What Do We Tell the Partners and the Staff? Announcing a Clinic AI Project Without Starting a Panic
The same AI project reads as a threat or a plan depending entirely on how it is announced. Four facts make the difference, and they are the same four at the partners' meeting, at the staff huddle, and in anything a patient ever sees.
What this settles
What you say about this project, to your partners and to your staff, before anyone hears it secondhand.
In this article
You have done the work. You know which workflow, you know what it should cost, you know how you will prove it works. There is one step left that most practice owners treat as an afterthought, and it decides more about the outcome than the software does: what you actually say, and to whom.
Get this wrong and you do not have an AI project. You have a rumour about cutting front desk hours, with a budget line attached to it.
The gap gets filled either way
“We are beginning an AI transformation to drive efficiency across the practice” is a sentence with no facts in it. It names no workflow, no owner, no number, and no end. And because staff cannot act on it, they interpret it, and they interpret it in the direction of their own risk.
The interpretation is always the same. Efficiency means hours, and transformation means the job they are sitting in. The absence of a specific answer is not neutral, it is an invitation, and the most capable people at your front desk have the most options and start exercising them first.
”Intake takes the front desk the whole morning. We are testing a tool to draft it before the patient arrives. Priya owns it. If it is not working by the end of October, we stop.”
One workflow, one owner, one number, one exit. There is nothing left over to build a story out of.
”We are beginning an AI transformation to drive efficiency across the practice.”
No workflow, no owner, no number, no end. The room hears: whose hours are getting cut, and is this why the vacancy was never filled.
Nothing in the second version is softer. It is not reassurance, and it does not promise anyone’s job is safe forever. It is simply specific enough that there is no room left to invent something worse.
The same four facts, in every room
The instinct is to write three different messages: an optimistic one for the partners, a calming one for the staff, a quiet one for patients who ask. That instinct is wrong, and not only for the ethical reason. Those three versions meet each other. A partner forwards the summary, a patient asks the receptionist about it, and the differences between your versions become the story.
Write one set of facts. Change what you lead with, never what is true.
The partners lead with cost, payback, and exposure, because they are deciding whether to fund it out of shared overhead and what it risks. The staff lead with what changes about their Monday, because that is the only part of this that touches them. Patients lead with accountability and privacy, because that is what they are actually asking about. Same project, same facts, different opening line.
What every version has to contain
The announcement
Specific enough that nobody has to guess the rest
A project people can repeat accurately
The fourth one does the most work and gets left out the most. A stop date converts an open-ended commitment into a bounded experiment, and bounded experiments are far easier to approve, far easier to staff, and far easier to end. It is the same discipline as deciding what a proof of value has to prove, carried into the announcement instead of left in the plan.
Check the message before you send it
Scorecard
Is this announcement ready to send?
Check every one that is true of the message as currently drafted.
0 of 6 checked
The question nobody asks you directly
Someone at that huddle is wondering whether this is the beginning of their hours disappearing. They will not ask. You have to answer anyway.
If the honest answer is that nobody’s role is at risk, say it in those words, once, plainly. If the honest answer is that roles will change, say that instead, because the alternative is that people find out later and conclude you knew. That is the version that costs you the staff you were trying to keep, and it is the same failure behind staff quietly refusing to use the tool you bought: trust spent early is very hard to earn back.
The strongest position is usually available and underused. Most first clinic AI projects are aimed at work nobody enjoys doing, and the specific claim “this removes the part of your job you complain about every single day” is both more credible and more accurate than “this will make us more efficient.” Say the specific thing.
A project with a named owner and a stop date reads as governance. A transformation with neither reads as a gamble, and every partner in the room prices it that way.
What to do this week
Before the meeting, write four sentences: the workflow, the owner, the number, the stop date. If you cannot write all four, the problem is not your communication, it is that the project is not defined yet, and announcing it in that state is how it becomes permanent without ever being decided. That is also the honest test of whether you are ready to spend, which is where this series started.
About to put an AI project in front of your partners? Book a free consult and we will pressure-test the four facts with you: the workflow, the owner, the number, and the date you would stop.
The decision brief
One page for the person you are going to forward this to.
- The decision
- What you say about this project, to your partners and to your staff, before anyone hears it secondhand.
- What getting it wrong costs
- A vague announcement does not stay vague. Staff fill the gap with the worst available reading, and you spend the next quarter managing a rumour about cutting front desk hours instead of running a project. Your best receptionist starts looking first, because in this labour market they have the most options, and replacing them costs more than the tool.
- What you can do this week
- Write four sentences about the project you are considering: the one workflow it touches, the person who owns it, the number that decides whether it worked, and the date you would stop. If you cannot write all four, you are not ready to announce it, and probably not ready to fund it.
- What to tell your team
- We are testing one tool on one workflow, the intake that eats the front desk every morning. Priya owns it, we will know by the end of October whether it works, and if it does not we stop. Nobody's job depends on the answer.
Frequently asked questions
What should I tell my partners about an AI project?
Four things, in their units: which single workflow you are touching, what it costs and when it pays back, who owns it, and the date and condition on which you would stop. Partners are not unsettled by small budgets, they are unsettled by open-ended ones drawn from a shared overhead pool. A project with a named owner and a stop date reads as governance. A slide about transformation with no number attached reads as one partner's enthusiasm, and gets priced that way.
How do I tell clinic staff about AI without scaring them?
Be specific enough that there is nothing left to invent. Name the workflow, say what changes about their actual Monday morning, and say plainly whether anyone's hours depend on the outcome. If the honest answer is that roles will change, say that too, because people discover it eventually and finding out late is what destroys trust. Vagueness is not kindness. It is the thing that starts the rumour, and a small clinic has no room to hide one.
Should we announce a pilot before we know if it works?
Yes, and framing it explicitly as a test is the advantage rather than the weakness. A pilot announced as a pilot, with a stop date, gives you permission to end it without it reading as a failure. A pilot announced as a transformation cannot be stopped without a partner losing face, which is how a practice ends up renewing a licence nobody believes in for three years.
What do I say if the project fails?
Exactly what you said you would say when you announced it, which is why the stop condition belongs in the first announcement rather than the last one. If you named the number and the date up front, ending the project is you keeping your word. If you did not, ending it looks like a retreat, and the next proposal, from you or from anyone else in the practice, inherits that scepticism.
Before You Spend
All 10 decisionsThe decisions an owner or CEO makes before the first dollar of AI budget leaves the building. One decision per article, each ending in a page you can put in front of your leadership team.
Found this useful?
Share this with your network on LinkedIn, it helps more than you think.
Enjoyed this read? Get the next one in your inbox.
When there is something worth your time, you will be first to know. No spam, unsubscribe anytime.
Keep reading
Who Owns This Once It Is Live? The AI Job Nobody Budgets For
Ordinary software fails loudly and someone files a ticket. AI fails quietly and keeps answering. That difference is why a clinic AI system needs a named owner, a number, a cadence, and the authority to switch it off, and why most practices discover they have none of the four.
Read article
Your Patients' Data Stays Yours: Adding AI Without Handing Over the Chart
The top reason a practice stalls on AI is not cost, it is not knowing where the chart goes. Here is how to add intelligence while personal health information stays in your environment and in Canada, plus the seven questions to ask any vendor before you sign.
Read article
When Do We Stop? Writing the Kill Criteria Before You Write the Cheque
A clinic pilot with no stated ending does not end, it fades, and keeps billing while it fades. Deciding the number, the date, and who calls it costs nothing on day one and is nearly impossible to introduce in month nine.
Read articleHave software that should be smarter?
Let’s map a free AI-transformation roadmap for your product.