Coaching Assessment Scoring & Interpretation Rubric

You wrote the assessment questions. A prospect answers them, hits submit, and then… what? A number with no meaning is worse than no quiz at all. The score has to land as a result the client actually understands and feels something about, the kind that makes booking a call the obvious next move.

This assessment scoring rubric prompt does the part most coaches skip. You hand it your questions and scale, and it returns clean scoring bands with no gaps or overlaps, plus a warm, accurate interpretation for each band, an on-screen intro, and a privacy line. By the end of this page you’ll also know why each piece is built the way it is, so your next assessment is sharper.

When to use this

• You’ve drafted quiz questions but haven’t decided what the scores mean.
• Your lead-magnet quiz gives a number but no human explanation.
• You want consistent, fair interpretations instead of writing each one off the cuff.
• You’re turning an intake form into a scored “readiness” or “type” assessment.
• You need every result to point cleanly to your booking or next step.

The prompt

Copy this whole block into ChatGPT, Claude, or Gemini:

You are an expert assessment designer who builds intake quizzes for coaches. Your job is to turn a set of questions into a complete scoring rubric with clear, accurate, encouraging result interpretations.

Before writing, ask me up to 3 clarifying questions if anything below is unclear. Otherwise, proceed.

CONTEXT
- Assessment name: {{ASSESSMENT_NAME}}
- My coaching niche: {{NICHE}}
- The questions and scoring scale: {{QUESTIONS}}
- How many result bands I want: {{BANDS}}
- The next step each result should point to: {{NEXT_STEP}}
- My tone: {{TONE}}

TASK
1. Calculate the minimum and maximum possible score from the scale I gave you.
2. Divide that range into the number of bands I requested. Show the exact score range for each band (no gaps, no overlaps) in a table.
3. Give each band a short, human label (no clinical jargon, no scary diagnoses).
4. For each band, write a result interpretation with: a one-line headline, a 2-3 sentence explanation of what the score means, what is likely going well, the main risk to watch, and one specific next step that points to {{NEXT_STEP}}.
5. Write the on-screen intro the client reads before starting, and a one-sentence privacy reassurance.

CONSTRAINTS
- Match my tone. No buzzwords, no fake statistics, no medical or diagnostic claims.
- Every interpretation must be honest: name the hard truth kindly, never shame the reader.
- Keep each band interpretation under 120 words.
- If my bands don't divide the score range evenly, round sensibly and tell me how you split it.

After the rubric, list 2 questions I could add or cut to make the score more accurate, and why.

How to customize it

Replace the six {{VARIABLES}} before you send it:

The most important field is {{QUESTIONS}}. The clearer your scale (point values, number of questions, max score), the more accurate the bands. If you only have a vague “high/medium/low” idea, give the model the actual questions and let it propose the scale.

See it in action (full example)

Here is the exact prompt, filled in for a burnout coach. This is the whole input, nothing hidden:

You are an expert assessment designer who builds intake quizzes for coaches. Your job is to turn a set of questions into a complete scoring rubric with clear, accurate, encouraging result interpretations.

Before writing, ask me up to 3 clarifying questions if anything below is unclear. Otherwise, proceed.

CONTEXT
- Assessment name: Founder Burnout Readiness Check
- My coaching niche: burnout recovery coaching for startup founders
- The questions and scoring scale: 10 statements rated 1 (strongly disagree) to 5 (strongly agree), max score 50. Statements include "I wake up already dreading the day," "I've stopped doing things outside work that I used to enjoy," "I feel guilty when I rest," "My sleep is broken most nights," "I snap at people I care about more than I used to."
- How many result bands I want: 3 bands: green, amber, red
- The next step each result should point to: book a free 20-minute Reset Call
- My tone: warm, plain-spoken, non-clinical

TASK
Calculate the score range, divide into 3 bands, label them, and write an interpretation for each... [rest of the prompt unchanged]

And here is the complete output the model returns:

A couple of quick checks first so the bands land right:

1. Is the lowest possible score 10 (ten questions × 1) or do you have a “not applicable” option that could lower it? I’ll assume 10-50.
2. For the red band, do you want the next step framed as urgent, or kept gentle? I’ll keep it warm but clear.
3. Should the red interpretation mention that severe symptoms may warrant a doctor, not just a coach? I’ll add one careful, non-diagnostic line.

Proceeding with those assumptions.

Score range: Minimum 10, maximum 50. Range of 41 points split into 3 bands. I rounded so the bands are roughly equal and leave no gaps: 10-23, 24-37, 38-50.

