Pilot charter
Objective. Turn a qualified problem into a customer-owned baseline, a testable hypothesis, a bounded pilot scope and success criteria agreed in writing - with capacity value kept separate from cash.
A one-page charter that makes a pilot provable instead of a vibe. It fixes the baseline in the customer's own numbers, states one hypothesis you can pass or fail, bounds the scope so the result is clean, and writes the success criteria both sides sign up to. This is the tool version of S08 - copy it, draft it with the customer, do not start a pilot without it.
See it in the demo
What it is and when to use it
A pilot charter is the written agreement that defines what a pilot will prove and how both sides will judge it. Use it before any trial, POC or paid pilot begins, and revisit it if scope or sponsor changes. Without it, a pilot ends in a dispute about whether it 'worked', because nobody agreed the baseline or the test up front. The charter is customer-owned: the baseline is their number, the hypothesis is their outcome, and the success criteria are what THEY will accept as proof - not what you hope to show.
The charter - copy this
ACCOUNT: ____ SPONSOR (economic owner): ____ PILOT OWNER (customer side): ____
PROBLEM (from the discovery brief, in their words): ____
BASELINE (customer's own number, today, with how it is measured and the source): ____
- If no baseline exists, the first pilot deliverable is to establish one. State that here.
HYPOTHESIS (one sentence, pass/fail): 'If we [action] for [scope], then [measure] will move from [baseline] toward [target] within [window].'
SCOPE: application(s) ____ ; team / process / payer ____ ; number of seats ____ ; what is explicitly OUT of scope ____
DATA / INPUTS REQUIRED for the measure to be valid (sources, completion/standard definitions, access/role scope, capture policy + comms, freshness, time zone): ____
MEASUREMENT METHOD (what we compare, how, who runs it, same denominator before and after): ____
SUCCESS CRITERIA (what the customer will accept as proof): ____
VALUE TYPE: capacity value ____ | cash recovered ____ (kept separate, never blended)
TIMELINE + REVIEW DATE: ____ RISKS / DEPENDENCIES: ____
How to fill it
Draft it with the customer, not for them. Pull the problem and baseline straight from the discovery brief (KIT-01); if the baseline slot was empty there, the charter's first deliverable is to build one, and you say so plainly. Write exactly one hypothesis - a pilot that tests five things proves none. Bound the scope hard: one team or one payer, a stated seat count, and an explicit OUT list, so the result is attributable. Agree the measurement method before the pilot starts, including that the denominator is identical before and after. In the value-type line, label freed time as capacity and only money actually collected as cash - keep the four numbers (recoverable, billed, collected, capacity) distinct. Get the sponsor to confirm the success criteria in writing.
The honest caveat
A charter does not guarantee the hypothesis passes - it guarantees you will know honestly whether it did. If the measure does not move, that is a valid, reportable result, not a failure to hide. Do not pre-load the success criteria with a number the product has not been shown to deliver; targets are the customer's ambition tested against evidence, not a Time Champ promise. Capacity freed in a pilot is capacity, not cash - presenting it as cash is the single most common way a successful pilot becomes a disputed QBR later. Any required input that is not live (an integration, a standard definition, a capture policy) is a dependency that must be resolved or the measure is not valid - carry it as a risk, not an assumption.
Interpretation check
A pilot result is only meaningful against a baseline the customer owns and a denominator held constant before and after. Freed capacity is capacity value, not cash - the two must never be blended, even in a winning pilot.
In the field
KnackGlobal pilot charter: baseline = current cost per clean claim and output per FTE on one payer segment; hypothesis tests whether clean output per paid hour improves within the window; the freed claims-capacity line is kept explicitly separate from any denial-recovery cash so the result stays defensible against the incumbent displacement narrative.
Practical task
Draft a full pilot charter for a qualified opportunity: state the baseline and its source, write one pass/fail hypothesis, bound the scope with an explicit OUT list, define the measurement method with a constant denominator, and separate the capacity line from the cash line. Have the sponsor (or a role-play sponsor) confirm the success criteria in writing.
Self-check
What four things must a charter fix before a pilot starts?
A customer-owned baseline, one pass/fail hypothesis, bounded scope (with an OUT list), and success criteria the customer accepts as proof - plus the measurement method and value type.
There is no existing baseline. Can you still run the pilot?
Yes, but the first pilot deliverable becomes establishing the baseline, stated explicitly in the charter. You cannot measure improvement against a number that never existed.
The pilot frees 400 hours of capacity. How is that written in the charter result?
As capacity value, in the capacity line - never as cash. Only money actually collected goes in the cash line; blending them creates a disputed QBR later.
Why only one hypothesis?
A pilot that tests several things at once cannot attribute the result; one bounded hypothesis with a held-constant denominator is provable.
Verified vs demo. The charter structure is reusable; the baseline, targets and success criteria are the customer's own and must never be invented or pre-loaded with unproven numbers. Required inputs that are not yet live are dependencies to resolve, not assumptions to make.