Expectations
"Be more senior" is not feedback.
Write down the handful of observable things you actually want from each report. Everything else in Impact Plan hangs off this one list.
You know what good looks like. They cannot read your mind.
Almost every unproductive performance conversation has the same root: the manager had a standard in their head and never wrote it down, so the report was working towards a target nobody had described. Then at review time they find out they missed it.
Writing it down costs an hour once and removes an entire category of problem. It also makes you think — most managers find that the first draft of their own expectations is vaguer than they expected.
Not this
- "Shows ownership"
- "Is a strong communicator"
- "Thinks strategically"
- "Raises the bar"
Nobody can act on these, and two people will never agree on whether one has been met.
This
- "Raises a risk while there is still time to act on it, not in the retro."
- "Leaves review comments that explain the reasoning, not just the objection."
- "Brings options to an ambiguous problem rather than questions."
- "Writes the runbook before the on-call rotation needs it."
Observable. Either you have seen it happen or you have not.
Two kinds, and you need both
What the role asks for
The expectations that come with being a senior engineer, or a designer, or an account manager on your team. They apply to everyone in that role, which is what makes a level mean the same thing for two different people — and what makes a promotion argument survive contact with calibration.
What you are asking of this person
Coaching expectations you write for one report, because of where they are right now. The engineer who needs to start delegating. The person you are quietly preparing for a lead role. These are yours, they sit alongside the role's, and they are visible to the person they are about.
Most managers start entirely with the second kind and add the first once it has spread to a second team. Neither requires HR to have built anything first.
The blank page is the obstacle, so skip it
Generate six to eight observable behaviours for a role in a few seconds, then cut what does not apply and rewrite what does. Editing somebody else's list is a far easier job than writing your own, and the list you end up with is still yours.
- Each one comes with a description and the skills that enable it
- The more context you give about the team, the less generic the output
- Everything generated is a draft you own outright and can rewrite
Three states, because five would be a fiction
Every expectation sits in one of three places for each person. Not a score out of ten, and not a five-point scale where nobody can tell a three from a four.
Not demonstrated
You have not seen it yet. Often that means no opportunity has come up, which is itself worth noticing.
Developing
Some evidence, not yet a pattern. This is where most of the useful coaching conversation lives.
Consistently demonstrated
Reliable, repeated, and backed by evidence you can point at. Reaching this lands on the person's timeline as a milestone, and raises the skills that behaviour depends on.
Both of you assess — the difference between what somebody thinks they are demonstrating and what you have actually observed is one of the most useful things a check-in can surface, and it is very hard to get at by asking directly.
Stop writing reviews from memory.
Set up your first report in about ten minutes. Free forever for teams of five.
No credit card. Your reports never pay.