A group programme priced as a single number asks the board a single question, and it is the wrong one. Yes or no, all of it or none of it, in a group where the honest position is almost always that a few markets are ready, a few are close and the rest are nowhere near it. Nobody has a way to say that out loud, because the price sheet has no row for it.
A proposal put to a European composite insurance group did the arithmetic the other way round. Three phases, and on each row the price, the duration and the indicative team in full-time equivalents, side by side. The middle phase, the part that travels into each national operating company, priced per visit rather than absorbed into the programme fee. Everything below is what was offered in a bid document. None of it is a delivered outcome.
The challenge
The work being priced had a familiar shape. At the group centre, build a structured account of where data analytics could move value across the insurance value chain, from underwriting and pricing through claims and servicing. Take that account into each national operating company, test it against local strategy and local conditions, and review what each market would need in capability and resource to act on it. Then aggregate the country outputs into one ranked list and an investment plan the board could act on.
Three months of elapsed time, one centre, several countries. Work of this shape is normally sold as one transformation with one number attached, because the value story is a group story and the fee follows the story. The cost of that habit lands on the buyer. A sponsor who wants the exercise but doubts two of the markets can only negotiate the total down, which cuts depth everywhere rather than footprint anywhere.
The schedule of charges in this proposal had four columns: phase, price, indicative team, timing. Preparation was offered at €67,500, around 2.1 FTE, five weeks, run from the group centre. Exploration was offered at €24,500 per country visit, around 3.2 FTE, one week per operating company. Prioritisation was offered at €37,500, around 2.1 FTE, three weeks. Materials, travel and subsistence sat outside the fee, charged in addition and as incurred.
Printing the team size and the duration next to the price is a small act with a large consequence. It hands the buyer a division they can do in their head, and it exposes where the effort actually sits. The heaviest staffing is not in the analytical build at the centre. It is in the country weeks, at around 3.2 FTE against 2.1 at either end, which is what travel, facilitation and same-week write-up cost. A procurement team reading three rows learns more about the delivery model than they would from ten slides about the delivery model.
The approach
The unit of scope became the country visit rather than the programme. That gives the price sheet a spine and a variable. The spine is the preparation at the centre and the prioritisation at the end, both fixed. The variable is one line, repeated once per market. Adding a market moves one line by €24,500. Dropping a market moves the same line the other way, and nothing else in the structure has to be reopened.
The reduced-scope path was written into the proposal rather than waiting to be extracted in a negotiation. The prioritisation price was stated as resting on an understanding that four to six operating companies would be in scope, a wider band than the indicative plan elsewhere in the document assumed. The exploration phase was described as commencing by operating company, priced on top of the preparation, with an explicit offer to discuss efficiencies from combining the exploration phases of several countries. The supplier raised the volume conversation first. That is not generosity. It is what you do when the cost driver is travel and you would rather the buyer see the driver than suspect a margin.
There is one honest tension in the table worth naming, because it is the sort of thing a careful buyer finds and a careless supplier gets caught on. The priced unit of exploration is one week per operating company, while the visit designed inside that week is three days. The gap is travel, the intake conversation before the team lands, and turning the session into an artefact while the room is still warm. A supplier who prices the three days and delivers the week has mispriced. A supplier who prices the week and says why has just told the buyer what the work is.
The team figure is checkable in the same way. Around 3.2 FTE for a country week reconciles against the visit team described elsewhere in the proposal: a senior insurance and digital practitioner to hold the room, a data scientist to rule out infeasible ideas while they are still being proposed, and an analyst to capture value and effort as the conversation happens. Three named roles, three-point-something FTE. The number is a description of who is in the room, not a rounding of a rate card.
A modular price only stays modular if the paperwork is. Here the offer, the terms of business and the schedule of charges were bound into one signable agreement, with work commencing on receipt of the signed authorisation form, and billing monthly against actual expenditure. A per-country line item inside a contract that must be re-executed every time the country list changes is decoration.
The outcome
This piece describes a bid, and the honest tense matters. Every euro figure above is an offered price, not a delivered outcome and not an invoiced one, and the client signature block in the copy we hold is blank. What the document produced was a commercial structure, and the structure is the thing worth carrying forward.
Two things about it have aged well. The first is that a per-market unit matches how readiness actually varies. Data readiness is not a group property. In any multi-market insurer, one market will have clean policy and claims lineage and a working feature store while the next runs the same value chain through systems that only settle overnight, so straight-through processing is not on the table there whatever the group roadmap says. Nothing in this proposal establishes which markets sat where, because the readiness work that would have answered it was offered and not run. That is the argument for buying exploration one market at a time: you find out in one week and stop, rather than discovering it in month seven of a programme funded on the assumption of uniformity.
The second is that the exercise ends at an investment plan rather than a platform decision. Nothing in the price sheet commits the group to a stack. That ordering still holds: choosing where models will be deployed, and how their inputs and decisions will be traced, is a decision that gets better the more countries have been looked at first.
One thing we would change. The country week in this design is mostly conversation, and conversation is the least reliable instrument for the question it is being asked. Whether an operating company can complete a policy change or a claim without a person touching it is measurable from the workflow logs. Process mining over those events answers it with evidence in the same week the workshop runs, and it settles the argument that otherwise runs on whoever is most senior in the room. When our Consult teams scope multi-market work now, the readiness half of the country week is an evidence exercise and the value half stays a conversation, because value really is a judgement and readiness really is not. The same applies to the cautious language-model pilots now appearing in document handling: whether one is worth funding in a given market depends entirely on whether the answer it produces has anywhere to go, and that is a question about the estate, not the model.
The lesson from the price sheet is simpler than any of it. Publish the unit, publish the team, publish the duration, and let the buyer assemble the programme they can actually defend.