The Rumsfeld matrix has four quadrants. The top two, what we know we know and what we know we do not know, are the comfortable ones. The trouble is in the bottom two. The unknown knowns quadrant holds tacit knowledge, suppressed findings and unstated assumptions. The unknown unknowns quadrant holds the real surprises: shocks at the level of the system, equity dimensions nobody thought to disaggregate by, and second-order consequences in neighbouring sectors.
Most programme design happens in the top row, and most evaluation failures come from the bottom row. This note is about the meeting that puts the bottom row on the page before launch, which is the pre-mortem.
What a pre-mortem is
A pre-mortem is a post-mortem for a project that has not yet failed. The format is simple. You gather the team in a room and give them this instruction: “Imagine it is eighteen months from now, and the project has failed. Write the obituary.” You set a timer and give everyone twenty minutes to write a draft alone, on paper.
Then each person reads theirs aloud, and nobody argues.
That is the whole exercise, and the structure does most of the work. People who would not raise a concern in a planning meeting will write it into a failure scenario, because the scenario gives them permission. The exercise does not ask whether the project will fail. It assumes that it has and asks what caused the failure.
Why the obituary is the unit
Asking people to write an obituary instead of listing risks has three effects.
The first is that the cause of death becomes concrete. A risk register says “stakeholder misalignment, medium probability, medium impact”. An obituary says “the District Collector transferred in month four, the new one had been briefed against the programme by the previous contractor, and we never recovered the access we lost in that meeting”. You can plan against the second. The first only gets filed.
The second is that the obituary forces a sequence. Real failure happens on a date, while a risk list has no dates. An obituary makes the team commit to when the failure begins, and that usually reveals a period the plan had assumed was safe.
The third is that the obituary makes it less costly socially to raise a concern. In a planning meeting, naming a risk means telling a colleague their optimism is wrong. In a failure scenario, naming a risk helps the team see something that nobody is defending yet. The pre-mortem turns a confrontation into shared work.
How it fits the matrix
Used with the matrix, the pre-mortem fills the bottom row directly.
Unknown knowns appear in obituaries that begin “we always knew that…”. This is tacit knowledge the team had and had not voiced. It includes the frontline staff who had run the programme before and were not invited to the design meeting, the findings from the last evaluation that were politically inconvenient, and the assumption built into the choice of district that nobody had checked against the political map.
Unknown unknowns appear in obituaries that begin “what we did not see coming was…”. These are surprises the team had not modelled, such as a consequence in another sector, a shock in a neighbouring system, or a population the sample frame could not reach. They are harder to bring out, but the pre-mortem still helps, because imagining failure makes the team look further ahead than the planning document did.
When to run it
There are three good moments.
The first is before the proposal is finalised. A pre-mortem at the proposal stage costs an afternoon and may stop the programme being designed around an assumption that fails in month three.
The second is before the evaluation is commissioned. A pre-mortem on the evaluation itself (“imagine the evaluation produced findings nobody used; why?”) is the most underrated meeting in MEL practice. Most evaluations fail at the point of use, and those failures can usually be anticipated in a pre-mortem.
The third is before scale. A pre-mortem on the scale plan (“imagine the pilot worked but the scaled version did not; why?”) catches assumptions carried over from a context that no longer applies.
The artefact
A pre-mortem produces a single sheet and no report. The sheet is the bottom row of the Rumsfeld matrix, filled in, and it stays pinned above the desk for the life of the project. Each line from the obituaries becomes either a planning change, meaning the team will now design against it, or a watchpoint, meaning the team cannot prevent it but will notice if it starts.
The Knowledge Map canvas at /canvas/rumsfeld-matrix is designed to hold this output. Print it, fill in the bottom row during a pre-mortem session, pin it up, and return to it in month three.
In most failed projects the cause of failure was visible early, often with a date attached. The pre-mortem gives the team a chance to see it before the project ends.