Process · The ODEG method
Every founder-led business reaches the same wall. The thing that got you here — the founder being the system — is the thing keeping you stuck.
When founders describe this, they usually describe a feeling, not a structure. ‘Everything still comes through me.’ ‘Nothing happens unless I push it.’ ‘We are growing but it is getting harder, not easier.’ Underneath the feeling is something tangible: there is no operating model, and the founder is being asked to be one.
What an operating model is, in practice.
An operating model is the set of answers a business has to a small number of unglamorous questions:
- Where does work begin, and where does it end?
- Who owns each handover?
- What is the default decision when nobody escalates?
- What is automatically true (a trigger fires, a number updates) versus what is true only because someone remembered?
- What is the right place to find out the answer to the most-asked question of the week?
When the answers are clear and the business runs on them, the founder gets to step back into the role they actually want — strategy, customers, hiring. When the answers are implicit and live in the founder’s head, every growth step makes the founder more, not less, central.
Why AI changes the maths.
Until recently, building a real operating model in a small business was prohibitively expensive. You needed a layer of middle management, internal tooling, integrations and dashboards that the unit economics did not support.
AI changed the maths. The cost of a senior-feeling internal tool collapsed. The cost of an integration between two systems that do not natively talk collapsed. The cost of an always-on chase for any repetitive task — invoices, leads, follow-ups, reordering, reporting — collapsed. A founder-led business can now afford the operating model it could not afford 24 months ago.
How ODEG runs this.
Five steps, in this order, every engagement.
- Audit — a clear-eyed read of the business: where it works, where it leaks time, money and momentum.
- Roadmap — a prioritised plan, with the highest-leverage moves separated from the noise.
- Build — AI workflows, automations, dashboards and integrations built around the systems already in place.
- Implement — roll-out, training and operational hand-over so the change actually lands with the team.
- Improve — continuous refinement: tighten what is working, retire what is not, extend what is useful.
Each step has a tangible output. Each step is timed in weeks, not quarters. The whole engagement is shaped so the founder is doing less by the end of it, not more.
More from the journal
Field note · Sip Champagnes
The numbers, before the story.
What happens when you redesign how a business works and then build the AI and automation to make it real. Audited management accounts, nine months in.
Read entryEssay · Operating model
Why most automation projects do not move the P&L.
After enough engagements, a pattern: businesses spend on AI and automation, see operational improvement, and then watch the bottom line refuse to move. Here is why.
Read entry