Reporting Progress
Most status reports are written for the writer — to demonstrate diligence, or to have something to point at later. A report worth the time it costs is written backwards from what the reader has to decide.
In short
- Lead with a verdict: will we make it, and should you worry. Everything else is supporting detail.
- Report what changed, not everything that exists. A tool already lists what exists.
- Every report should ask for exactly one thing, or it is a broadcast.
- Bad news goes first and goes early. Credibility is built on the reports where something was wrong.
What reporting is for
Three jobs, and they need separating because a report trying to do all three at once does none of them:
- To let someone decide. The sponsor needs enough to approve, redirect or leave alone. This is the primary job and it usually needs under a hundred words.
- To coordinate. Neighbouring teams need to know what changed that affects them. Better served by a notification when it happens than by a weekly summary.
- To create a record. A trail of what was known when. Valuable, and mostly a by-product of doing the first two consistently rather than a reason to write anything extra.
Notice that none of them is “to show that we’ve been busy”. That instinct is understandable and it produces the reports nobody reads, because effort is the least useful thing you can report.
The shape that works
Status labels that mean something
Green, amber and red are near-universal and nearly useless, because nobody agrees on what they mean. Two rules fix most of it:
- Define them once, in terms of the date. Green: will hit the date with the plan as it stands. Amber: will hit it only if something specific happens — and say what. Red: will not hit it without a decision from someone other than the team.
- Amber requires a named action. An amber with no ask is a green with anxiety attached. If nothing is required of anyone, it’s green; if something is, say what and by when.
The failure to watch for is the watermelon report — green on the outside for months, then red without ever passing through amber. That pattern doesn’t mean people are lying; it usually means the definitions were never agreed, so “green” drifted into meaning “nobody has escalated”.
Numbers and words together
Quantitative data says what; qualitative says why it happened and what it means. Neither works alone. “62% complete” tells a sponsor nothing actionable. “It’s going well” tells them less.
Two habits keep numbers honest in a report:
- Pair every number with a comparison. “3 of 8 buffer days used” beats “3 buffer days used”, because the reader shouldn’t have to remember the denominator.
- Prefer countable things. Deliverables accepted, days of buffer left, items blocked. These can’t be adjusted by optimism the way a percentage can — see 6.1.
Adjusting for the reader
| Reader | What they actually want | What to send |
|---|---|---|
| Sponsor | Will we make it, and is anything needed from me? | The card above. Weekly, same day. |
| Steering group | Decisions taken, decisions needed, money | Monthly. Exceptions and choices, not activity. |
| The team | What changed that affects my next task | Continuous, on the task itself. |
| Neighbouring teams | Anything touching their dates | An event when it happens, not a weekly digest. |
| End users | When it arrives and what changes for them | At milestones. Silence in between is fine. |
Two of those five rows are events rather than reports. That’s the useful realisation: a lot of what gets bundled into a weekly report is really a notification that should have gone out on Tuesday when the thing actually happened.
Delivering bad news
This is the part that decides whether anyone trusts the other reports, and the mechanics are simple enough to write down:
- Same day, not next Friday. A slip that a sponsor learns about a week late is two problems: the slip, and the week.
- The sponsor hears it from you first. Hearing it from someone else converts a manageable problem into a relationship one.
- Lead with it. Burying it in paragraph four teaches people to read your reports sceptically for the rest of the project.
- Bring the options. Problem, impact, two or three courses of action with their costs, and your recommendation. That turns a five-day email exchange into a five-minute decision.
- Say what you already did. Not defensiveness — information. It tells the reader which options are already exhausted.
The general principle: a report that only ever says things are fine carries no information, because it would say the same thing either way. The value of your green weeks comes entirely from having reported the amber ones honestly.
How this looks in AB
In AB Projects, status, progress and the project dashboard make the actual state visible without anyone writing a weekly report. Adaptive Cards in the linked Teams or Slack channel deliver “what changed this week” automatically, and when someone asks “when did this slip?”, the change-history tab is the audit trail with the answer.
That changes what a written report is for. When anyone can see the raw state at any time, the report stops being a data transfer and becomes the two things a dashboard genuinely cannot do:
- Interpretation — what the changes mean for the date, which a list of task movements does not say.
- The ask — the one decision you need from the reader this week. No dashboard has ever requested anything.
Which is why the card in the diagram is short. Almost all of it is the part a tool can’t generate.
Terms used on this page
- Status report
- A periodic summary of progress, risks and decisions needed. Valuable when it interprets and asks; wasted when it lists what a tool already shows. More →
- RAG status
- Red, amber, green labelling of project health. Only meaningful if the three are defined in terms of the date and agreed in advance. More →
- Watermelon report
- A status that stays green on the outside while the project goes red underneath, then flips without warning. A symptom of undefined labels rather than dishonesty. More →
- Exception report
- A report sent only when something crosses a threshold, rather than on a schedule. Efficient, and dependent on the thresholds being agreed. More →
- Dashboard
- A live view of project state that people pull from. Replaces the data-transfer part of reporting and none of the interpretation. More →
- Variance
- The gap between plan and actual, in time, cost or scope. What a report exists to explain rather than merely display. More →
Related reading
4.3 Communication Planning
Who gets what and how often — the decisions this page assumes have been made.
6.1 Monitoring Deliverables
Where the countable numbers in a good report come from.
4.2 Stakeholder Management
Why regular, boring, honest reporting is the thing that actually builds trust.