How to Write a Scorecard Action Plan People Act On

How to write a scorecard action plan: fill one cell for every dimension and level, give each cell a short write-up and about four actions, and write the weakest-dimension cells first. Each action names what people at that level get wrong, the first move, and roughly what it costs.

How to write a scorecard action plan (and why it is the product)

Here's the uncomfortable truth about your scorecard. The questions are not what you're selling.

The questions produce a coordinate. This person sits at level three on dimension four. Anyone can specify that arithmetic. What the person actually receives is a piece of writing about that coordinate, plus things to do about it. That is the product. Everything else is plumbing.

Most builders get the scorecard action plan backwards. Building the measurement feels rigorous. Writing the advice feels like copywriting. So they do the first with great care and the second never. One live scorecard in the guide's count carries a hundred and forty-five questions and not a single recommended action. It tells you exactly where you stand and nothing about what to do. Precision with no advice is a quiz.

You're a solo operator. You have limited hours. You will be tempted to treat the action plan as the bit you'll "add later". Don't. Later never comes, and meanwhile every person who completes it gets a number and a shrug.

The grid you owe every reader

An action plan is a grid. Dimensions run one way, levels the other. Every cell needs its own write-up and three to five actions. Real working instruments run at roughly four actions per cell, so budget four.

Do the sum before you commit to a structure. Five dimensions and five levels is twenty-five cells. At four actions each, that's a hundred actions. Seven dimensions and five levels is thirty-five cells. Eight dimensions and seven levels is fifty-six cells, and you will abandon it, probably after you've built every question.

So choose the grid you are willing to write. Then let that decide the shape of the scorecard, not the other way round. If twenty-five cells is a month of evenings and you don't have a month, cut a dimension now. It's cheaper than cutting it at cell nineteen.

If you're still deciding on the number of levels, how to set scorecard band boundaries covers where the cuts go. Here the point is only that every level you add is another row of writing you owe.

Free guide · 28 pagesBuild a scorecard that tells the truth.The Scorecard Playbook: an assessment that separates people and gives you a reason to talk.Get the free guide →

Write the weakest cells first

Not all cells are read equally. A good results page sorts dimensions weakest first and badges the top one "start here". Think about what that does. The person's worst dimension is the one they read. Their strongest one is the one almost nobody reaches.

That tells you where to spend your evenings. If you're going to write twenty-five cells at uneven quality, and you are, because it's evenings, know which five carry the product. They're the weakest-dimension cells. Write those first, write them to your highest standard, and let the strong-dimension cells be shorter and plainer.

It also tells you what a "good" cell sounds like at the bottom. Someone who scores badly on a dimension is not reading to be flattered. They're reading to find out what's wrong and what to do on Monday. Give them that.

What a good scorecard action plan actually says

Anyone can write "improve your onboarding". It costs nothing and it changes nothing. The cell earns its place when it says three things:

• What people at this level on this dimension usually get wrong.
• What the first move is, concretely.
• Roughly what it costs, in time or money.

That's the difference between a report someone acts on and a report someone skims. It is also the one part an AI cannot write for you. It needs the industry knowledge you've built over years of doing the work. The arithmetic is generic. The advice is where you are visible, or visibly absent.

Those are examples of the shape, not advice for your field. Your version will be better because yours comes from the mistakes you've watched people make.

Write the follow-up before you write a question

Here's the cleanest discipline I know. Write the follow-up email for every result before you write a single question. If two results would get the same email, you have one fewer result than you thought.

That sounds like a time sink. It's a time saver. It forces the real question: what would I say to someone at this level that I wouldn't say to the next level up? If you can't answer, the levels aren't levels. They're decoration.

The same logic applies inside the plan. A level that produces the same next step as the level above it is a cell you wrote for nothing. Merge them. You've just saved yourself a row.

Also decide what the result triggers. A result produces a tag, and the tag is a socket almost anything can plug into: an email sequence, a task, a notification. When you wonder whether the scorecard can do something for a given result, the answer is nearly always to tag it and let something else do the work. Design the plan and the follow-up together. How you name the levels matters here, because the name travels into your inbox, and how to name scorecard result bands shows why identities beat grades.

Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →

Say something uncomfortable

You will be tempted to soften every cell. Most people get a reassuring result, you feel generous, nobody gets upset. It destroys the whole thing.

A result that cannot be uncomfortable produces no urgency, no credibility and no reason to act. And the people you actually want spot flattery in seconds. They discount everything after it.

So make sure at least one result tells someone something true that they didn't want to hear. A useful test: would someone who scores badly still forward it to a colleague? If yes, you were honest and kind at the same time. If they'd bin it, you were just kind.

There's a related limit. A scorecard can't manufacture concern about a problem nobody has. If your people don't already worry about the subject, no amount of good writing makes them, and you'll collect completions with no conversations. Check that before you write four hundred words of advice about it.

Free guide · 28 pagesBuild a scorecard that tells the truth.The Scorecard Playbook: an assessment that separates people and gives you a reason to talk.Get the free guide →

Your Monday

You don't need a free quarter. You need four decisions, in this order.

1. Write down your cell count. Dimensions times levels. Put the number where you'll see it.
2. Write the follow-up for each result. If two match, merge the levels.
3. Write the weakest-dimension cells. Four actions each, each naming the usual mistake.
4. Read one cell aloud as a stranger. Delete anything you'd skim.

Then, once real people have read your scorecard action plan, go and look at which parts of the advice get read and whether the questions behind it do any work. How to find dead questions in a scorecard picks up from there.

Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →

Frequently asked questions

How many actions should each result have?

Three to five per cell, and about four is the working average. Fewer feels thin. More and nobody reads them. Make each one specific enough that the reader could start on it the same day.

Which cells do I write first?

The weakest-dimension cells. A good results page puts the worst-scoring dimension first, so that's the writing most people read. Write those to your best standard and keep the strong-dimension cells shorter.

Can AI write my scorecard action plan for me?

It can draft the wording, but it can't supply the knowledge. The valuable part is what people at a given level typically get wrong, and that comes from experience in your field. Draft with help if you like. Then rewrite from what you know.

What if the results are mostly good news?

Then the scorecard is probably flattering people, and that makes it worthless. Check that at least one result says something uncomfortable and true, and that a person who scores badly would still pass it on.

The full method, with the checklists, is in The Scorecard Playbook, a free 28-page playbook.