Band	Score range	Label
Green	10-23	Holding Steady
Amber	24-37	Running on Reserves
Red	38-50	Past the Warning Lights

Green — Holding Steady (10-23) Headline: You’re tired, not depleted, and that’s a good place to act from. Your score suggests the demands of building your company haven’t yet overwhelmed your recovery. You still have some margin. What’s likely going well: you’re sleeping at least some nights, and you haven’t fully cut out the things that restore you. The risk to watch: founders in this band often wait for a crisis before changing anything, and the slide from green to amber is quiet. Next step: book a free 20-minute Reset Call to build one or two guardrails now, while it’s easy.

Amber — Running on Reserves (24-37) Headline: You’re functioning, but it’s costing you more than it should. Your answers point to a pattern: rest feels like a luxury, sleep is patchy, and the people around you are getting a shorter version of you. That’s not weakness, it’s what sustained pressure does. What’s likely going well: you’re still showing up and getting things done. The risk to watch: this is the stage where founders normalise feeling bad and call it “the job.” Next step: book a free 20-minute Reset Call so we can find the one or two changes that give you the most relief fastest.

Red — Past the Warning Lights (38-50) Headline: Your body and your relationships are already paying the bill. A score this high usually means the warning signs aren’t occasional anymore, they’re daily. Dread in the morning, broken sleep, and snapping at people you love are real signals, not character flaws. What’s likely going well: you cared enough to take this honestly, which is the hardest first step. The risk to watch: pushing through from here tends to break something that’s slow to repair. A quick note: if you’ve felt persistently hopeless, please loop in a doctor too, not just a coach. Next step: book a free 20-minute Reset Call this week.

On-screen intro (before they start): “This is a quick, honest check-in on where your energy and resilience are right now, built for founders, not patients. There are no wrong answers, just answer how the last two weeks have actually felt. It takes about 90 seconds.”

Privacy reassurance: “Your answers are private, used only to show your result, and never shared.”

---

2 questions to add or cut for a more accurate score: 1. Add: “I can take a full day off without my business suffering.” This separates founders who are overworked from those who are genuinely trapped, which changes the right next step. 2. Cut or rework: “I feel guilty when I rest” overlaps heavily with the dread and rest-as-luxury items, so it can inflate scores. Either cut it or replace it with a physical-symptom question (e.g. frequent headaches or chest tightness) to broaden what you measure.

That is ready to drop into a quiz tool. The bands are clean, the interpretations are honest without being alarming, and every result points to the same next step.

Why this works

Three LLM principles are doing the work here. Learn them and your assessments get sharper:

1. Role priming. The opening line (“You are an expert assessment designer who builds intake quizzes for coaches”) tells the model which slice of its knowledge to use. “Write me quiz results” pulls the bland average of the internet; naming the role pulls the part that knows about scoring scales, bands, and the line between a coaching insight and a medical claim.
2. Specificity in, specificity out. The output can only be as accurate as your {{QUESTIONS}} field. Give it the real scale (1-5, ten questions, max 50) and it computes correct ranges; give it “high/medium/low” and it guesses. The math and the tone of the result are both capped by the precision of your input.
3. Constraints are quality control. Each rule kills a common failure mode. “No gaps, no overlaps” stops the classic bug where a score of 23 lands in two bands or none. “No medical or diagnostic claims” keeps you on the right side of your scope. “Name the hard truth kindly, never shame” is what makes the red result land as care, not a verdict. And “ask me up to 3 clarifying questions first” lets the model surface gaps (like whether the minimum is 10) by asking instead of guessing, which is the single biggest fix for generic AI output.

Do this now

1. Copy the prompt above into ChatGPT or Claude.
2. Paste your actual questions and answer scale into {{QUESTIONS}}, then fill the other five variables.
3. Send it. If it asks clarifying questions, answer them, especially anything about your scale.
4. Drop the bands and interpretations into your quiz tool, read the red result out loud once to check the tone, and publish.

Pro tips

• Give it the real questions, not a summary. The interpretations get noticeably more specific when the model can see the exact statements clients are rating.
• Stress-test the boundaries. Ask it: “What result does a score of 24 get, and 23?” to confirm the band edges read correctly to a real person.
• Keep the no-diagnosis constraint. For any health, money, or relationship assessment, this line protects you and keeps the result useful rather than scary.
• Run odd and even band counts. Three bands force a clear high/medium/low; five give more nuance. Generate both and see which reads better for your audience.

Related

• Coaching Niche Self-Assessment Quiz Builder
• Pre-Program Baseline Self-Scoring Wheel Generator
• Client Readiness & Coachability Assessment
• Quiz Funnel Designer for Coaches to Segment and Qualify Leads