Status Reporting That Does Not Contradict Your Own Register
The purpose of a status report is to let somebody decide something. Most of them are written to demonstrate that work occurred.
A status report exists so that a person who is not doing the work can decide whether to intervene. Everything in it should serve that decision. Most reports serve a different and unstated purpose, which is to demonstrate diligence, and reports written for that purpose are long, reassuring, and unread.
The diagnostic is simple. If your weekly report has never caused anybody to do anything, it is not a management instrument, and the hour spent producing it each week is a cost with no return.
The contradiction problem
The most damaging failure is not a poorly written report, it is a report that disagrees with the firm's own records. The key risks section says three things; the register says eleven, two of which are severe and absent from the report. The report says amber on schedule; the plan shows a milestone that slipped a fortnight ago.
This happens for a structural reason rather than a careless one. The report is written by a person recalling the week, and the registers are updated by different people at different times. Retyping is the mechanism of drift, and drift is guaranteed given enough weeks.
The fix is that the report selects rather than restates. The key risks are chosen from the live register, so they carry the register's current score and cannot be stale. The schedule position is read from the plan. The commercial position is read from the ledger. The narrative is written by a human, because judgement cannot be assembled, and the facts are assembled, because judgement is not required to copy a number correctly.
What belongs in the report
- The overall position, as a rating, with the reason for it in one sentence. A rating without a reason invites a conversation about the colour.
- What changed since the last report. Not what happened, what changed, because a reader who read the last one needs the delta.
- What is now waiting on the client, which is usually the most actionable thing in the document and is usually buried.
- The three risks that would most change the outcome, selected from the register rather than recalled.
- The commercial position against plan, including approved changes, because the fee has usually moved and the reader may not know.
- The decisions being asked for, stated as decisions with the options, not as topics for discussion.
What does not belong
A list of activities completed does not belong. It is the largest section of most reports and it serves the author rather than the reader, who does not need to know that four workshops occurred unless something came out of them.
A rating with no threshold does not belong either. Amber means nothing unless the reader knows what would make it red, and firms that do not define the thresholds produce ratings that vary by author rather than by engagement. Two managers reporting the same situation should produce the same colour, and without written thresholds they will not.
Anything the reader cannot act on and would not act on is a candidate for removal. That test alone usually halves a report.
The internal and the client version
These are two documents with two audiences and they should not be the same file with rows hidden. The internal report can discuss margin, resourcing difficulties, and the fact that the client sponsor is the obstacle. The client report cannot.
The mechanism that keeps this safe is marking audience at the point of writing rather than filtering before sending. An item written as internal stays internal, and the client version is assembled from what is marked for the client. Filtering at the moment of sending depends on somebody remembering, every time, under time pressure, and the failure mode is a single paragraph reaching a client that should never have left the firm.
It is also worth being explicit about what makes a report visible to a client at all. Issued, addressed to the client, and not marked confidential is a reasonable set of conditions, and stating them plainly is better than relying on a convention, because a convention is what breaks when somebody new writes the report.
Cadence and length
Weekly is right for most engagements in delivery and wrong for long research phases where nothing changes week to week and the report becomes an exercise in finding something to say. Match the cadence to the rate at which the answer to should we intervene actually changes.
On length, one page for the client and two for the firm is a reasonable discipline. The constraint is useful precisely because it forces selection, and selection is the work. A report that grows without limit has stopped being edited.