From 2,000+ engineers billed by the hour to a catalogue a client can buy.
A large engineering services organization had deep expertise in five areas and almost no way to sell it as anything other than people. This is the operating model that changed that — and the instrument that keeps it honest.
- Organization
- 2,000+ engineers, five practice areas
- Role
- Designed the operating model; leads the Infrastructure CoE
- Span
- 2016 → now; the offering model from 2025
- Status
- Running — the model is in use and published as a live instrument
Deep expertise that a client could only buy by the hour.
- Expertise existed, but a client could only buy people by the hour — never an outcome.
- Two practices could claim the same work, and nothing settled who owned it.
- A client problem bigger than one practice had no route through the organization.
- Ideas for new offerings arrived from everywhere and were judged by opinion.
- Nobody could say, on evidence, how mature the commercial organization actually was.
One accountable name at every level.
Five practice ecosystems under a single portfolio. Capability areas belong to exactly one practice; offerings are rooted in a capability. Nothing floats.
The connected ecosystems, and the portfolio across themCoE Head Portfolio strategy, investment, lifecycle governance, ownership clarity
One practice area — Infrastructure, Quality Engineering, and so onCapability Owner Specialist expertise, standards, reusable capability assets
Named expertise, belonging to exactly one practiceCapability Owner Depth, knowledge, the standard the work is held to
A commercial proposition rooted in a capabilityOffering Owner Positioning, roadmap, assets, proof, performance — for one offering
Three layers, nineteen stages, one artifact each.
The layers run in parallel, not in sequence. Each stage hands over exactly one named thing — which is what turns a diagram into an answer to "what do I produce next?".
Offering Lifecycle
Is this worth building, and should we keep investing in it?
6 stages- IdeateOffering Idea Brief
- ValidateValidation Scorecard
- ProductizeOffering Definition Pack
- LaunchLaunch Plan
- MeasureOffering Health Dashboard
- Evolve / RetireLifecycle Decision
CoE Sales Motion
How do we take the capability to market?
6 stages- Spot itClient Signal Library
- Open itConversation Starter Kit
- Position itValue Proposition
- Scope itOpportunity-to-Offering Map
- Prove itProof Pack
- Grow itGrowth Map
Cross-CoE Solutioning
What happens when the client problem is larger than any one offering?
7 stages- DetectCross-CoE Opportunity Candidate
- FrameClient Problem & Outcome Brief
- AssembleCapability Composition Map
- ShapeIntegrated Solution Blueprint
- CommitSolution Governance Model
- DeliverIntegrated Outcome Dashboard
- ExpandExpansion Map
Four kinds of decision. One owner each.
Lifecycle, positioning, methodology, roadmap, assets
Investment, overlap, merging, retirement, ownership disputes
Architecture, cross-workstream trade-offs, integrated scope
Standards, specialist methods, reusable assets
Programmes like this stall in year two because people stop arguing about the work and start arguing about who has the right to stop it.
Eleven questions an idea has to survive.
Each is scored on evidence — none, weak hypothesis, some, strong — so a "no" is a reading, not a mood.
Seven levels, each one checkable.
Not a maturity model to be scored in a workshop — each level states a condition the registries either meet or do not.
The model does not live in a document. It is published as a working instrument that reads the registries and reports what is visible from them. "This level is not visible from here" is a statement about the data, not a verdict on the organization — which turns an argument about the score into a question about what to write down.
Built, and running.
- Expertise in five practice areas is expressed as offerings a client can actually buy.
- Every offering has one named owner, a lifecycle state, and a stop condition.
- Four classes of decision have exactly one owner each — disputes are about work, not authority.
- New ideas face eleven written questions and an evidence scale, not a committee mood.
- Cross-practice client problems have a route, with named hand-offs and one artifact per stage.
- Maturity is read from the registries rather than asserted in a deck.
If you are trying to move an engineering organization from selling people to selling outcomes, this is the work I do.
Scope a transformation →