Skip to content

Bounded Contribution Pilot

Method — a bounded, reversible trial — instantiates Net-Additive Contribution Intake

Admits an uncertain contribution into a small, reversible sandbox with pre-set success, stop, and handoff criteria, so its real net value is observed before any full commitment.

Some offers can't be priced from the outside — the benefit is plausible, the coordination cost is guesswork, and the only honest answer is "we'd have to try it." Bounded Contribution Pilot is the mechanism for exactly that answer. It admits the contribution into a deliberately small, reversible slice of the real system — one team, one site, one quarter — with the success, stop, and handoff criteria written before the trial starts. Its defining property is reversibility: the pilot is scoped so that if it fails, the system can walk away having lost only the bounded stake, and so that whether it worked is observed rather than argued. It converts a "defer" disposition into evidence, and along the way it produces the one thing no paper estimate can — a track record of whether this contributor actually delivers.

Example

A city parks department is offered an ongoing trail-maintenance program by a newly formed volunteer group: monthly work parties, their own tools, "no cost to the city." The benefit is real if it materializes; the risk is that rangers end up supervising, insuring, and cleaning up after enthusiastic amateurs. Rather than decline or fully adopt, the department runs a bounded pilot — one park, one season — with criteria fixed up front: success means measured trail-condition scores improve without pulling ranger hours off priority work; the stop trigger is any safety incident or supervision load above a set threshold; the handoff question is who owns coordination if it scales.

By season's end the sandbox has answered what no proposal could. The trails did improve, but the group's turnout was erratic and the supervision cost ran higher than promised — a concrete reliability read on this particular partner. The disposition isn't a clean yes; it's reshape and re-pilot: continue in the one park with a named ranger liaison and a minimum-turnout commitment before any city-wide rollout is discussed.

How it works

  • Scope for reversibility first. The trial is sized so failure is survivable and cleanly undoable — a bounded stake, a single site, a fixed clock — before anything else is decided. Reversibility is the design constraint, not an afterthought.
  • Pre-commit the three gates. Success, stop, and handoff criteria are written before the pilot opens, so the trial reads out as a measurement rather than a negotiation once people are attached to it.
  • Instrument the contributor, not just the output. The pilot watches reliability — turnout, responsiveness, promises kept — because a contribution's real cost is inseparable from who has to be chased to realize it.
  • Exit on the criteria. At the clock or the trigger, the pilot graduates, re-pilots, or unwinds against the pre-set gates — not against how attached everyone has become.

Tuning parameters

  • Blast radius — how large and how reversible the sandbox is. Bigger pilots yield more representative evidence but cost more to unwind; keep it as small as still tests the real question.
  • Trial length — long enough to see the friction that only shows up over time (the third month, the first holiday), short enough not to become a de facto adoption by inertia.
  • Success-criteria strictness — how demanding the graduation bar is. Set loose, the pilot rubber-stamps; set unrealistically tight, it kills contributions that would have matured.
  • Stop-trigger sensitivity — how fast a safety, cost, or displacement signal aborts the trial. The tighter this dial, the more the pilot protects primary work at the cost of ending some trials early.

When it helps, and when it misleads

Its strength is that it buys evidence and optionality on offers too uncertain to price — you learn the real coordination cost and the contributor's real reliability while your exposure is still bounded and reversible. It is the honest home for "we don't know yet," and it defuses the pressure to either adopt on faith or refuse on suspicion.

Its classic failure is pilot purgatory: a trial that never ends, quietly becoming permanent adoption without anyone deciding it should — the reversibility spent by drift. Close behind is the unrepresentative pilot (too small, too favorable, too supervised to resemble real operation) and the pilot run to manufacture a yes, scoped so it cannot fail. The discipline that guards against all three is to fix success, stop, and handoff criteria in writing before opening the sandbox, and to force an explicit graduate / re-pilot / unwind decision at the clock rather than letting the trial decay into a default.

How it implements the components

Bounded Contribution Pilot realizes the trial slice of the archetype — the machinery for deciding under irreducible uncertainty:

  • bounded_trial_boundary — defines the reversible sandbox: its scope, its stake, its clock, and the pre-set success/stop/handoff gates that make it a measurement rather than a commitment.
  • contributor_reliability_signal — the trial doubles as a probe of the contributor, yielding a first real track record of whether they deliver on what they promise.

It does not price the offer in the first place or issue the disposition (Contribution Net-Value Review), secure the permanent owner and budget a graduating pilot then needs (Sponsor-Required Acceptance Protocol), or write the working terms it runs under (Contribution Onboarding Packet).

Notes

A pilot only tells the truth if it can fail cheaply. If unwinding the trial would already cost more than the system can bear — data entangled, a contributor publicly committed, a dependency taken — the pilot has lost the reversibility that was its whole point and has become an adoption wearing a trial's label. Guard the exit as carefully as the entry.