery action, an owner and a date. The project can still handle it.Be aware; expect a follow-up
RedThe variance cannot be resolved inside the project, or a tolerance has been breached.Take a decision — scope, money, time or authority

Read that way, the colour is not a score for the team. It is a request. Green requests nothing, amber requests attention, red requests a decision. A programme where red is treated as failure will not produce a red until it is far too late to act on one.

The colour needs a written rule

Most RAG dashboards lose their meaning because the colour is a feeling rather than a threshold. Two projects, the same slippage, different colours, because one manager is cautious and the other is not. Once that is true across a portfolio, the dashboard stops distinguishing anything.

The correction is to write the thresholds down in the charter or the reporting standard, before the first report. They are usually a variance against baseline:

The tolerances themselves — five per cent, ten per cent, two weeks — are set by the organisation and by what the project can absorb. The point is that they exist and are the same next month.

Per-dimension colours, then an overall

A single colour for a whole project compresses too much. Schedule green and budget red is a real and common position, and one colour cannot say it. Most reports carry a colour per dimension and one overall, with the overall following an explicit rule — commonly the worst of the individual colours, or the worst that cannot be resolved inside the project. Whichever rule is used, it needs to be stated, because otherwise readers assume an average and the overall reads as better than the parts.

Direction of travel

The colour is a position; the arrow next to it is the trend. Amber and improving is a project doing its job. Amber and worsening is a project about to go red and worth a conversation now. The arrow is compared with the previous report, not with plan, and it is the field most often left off a dashboard that has room for it. Including last week's colour alongside this week's makes the change visible without anyone having to remember. It belongs on the same line as the colour rather than in the narrative underneath it.

Amber drift

The failure mode of RAG reporting is a project that has been amber for eleven weeks. Amber is meant to be a transitional state: there is a variance, there is a recovery action, and by the date on that action it either resolves back to green or escalates to red. Persistent amber usually means the recovery action has no date, or has a date that keeps moving.

Two practical countermeasures. Record the number of consecutive weeks at the current colour on the report — a visible counter changes behaviour more than a policy does. And require every amber to name the recovery action, its owner and the date it resolves by, so amber costs something to declare.

Who sets the colour

The project manager sets it and the sponsor can challenge it. Where a PMO applies rules to reported data — schedule variance, budget variance, defect counts — the colour becomes calculated rather than declared, which removes the argument and some of the judgement. Most organisations end up with a mixture: calculated by dimension, with the overall set by the PM and explained in one sentence.

What breaks the system is a colour negotiated downwards before publication. Once a report has been amber and appears as green, the whole dashboard becomes a description of how much bad news each manager is willing to write down.

What red obliges

Red is only useful if something follows it. In most governance models a red status obliges a recovery plan within an agreed period and an item on the next steering agenda with a decision attached. Without that, red is just a colour, and the incentive to avoid it is entirely one-directional. Where a project has been red long enough to need structured intervention, the health-check-first sequence in the Project Recovery Plan is the usual shape of the response.

The Weekly IT Project Status Report (Excel) derives the RAG from the underlying schedule, budget and risk data rather than from a dropdown, so the colour and the numbers cannot disagree. At portfolio level the PMO Portfolio Dashboard applies one rule across every project in the register, which is the only way a multi-project RAG view stays comparable. The reporting habits that undermine both are covered in the status report guide.

Questions

Does RAG have a fourth colour?

Some organisations add blue or grey for complete, and a few use double-red for critical. Each addition costs comparability, so most keep three states and put the nuance in the direction arrow.

Who decides between amber and red?

The rule decides, and the project manager applies it. The distinction that works in practice is whether the variance can be resolved inside the project or needs a decision from outside it.

Should a project be red on its first report?

It can be. A project starting without a signed-off scope, a sponsor or funded resource is describing a real position, and reporting green because it is early is how that position stays invisible.

How is the overall colour derived?

Usually the worst of the individual dimensions, or the worst that cannot be resolved within the project. Either is defensible as long as the rule is written down and applied the same way each period.

Questions · All 36 templates