What Good Looks Like:How to Write It Down for a Hire
What good looks like, written down: skip the job description, say what they produce and how they work, and make it specific enough to point at in a chat.
To write down what good looks like, skip the job description. Write a short statement of what this person is expected to produce and how they are expected to work, specific enough that either of you can point at it. Then check they agree it is fair.
How to write down what good looks like in one page
Right now, what good looks like lives in your head. All of it. The level of detail you expect, how fast replies should go, how a customer should be handled. You know it so well you've stopped noticing it's there.
They can't see it. Nobody can infer it from outside. New people won't ask, either, because asking feels like admitting they haven't guessed. So they work to a standard they've invented, and you judge them against one they've never seen.
Then you give feedback, and it's your impressions. Legitimate. Unarguable. And useless, because they have no way to know whether they're missing an expectation or a mood.
Writing it down fixes that. Not because paper is magic. Because a shared reference changes what the conversation is. It's no longer you against them. It's the two of you looking at the same page.
A job description is the wrong document
Most people reach for a job description. Resist.
A job description lists duties. Answer enquiries. Manage the inbox. Produce weekly updates. You can do every one of those things badly and tick every box. It tells you what the role contains, and nothing about what good looks like.
| Job description | What good looks like | |
|---|---|---|
| Describes | Duties and tasks | What they produce and how they work |
| Answers | "What is this role?" | "Am I meeting the standard?" |
| Useful in feedback? | Barely. Everything on it can be done badly. | Yes. You can both point at a line. |
| Written for | Hiring and paperwork | The weekly conversation |
| Changes | Rarely | Whenever the work or your expectations do |
Pick the right column. You aren't filing paperwork. You're giving someone a map of what you want.
Free guide · 21 pagesAmbiguity is the cruelty.The Performance Conversation Playbook: how to tell someone the truth about their work.Get the free guide →What goes on the page
Two halves. What they produce. How they work.
What they produce. The outputs. Not "handles customer messages" but what a good outcome of that looks like: a reply that answers the question asked, sent within a time you name. The thing you'd be happy to receive.
How they work. The standards that sit underneath the outputs. How much detail. How they handle a customer who's upset. What they do when they're stuck, whether they ask or struggle quietly. How quickly they tell you bad news.
Keep it short. A page, give or take. If it runs to many pages, nobody will read it, and a document nobody reads is a document you've just made for yourself.
Make every line specific enough that you could point at it. "Communicates well" fails the test. "Tells me within a day if a deadline is at risk" passes. A line that can't be pointed at will turn straight back into an impression the first time you use it.
Start from the document you already have
If you brought this person in with a ninety-day outcome document, you've already written half of this. Same artefact. You're extending it past the first quarter.
That is the easiest way in. Take what you agreed for the first ninety days, and ask what it should say for the next stretch. What have they proven? What's now expected as standard? What's the next step up?
If you didn't write one, don't wait for the perfect version. Write down five things you'd say if you were describing a great month. Then turn each into a sentence you could point at. That's a start, and you can refine it as you learn.
They have to see it, and agree it is fair
Here's the step people skip. The document isn't finished when you've written it. It's finished when they've read it and said "yes, that's a fair description."
If they don't agree, you've learned something valuable. Maybe it contradicts what you told them at the start. Maybe it assumes things about their hours or tools that aren't true. Maybe you've set a bar that no one told them about. Better to discover that now than in the middle of a hard conversation.
Agreement doesn't mean they get to rewrite it. It means you hear the pushback, change what's fair to change, and explain what you're keeping and why. That conversation is the first feedback conversation, and it's a low-stakes one. Good practice for both of you.
How it changes every conversation after
With the page in place, feedback gets simpler. You can say, "Here's the line about replying within a day. On Tuesday's message that didn't happen. What got in the way?" You're not arguing about whether you're being unreasonable. You're pointing at something you both agreed to.
It works the other way too. They can point at it. "This job wasn't in the description" is a legitimate challenge, and it should be. If a standard isn't written down, it isn't fair to hold them to it in a hard moment. That is the point.
It helps with giving specific feedback, because the reference gives you the language. And it makes the regular one to one slot useful, because there's always something to check against.
Three ways it goes wrong
It's too vague. "Takes ownership." "Shows initiative." Lovely words. You can't point at them, so you're back to impressions. Replace each one with a behaviour you could see.
It's a wish list. Everything you'd love in a perfect person, written as if it were the minimum. Then everyone misses it, all the time, and the document becomes a stick. Describe what good looks like, not a fantasy.
It's written once and forgotten. The work changes. Your expectations shift. A page that no longer matches reality is worse than none, because it makes the next conversation feel arbitrary. Look at it when you do your lighter periodic review.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →A quick test before you hand it over
Read each line and ask three questions.
- Could I point at an example of this, good or bad, from the last few weeks?
- Would they say it is a fair description of the work as it really is?
- Would I hold anyone else to the same line?
If a line fails the first, make it concrete. If it fails the second, talk to them. If it fails the third, you've found a bar that only exists in your mood, and you should drop it or fix it before it causes a hard conversation.
Why it matters more as you add people
There's another reason, and it only shows up with the second person. Your bar drifts. A different mood, a different month, a different relationship, and you hold identical work to different standards without noticing.
The page is one defence. The other is a question you ask before any assessment: would I say the same thing about the same work from the other person? It takes ten seconds. At this size, it is the entire calibration process you need.
Written standards also turn what's in your head into something that works when you're not there. Every specific piece of feedback moves a standard from your head into theirs. Enough of them and "how we do things here" holds without you in the room. That is how one person's judgement scales.
Where it lives
Somewhere you both can open in ten seconds. A shared document. A page in the folder you already use. Not a file buried in a drawer.
Then use it. Open it in the periodic review. Mention a line in a regular chat when it's relevant. A standard that appears only when something has gone wrong teaches people to fear the page. A standard that appears when things go well teaches them it's a tool.
And be ready to be wrong. Sometimes the conversation shows that your line was unreasonable, or that you never told them something it assumes. That's not a failure of the document. That's the document doing its job: showing you a gap while it's still cheap to close.
Free guide · 21 pagesAmbiguity is the cruelty.The Performance Conversation Playbook: how to tell someone the truth about their work.Get the free guide →What good looks like is not your goals
One caution. What good looks like is not your goals document. Goals are what should be achieved, deliberately beyond comfortable reach. This page describes the ordinary standard of the work.
Mix them and you'll get the same two problems as always: cautious targets, and bad news hidden until late. Keep the two documents separate and say so. The other three conversations they sit next to are covered in feedback versus development versus pay.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →Frequently asked questions
How long should it be?
Short. A page is a sensible target. If it is long enough that nobody would read it, it will not work as a reference in a conversation.
Is it the same as a job description?
No. A job description lists duties. This describes what they produce and how they work, specifically enough for either of you to point at it.
What if they think it is unfair?
Good, because you hear it now rather than during a hard conversation. Listen, change what is fair to change, and explain what you are keeping and why.
When should I update it?
Whenever the work or your expectations change, and at least when you look back over a longer stretch. It extends the outcome document you started with, past the first quarter.
The full method, with the checklists, is in The Performance Conversation Playbook, a free 21-page playbook.
How trapped are you, really?
Take the Corporate Suffocation Index, the free solopreneur readiness assessment.