Build the Capability into the Organisation

Colleagues not long in the business often ask: who taught you to work this way?

It is not an easy question to answer. Nobody taught us. It accumulated over eight years, one after-action review at a time. And where should what accumulates be kept? That is the real subject here. Let me state the view up front: capability that grows on a person walks away; capability built into the organisation stays. These are two kinds of firm.

A New Generation of Tools Means a New Generation of Organisational Method

Start with the old rule: better tools of production always change the relations of production. Toyota had TPS and lean production, Motorola had Six Sigma, Intel had OKRs, Haier had Rendanheyi, Amazon had the flywheel. Five companies, five organisational methods, each standing on a generation of new tools. This time it is AI's turn.

So what is the method for the AI generation? We are still writing it. It has a name: the Orchestrator Model, the Glacier Model (still being defined). An orchestrator arranges human judgement, the speed of tools and the sequence of a project into one schedule. In other words, the tools handle fast, and people handle right.

The most practical use of AI is distillation: reproducing at speed the good things you have already seen. But distillation itself needs a view. You need a ruler first, so that you know which one counts as good. Taste comes before tools. Otherwise what comes out is only faster mediocrity. On this we are fairly confident.

People Give the Organisation Its Frame, the Organisation Feeds Them

For us this comes down to two concrete things.

First, put the best young people in the load-bearing positions. Colleagues one or two years in can follow three to five projects through a full cycle, handling financings that run into the billions of yuan. Why? Because eight years of method, sequence and standards have settled into the organisational model and hold them up first. As things stand, the new generation running projects has better backgrounds and more anxiety: they missed the years of high multiples, so their ability is greater and their expectations lower. But anxiety is not a flaw. What the organisation can give is not multiples but method, tempo and positive feedback. The organisation is an amplifier.

Second, turn judgement into a public asset. Price-band data and front-line market feedback are recorded straight away as an organisational asset available to everyone, rather than staying in one person's head and coming and going as that person does. The plainest component of the Glacier Model is the Stop Doing List (the subtract-only list): every pit worth avoiding written down as a prohibition, reread by everyone at every review. Methods can iterate; boundaries only grow. The list only takes entries; it never gives them back.

But Can It Be Carried Away?

Someone will ask: if the methods are all written out in the open, what happens the day a colleague leaves? People do leave.

But that worry is worth turning around and costing out. If one person really can carry an organisation's resources away, then those resources were growing on people all along, not in the organisation. The fear is in fact a very accurate diagnosis. What does it measure? Where your things have settled. Put the other way: if the resources really sit on the organisation's side, people moving is no threat. We do not need to keep a guard up.

This is also why the partners stay on the front line running projects. Judgement and relationships both start to dilute the moment you step back behind the scenes. This is not a guard against anyone; it is that only the front line grows things worth keeping. The bosses sit at the table.

In this round of tool change we do not only want to use the tools well. We want to answer the organisational question behind them. That question has a name. Next piece: the name.