Published 31 January 2024 · Updated 11 October 2026
On the operating-model redesign of a 10,000-person business and technology services company — 22 business lines, sitting underneath a major cost-reduction and reinvestment program — the request in the first workshop was the same one I get almost every time: show us the new org chart.
I held it back for months. That sequencing is the single biggest reason the model was still standing a year after we handed it over.
For many people the bottom line for operating model design is the org chart. That is exactly why it is the thing everyone wants answered first. When I’ve led this work, the org chart has been the last artefact I produce, deliberately.
Operating model design is primarily about answering how to get the flow of value to your clients as efficiently as possible, while making the best use of your capabilities — the strengths of your people, the assets you already own — in the context of the business strategy and the environment around it.
There are a lot of questions and scenarios to work through before you get anywhere near “and what grouping of people and accountabilities would best support that?”
Tensions are the heuristic
The question I keep coming back to is: where are the tensions in this operating model? Which are bad tensions that need to be designed out, and which are good tensions worth keeping?
When I led the target operating model for a national smart-meter data operator serving more than 30 million homes and businesses, the useful arguments were never about boxes. They were about where we deliberately accepted friction — a handoff between two teams, a checkpoint before a release — because removing it would have cost us something more important. Deciding those on purpose is most of the design work.
That tension map is a good Pareto heuristic for whether you have done enough design, and it is worth revisiting once the org chart is done too. It is also the artefact that makes the org chart defensible when someone challenges it.
What it cost us when we got the sequence wrong, on earlier programs, was rework: a structure agreed in week two, and a value flow that turned out not to fit it, so the structure had to be unpicked in front of the people who were about to be moved into it.
If you’re the accountable executive
- Don’t approve an org chart until the value flow, the capabilities you are betting on and the deliberate tensions are written down and agreed. If you only get one of those, get the tensions.
- Ask the design team to show you the options they rejected and why. A single recommended structure with no visible alternatives usually means the design work hasn’t happened yet.
- Set the sequencing in the mandate, in writing. Say explicitly that the structure is the last decision, then hold the line the first time it gets uncomfortable — it will.