Kapari
The bench, taken apart

When everyone says yes, our engine raises a flag.

A board that applauds stops asking questions. That is exactly when bad decisions get through. Kapari does the opposite: the wider the agreement, the harder it pushes to find out how this could still fail.

iThe premortem fires on its own. You do not have to ask for it.
What the research says

It is a year from now. The project failed. Why?

In 2007 the psychologist Gary Klein published a brutally simple method in the Harvard Business Review. Before launching a project, gather the team and state one thing: it is a year from now, the project failed, it was a disaster. Everyone writes down why.

The shift is tiny and it changes everything. Asking "what are the risks" invites everyone to reassure themselves. Announcing failure as a done deal forces them to explain it. Klein measured that this flip surfaces far more reasons, especially the ones nobody dared to say.

The problem he targets has a name: once a group agrees, it stops looking. The doubts stay in people's heads. Nobody wants to be the one who kills the mood.

How it is coded

The wider the agreement, the harder the engine pushes.

Kapari measures how much the panel agrees. When support runs wide, it does not simply hand you good news: it triggers the premortem and shows you how this could still go wrong.

It is automatic. You do not request it, and you cannot skip it. That is deliberate: a safeguard you have to remember to switch on is useless, because you only remember it when you already have doubts.

The premortem is not written out of thin air. It starts from the panel's own voices and from what holds them back, including the ones who ended up saying yes.

What this changes for you

The day your board applauds, you still get the list.

You know the scene. You present, everyone nods, the meeting ends early. Three months later someone tells you they had a doubt but did not want to break the momentum.

The engine asks in their place. You leave with the possible reasons for failure, written down, even though the panel was in favor.

And you stay in charge: those reasons are leads to check, not a verdict. What you do with them is your call.

The source

The work this choice rests on.

Klein, G. (2007), "Performing a Project Premortem", Harvard Business Review. Klein also established, with Daniel Kahneman, the conditions under which expert intuition is reliable and the conditions under which it is not: that work is implemented elsewhere in the engine.

The bench, taken apart

More from this series

A good decision can end badly. So what do you grade?

Our tool tells you when not to trust it