Every fixed-price proposal has one number that decides whether the price is survivable, and it is almost never the price. On this one it is a shape. The most expensive person on the team is booked for five days. The cheapest is booked for seventy. Between them sits a design lead at thirty-nine. Read the day counts against the rate card and effort runs exactly inversely to seniority, with a spread of roughly 2.6 times from the top rate to the bottom.
A European travel and tourism group had just completed a structural blueprint for a newly created digital function. The next phase was the work of turning a structure chart into something an organisation can actually run on: the processes underneath it, the people mapped onto those processes, and the governance that keeps both honest. Roughly thirty processes described at a working level of detail, five of those taken a level deeper, role mapping at two tiers of the hierarchy, a competency model, a governance cadence, and a prioritisation process for the idea portfolio. One quarter. One fixed fee.
The challenge
That scope does not fit a quarter at any sensible rate if the work is invented from scratch, and the proposal says so before it says anything about money. The delivery philosophy is stated first: operate at pace by building on the work the client had already done internally, plus a pre-built library of process templates and reference flows. The fee buys assembly and judgement rather than invention. That is the honest version of the pitch, and the part that makes the arithmetic possible.
The second condition is stated just as plainly and is more interesting. Continuity of the people who ran the previous phase was made an explicit part of the offer. Context, not capacity, is the scarce input on operating-model work. Whoever sat in the rooms where the blueprint was argued knows which decisions were settled, which were parked, and which will be reopened the moment someone writes them down wrongly. That knowledge does not transfer through a handover document, and rebuying it costs more than any day rate on the slide.
So the team has two jobs that pull against each other: hold the context, which is expensive, and produce volume, which at senior rates does not fit a fixed fee. The proposal separates them and staffs them at different grades.
The approach
The senior tier is booked for five days across the whole quarter. That is sponsorship and assurance: presence at the moments where a decision needs weight behind it, and nothing else. It is under a tenth of the working days in the window.
The design and workshop lead is booked for thirty-nine days, which is more than half the assumed window. This is the tier that carries the client relationship, runs the sessions, and makes the design calls on the global operating shape and the structure beneath it. It is also the tier where continuity was made a condition, because it is the tier holding the previous phase in its head.
The production tier is booked for seventy days. That tier writes the process flows, produces the role descriptions, and turns workshop output into artefacts people can be trained on. It is the newest addition to the team and it carries the largest single share of the effort by a wide margin.
Put the day counts next to the rates and the mechanism is visible. The blended rate across the engagement lands about 1.4 times the bottom rate and a little over half the top one. It sits near the floor of the spread, not in the middle, because the tier with the most days has the lowest rate. Without that weighting, a fixed price against thirty processes in a quarter does not close. Flatten the pyramid and the same 114 days becomes a number nobody signs.
Which is also where the quality risk lives, and the proposal does not pretend otherwise. The client receives most of its deliverables, by volume, from the tier with the least context and the least tenure on the problem. The design decisions are made by people who were in the room. The written record of those decisions, which is what everyone downstream will actually read and be trained against, is made by someone who was not. Every gap between what was decided and what got written is a gap that surfaces months later, when a process owner reads a flow and does not recognise the work it describes.
There is a scheduling tell here too. The proposal assumes a fourteen-week window, which is seventy working days, and it books the production tier for seventy. Every working day in the window is sold before the work starts, on exactly the tier carrying the most, with nothing held back for a public holiday, a slipped workshop or a day of illness. The two dates the window runs between enclose eighty-nine calendar days, which is twelve weeks and five days rather than fourteen, so the buffer is not merely thin, it has already been spent. Slack that is not on the page does not exist, and it is worth counting in any fixed-scope proposal before signing rather than after.
The uncertainty is handled the same way the effort is: by shaping it rather than by hoping. Forty-five further days sit outside the core price as a dated option, twenty-five to take five key processes to the deeper level and twenty to work through the headcount and role data properly. That is thirty-nine percent of the core day count held in reserve, priced up front, with a fixed date by which the client must decide. It is a better instrument than an open change request, because both sides know the number and the deadline before anyone commits.
The two option rates match none of the three core rates, which quietly implies two grades not otherwise on the engagement. A reserve priced at grades you have not met is a reserve you cannot fully evaluate.
The outcome
This is a proposal. It produced a priced plan, a staffing shape and a dated option. It did not produce a delivered operating model, and nothing here should be read as one.
What survives the era is the shape, and two things we would do differently with it now.
The first is where the production days go. The scope rations detail deliberately: about thirty processes at a working level, five taken deeper, and those five are mostly the ones that did not previously exist. That ration is correct in principle and expensive in practice, because the bulk of the production budget still goes to documenting processes that already run. Inherited processes do not need to be interviewed into existence. They are already written down, in the event logs of the systems that execute them. Process mining over those logs produces the working-level description with a source attached, and it frees the workshop budget for the processes with no incumbent to observe. Our Consult engagements now start with that separation, because it moves days from transcription to design without changing the price.
The second is what happens to the artefacts afterwards. An operating model that lives in a slide pack decays from the day it is approved. The same flows, roles and gating steps can sit in the systems that carry the work, with the lineage to show which version of a process produced which decision. That is Platform work rather than document work, and it is the difference between an operating model people were trained on once and one they are running.
Neither of those changes the finding. Effort runs inversely to seniority on this kind of work, it has to, and the price depends on it. The discipline is knowing that the inversion buys affordability by moving the written record down to the tier furthest from the decision, then closing that distance deliberately rather than assuming it away.