Every serious organisation eventually runs into the same question: who decides what, on what basis, and how the choice will hold. Programmes are launched, governance is drawn up, funds are opened, trade-offs commit years. And yet the structure that carries these choices is most often improvised: an inherited organisation chart, a spreadsheet, one more meeting.
It is not a shortage of intelligence: the people in these rooms are capable. It is not a shortage of information: they are drowning in it. What is missing is the architecture. The criteria, bodies, rules and circuits that make a choice well made, then defensible, then repeatable without starting over. What binds an organisation for the long term deserves at least the rigour we give a load-bearing wall.
The Garel Lab exists to structure what is at stake.
Our method rests on three refusals.
We refuse the single point of view.
Every structure we design is confronted with several schools: decision models, game theory, simulation of the actors. When they diverge, we do not hide it, we measure it. Disagreement is not noise. It is the most honest information a leader will receive.
We refuse the untested structure.
Before it is handed over, every structure is attacked: simulated actors looking for the flaw, weightings shaken, adverse scenarios. A governance that does not survive our own assault has no reason to survive a board, a funder or an opposing party.
We refuse the untraceable.
Everything that leaves the laboratory is argued and sourced: a document that can be reopened years later and still defends itself. A structure is not finished when it is drawn. It is finished when it can be explained.
We work for the rooms where error costs more than rigour: institutions whose rules must hold before the law, leadership teams that commit the future of the house, funds and programmes that allocate on the strength of a few criteria. We do not hand them an opinion. We hand them a structure, and the models that make it hold, tested first on our own cases.
Others advise. We structure.