Foray 246 · Essai 1 · 24 October 2026
The director at the handover door
The thing Mara is good at may be hiding the thing she could change.
Mara Venn directs a consortium’s six small museum retrofits across a coastal county. Heat pumps, roof repairs and improved collection stores must fit around school visits, fragile objects and brief winter closures. She is trusted because she can make a contractor’s “finished” mean something usable to a conservator and a caretaker. Her week is full of exactly that translation.
On Tuesday, a roof package is complete except that the humidity logger has no named owner. On Thursday, a new plant room is ready but the weekend custodian has never rehearsed an alarm. Neither problem belongs neatly to the construction package. Mara makes a call, repairs the handover and keeps the opening date. Three towns away, the next museum will need the same conversation.
Her apparent problem is lack of time. Her possible opportunity is that the same acceptance boundary is being rebuilt by hand on every scheme. A reusable rehearsal—people, evidence, operating exceptions and a clear account of what has not been accepted—might improve delivery and create a bounded service. But a recurring irritation is not yet demand, and a useful service is not necessarily Mara’s business to own.
A good present can be a trap
The present arrangement rewards everyone locally. Contractors finish their packages; museum managers protect opening dates; the consortium sees a reliable director. Mara’s rescue work absorbs the disagreement between those definitions of success. Removing her from the loop tomorrow could make things worse. The system will not change simply because another diagram is more attractive.
We therefore look for a nearby arrangement: something she could reach using relevant access and capability, with a transition small enough to try. A “partial equilibrium” here means a provisional pattern of work that could continue while selected surrounding conditions stay fixed. It is a heuristic. This essai does not estimate market supply and demand, prove strategic best responses or establish a Nash equilibrium.
An inspectable opportunity landscape
Six ways the work might settle
Choose a possibility, then compare it across the three views. These are authored ordinal judgements from 1 to 5, not measurements, currency or probabilities. The pale marks show a one-band sensitivity span, not statistical confidence. Straight lines connect categories for reading; they do not estimate the difficulty between them.
Changing the example only changes this page’s temporary view. Nothing is saved, sent, purchased or commissioned. Reload returns to the opening assumptions.
How to read the dimensions
Reachability: existing access and capability, higher means closer. Net working reward: an assumed blend of useful outcomes, sustainable earning and a better working week after routine effort; it deliberately remains a contestable judgement. Reuse: how much the method could travel across schemes. Setup burden: effort and commitment before useful delivery, higher means more burden. Other actors: reliance on agreements beyond Mara’s control, higher means more reliance.
No weighted total is calculated. The dimensions are not commensurate, and a high point is neither a recommendation nor an expected return. Inspect every number and counterfactual rule.
The neighbouring opportunity
B, one acceptance rehearsal, is the smallest useful disturbance. Before one museum’s handover, Mara brings the contractor, custodian and collections lead into the same room. They rehearse a cold Monday: an alarm, an absent colleague, a missing record. Each participant names what they must receive, what evidence they will inspect, and who can decide that it is adequate. The result is a short exception-and-acceptance record with named owners, rather than a general improvement workshop.
The proposed result does not certify engineering, replace the client’s acceptance authority or promise that all defects have been found. Its boundary matters commercially as well as operationally. Someone could buy the rehearsal and its record without buying an open-ended promise that Mara will solve every handover. Actual acceptance remains with the authorised people.
C, a consortium service, becomes plausible only after the second use. Its reusable part is the preparation, rehearsal structure and evidence pattern. Its local part is the museum’s particular objects, equipment and people. The hidden combination is operational rehearsal plus a portable service boundary plus a buyer spanning several schemes. None of those ingredients is new. Their fit in this invented setting is the proposition.
D, a partner-led version, may be better for Mara than ownership. An existing commissioning practice could supply facilitators and contracting capacity while Mara contributes the difficult boundary questions. That exchanges some control and potential margin for a less fragile working week. The map’s higher reward is an assumption to challenge: if the partner expects Mara to do all the exception handling, the apparent freedom disappears.
Three tempting wrong turns
A · Optimise the rescue
Templates and automation could make Mara’s present work faster. This is a valuable baseline and may be the right answer. It becomes a trap only if released time fills with more rescues while the acceptance gap remains.
C · Sell before repeatability
One successful workshop may depend on Mara’s standing. Before calling it a service, a second facilitator must deliver a useful result under a bounded brief. An attractive first result can conceal an unscalable personal role.
E · Build the platform
Software is a larger leap: records, support, procurement and sustained adoption all need owners. A polished portal can move the bottleneck from handover translation to data maintenance. Here its setup burden is much greater than its evidence.
The route matters as much as the destination
The path is not “notice gap → launch company”. It is a sequence of changed beliefs and agreements. The first useful move is a protected rehearsal inside the existing delivery remit, subject to that remit actually allowing it. The second is a genuine choice by another sponsor. The third is delivery by someone other than Mara. Each step can fail without making the preceding learning worthless.
Make one boundary visible
In one handover, record the disputed acceptance questions before and after the rehearsal. Keep preparation time and displaced work visible.
Stop: nobody with acceptance authority will participate, or the rehearsal duplicates an effective existing review.
Ask for the second use
A second museum sponsor chooses whether the same bounded result is worth protected staff time or an agreed fee. Praise alone is not demand.
Revise: if only Mara wants it, improve the present workflow rather than assuming a market.
Remove the heroic dependency
A second facilitator runs the same kind of rehearsal. Compare usefulness, unresolved exceptions and total effort, including Mara’s hidden support.
Stop expansion: if the result still depends on unlimited access to Mara, retain it as a personal method.
Who has to find the new arrangement worthwhile?
| Actor | Reason to participate | Reason to refuse | What the probe must reveal |
|---|---|---|---|
| Museum sponsor | A clearer opening decision | Another meeting with no authority | Does the rehearsal change a real acceptance decision? |
| Contractor | Fewer ambiguous closeout demands | A rehearsal becoming unpriced scope | Can exceptions remain explicit without expanding the package silently? |
| Custodian and conservator | Usable evidence and rehearsed exceptions | Responsibility without time or training | Can they actually perform the receiving role? |
| Consortium | Learning travels between schemes | Budget is allocated by individual project | Who sponsors work whose benefits cross project boundaries? |
| Mara or delivery partner | A repeatable bounded result | Invisible support swallows the margin | Can someone else deliver it within the agreed boundary? |
This is why the map includes reliance on others. A destination that looks close in capability space can be far away in permission space. Conversely, a modest agreement spanning two projects may matter more than a substantial improvement in the tool.
What this first essai finds—and leaves open
The strongest candidate is a small acceptance rehearsal with a deliberate second-use test. The interesting possibility lies in converting repeated translation into a shared, bounded practice. The strongest objection is that the demand may be for Mara’s judgement and authority, not for a reusable result. That objection is exactly what the first probes must separate.
The novelty audit is deliberately modest. The invented brief already mentions handover, templates and coordination. “Faster switchboard” is a known improvement. The rehearsal service and partner route are underexploited combinations. The platform is an additional candidate not mentioned in the brief, but it is familiar in the wider world and weakly justified here. The strict unmentioned lane does not automatically produce the best opportunity.
A later model might discover a better actor boundary, represent sequences instead of points, challenge the composite notion of reward, or show that “equilibrium” is the wrong frame. A real enquiry would need evidence from receiving teams, current provision and actual sponsors. More expressive maps should make that uncertainty easier to inspect, not easier to forget.