Productising a Service That Is Not Supposed to Be a Product
How our organizational assessment method works: hundreds of voices coded, a framework emerging from the evidence, findings wired into strategy as experiments.
The problem with selling judgement by the day
Enterprise coaching is a hard thing to buy and a harder thing to sell. It is sold by the day, the days are finite, and most of them do not go where the client thinks they go. They go on legwork: reading several hundred survey comments, coding them, counting them, cross-checking them, drawing the charts, building the deck. The part the client is paying for, someone experienced sitting with their leaders and telling them what is going on and what to do about it, arrives last, when everyone is tired.
The standard way out is to package the front end as an assessment. Forty questions, a maturity score, a benchmark, a heat map. It scales very well, and it has one property that does not go away: ask forty questions and you get variations of forty answers. The instrument finds what its author decided to look for, and what the author decided to look for is usually what the author sells. Ask for a team-agility score and you will get one, whether or not the team is where the problem is.
We wanted the opposite on both counts. An assessment that starts from the business and sees what surfaces. And a way of delivering it that spends the consultant on the client, not on the spreadsheet.
The bet: split the work
The decision was to draw one line through the engagement. Everything labour-intensive goes on one side and is automated. Everything that depends on trust, presence and judgement stays on the other side and stays human: meeting the client, the stakeholder interviews, facilitating the session where a leadership team agrees what it is actually trying to do, the coaching that follows.
Today most of the labour-intensive work is automated, and the time it used to take goes to the high-touch side. The point of the model was never to remove the consultant. It was to give the client more of one.
How an engagement runs
No single step below is the point. The point is that it is one chain, and it runs the same way every time.
Open questions in. We ask few questions and leave room to talk: short surveys of scored questions with a comment box beside each, and open interviews with leaders. The scores tell us where to look. The comments and the interviews tell us what is there.
The words are coded, not skimmed. Four steps run over everything everyone said: meaning extraction, which breaks a comment into the separate things it is saying; sentiment analysis; qualitative scoring; and framework coding, which places each statement in the reference models it bears on. A mid-sized engagement yields around five hundred coded statements beside some fifteen hundred scores; a large one, several times that. This is ordinary qualitative research method, the kind a research team does by hand and stops being able to do somewhere past a hundred voices. Every published number is reconciled against its source, and the client’s name is removed at the door, so nothing downstream ever sees it.
The framework comes out of the data. A survey goes in with the dimensions we designed. It comes out with more. In one engagement about half as many again surfaced from what people chose to write, themes nobody had asked about: ownership vacuums, duplicated work, the different speeds at which parts of the organisation lose and replace people. In one case the best-scoring capability in the whole survey turned out to hide a single point of failure, and it was found because of a note respondents had written into their own comments (the team-resilience study). Nothing in the numbers asked for that page. The words did.
The organisation becomes a map. The dimensions, their scores, how often people raise them and how they feel about them become a network: what moves with what, and where the pressure collects. We call this picture the system, and it is what the rest of the engagement works on. The tooling holds one line of research discipline on our behalf: a correlation is never written up as a cause. The network shows what moves together. Why it does is a judgement, made by a person, with the client in the room.
One body of evidence, many lenses. Because the evidence is coded once, it can be read through any reference model without being collected again. Each statement carries its place in several at the same time: the agile principles, the scaled frameworks, Scrum and Kanban, and the seven alphas of the Essence kernel (opportunity, stakeholders, requirements, system, work, team, way of working). A client who runs SAFe sees their organisation against SAFe. A client who runs nothing in particular sees it against Essence, which was designed to be neutral about method. The assessment measures once and lets the evidence choose the lens.
The strategy, played back. Leaders can nearly always state the goal, not always how to get there. From the stakeholder interviews we build a strategy map on the four classic perspectives and play it back as four lines of plain text. For one client, a business with a growth target everyone could quote, it read in substance: by fixing how we organise, hire, train and keep people, we enable better decisions, predictable delivery and an efficient operation, which gives us the capacity to serve more customers, in more places, at higher quality, which is how the growth target is met. Under each line sat the strategic objectives to execute against. The reaction is nearly always the same: the leadership team recognises it at once, and had not been able to say it themselves.
Findings become experiments. Each significant finding is wired into that map. Technical debt, to take a common one, is not a finding on a slide. It is a process-level objective that feeds delivery predictability, quality and speed, which feed the customer outcomes, which feed the number at the top. Each becomes a change experiment with a hypothesis, a leading indicator and lagging indicators, linked to backlog items a team can start on Monday.
Every link is traceable. A line on the board’s strategy map connects, step by step, to a sentence someone typed into a comment box. When someone asks “who says?”, there is an answer.
Then the map is updated. An engagement is a loop, not a report: make the change, measure, keep it or drop it, and assess the system again.
The part that stays human
None of this replaces the conversation. It earns it. The machinery cannot sit across from a chief executive and notice what they avoid saying. It cannot hold a leadership team in a room until they agree what the objective really is. It cannot judge that a correlation is a cause, or decide which of six true findings this organisation can bear to hear first. Those are the consultant’s work, and they are the reason to hire one. They used to be what was left over after the legwork. Now they are the engagement.
A product, not a person
Putting a repeatable pipeline under the engagement changed what the service is. It has defined stages, defined outputs and a defined hand-off between the automated and the human parts, which means it can be delivered by people other than its author. It is offered direct, and white-labelled through channel partners who deliver it under their own name. For a partner firm the report is a beginning: every change experiment on the strategy map is a piece of work somebody has to do.
The same engine, pointed elsewhere
The platform this is published on is the same method pointed at a field of knowledge instead of an organisation: gather widely, let the structure emerge, keep each fact in one place, render many views from it, keep a human signature on what goes out. The research series, Measured in the field, is the first series of studies published from this work, anonymised: samples of what the assessment’s analysis stage produces. More follow as further engagements are written up. How the operation underneath was built and run is its own story, told in the field report.
A journey, not a destination
The harness that runs the platform, agents working under review with a person signing what matters, became stable this year, and it is only now being turned on the consulting archive: years of engagements and recorded sessions, enough material for a hundred or more studies of the kind the research series holds today. Assessing the same organisation again over time, so that a theme can be watched moving, is the natural next step. The direction is the one that has held throughout: whatever becomes routine is pushed down to the machine, and the attention that frees up moves to the next hardest thing.
Limits
The figures quoted here are illustrations from individual engagements, rounded, not benchmarks. The network shows correlation; every causal claim in a client report is a human judgement and is presented as one. And this is an account by the person who built it, which makes it checkable and also partial.
If this is your problem
Two kinds of people tend to write to me. A leadership team that can state the goal but not the path. And a firm that wants to offer this kind of assessment to its own clients, under its own name. Either way, write to me.