Pilot design for a first coaching buy.
Population, success criteria, kill-criteria, and the claims you refuse. Not an RFP checklist. Not launch-week ops.
Pilot design for a first coaching buy is the population, success criteria, kill-criteria, and claims you refuse before you sign. It is not an RFP checklist. It is not the two weeks after signature. If you skip this page, you will evaluate a vendor on vibes and then call the mess a product problem.
The internal sell for this cycle is getting executives to approve a leadership development budget. Make-or-buy of the labor is build vs buy coaching capacity. This page is the first contract's design, once you already know you are buying a wave.
A pilot without kill-criteria is a rollout you are afraid to name.
Before the contract, not the RFP.
An RFP asks vendors to perform a catalog. A pilot design asks you to perform a decision: who, what good looks like, what would stop you, what you will not say in the readout. Security questionnaires still matter. They are not this page. Architecture, once you are buying a platform, is choose a platform by architecture.
Launch ops start after you sign. Launch a coaching program people actually book. Do not write SSO into the success criteria. That is hygiene.
Pick a population, not a sample of volunteers.
Volunteers teach you about volunteers. A named job (first-time directors, a function, a site) teaches you whether coaching works for that job. Access rules for a nominated wave are nomination vs open enrollment. For a first buy, nomination is usually cleaner. Open enrollment as a first pilot is how you learn about marketing, not about a job.
Do not fill the pilot with PIPs only. You will prove that coaching looks like punishment. Mix stretch and support, or pick one job that is not already a file.
Success criteria you can observe.
Holds exist past session one
The population can get a second and third conversation. If they cannot, you do not have a pilot. You have a demo.
Three named behaviors
Written before kickoff. Observed by people who work with the manager. Not a smile sheet.
The owner hours were real
Someone ran the weekly look. If the pilot only worked because a founder heroed it, you learned nothing about the company.
Day-30 and day-90 reads belong on the 90-day adoption playbook. Write the criteria now so those meetings are not a scavenger hunt. SCALE can run the 1:1 room. See SCALE. The criteria are still yours.
Kill-criteria that are allowed to kill.
| Kill if | Because |
|---|---|
| Second holds never appear after you fixed match and calendar | The design does not fit the week |
| The sponsor ghosts the check-ins | You will not survive a rollout |
| Confidentiality is treated as optional | Trust will not recover in a larger wave |
| No one has hours to run the week | You bought a catalog |
Killing a pilot is a success if the criteria were honest. Stretching a dead pilot into a three-year contract is how coaching gets a bad name internally.
What you will not claim.
Write it in the pilot charter. We will not claim company-wide ROI. We will not claim a control group unless we designed one ethically (see control-group ethics). We will not publish a utilization trophy. We will claim holds and named behaviors for this population, or we will claim that we stopped.
That charter is what you take back to the ELT. It is also what keeps a vendor from writing your readout for you.
Write the pilot charter before the paper.
Bring the population you have in mind. We will write success, kill, and the claims you refuse.
Book a strategy callFrequently asked questions
How do you design a pilot for a first coaching buy?
Name a population small enough to operate, write three observable success criteria, write the kill-criteria that would stop a rollout, and list the claims you will not make at day 90. Then sign. Do not start with a 40-page RFP.
Who should be in a first coaching pilot?
A wave that shares a job (new directors, a function that just inherited managers, a site with a known gap). Not a random sample of "interested" people, and not only the people already on a performance plan.
What are good kill-criteria for a coaching pilot?
Examples: first-to-second holds never appear after you fixed calendar and match; the sponsor disappears; confidentiality is breached; you cannot staff the weekly owner hours. Kill-criteria are about operability and trust, not a mood survey.
What should you not claim after a coaching pilot?
A company-wide ROI, a culture transformation, and a utilization percentage you will later pretend is a benchmark. You may claim whether this population held sessions and whether named behaviors showed up.
Is this the same as launching the program?
No. Pilot design is what you decide before the contract. Launch is the two weeks after signature. Adoption is the 90 days after people book. Do not collapse the three.