What if I only get to rung 2?

Then present rung 2 and say what you would need to reach rung 3. A presenter who names the limit of their own evidence is more trusted than one who leaps to a recommendation.

Is this not just adding spin?

The opposite. Spin skips rung 2's assumption check and the objection step. This pattern makes both explicit, which is what stops a finding from being over-claimed.

Slides & storytelling

The so-what ladder

Take a raw finding and push it up three rungs — what it means, why it matters, what to do — so your presentation ends in an action, not a number.

Use this when you have a result and someone is about to ask “and?”

The pattern

Copy this and replace everything in [SQUARE BRACKETS] with your own specifics.

Prompt template

Finding: [THE RAW RESULT — a number, an observation, a measurement.]
Context the audience already has: [WHAT THEY KNOW, and what they will not be surprised by.]
Audience: [WHO, and what they control.]

Climb the ladder, one rung per line:
1. What it says — restate the finding precisely, with its limits.
2. What it means — the interpretation, and what would have to be true for that interpretation to hold.
3. Why it matters to THIS audience — in terms of something they actually control.
4. What to do about it — one specific action, with who does it and when.

Then: name the strongest reason a smart person in this audience would reject rung 2, and how I should answer it.

If the finding cannot honestly support rung 4, say so and stop at the highest rung it supports.

Why each part is there

What they will not be surprised by: A finding is only interesting relative to what the audience already believes. Without this you present the obvious as though it were news.
What they actually control: Rung 3 is where most presentations fail. Mattering in the abstract is not mattering — it has to touch a decision this room can make.
What would have to be true: Exposes the assumption smuggled into the interpretation, before someone in the audience finds it for you.
The strongest objection: You will be asked. Rehearsing the strongest version rather than the easiest is the difference between an answer and a stumble.
Stop at the highest rung it supports: The honesty valve. Some findings genuinely do not license an action, and claiming one anyway is how a good result destroys your credibility.

Before and after

The lazy version

Help me present my finding that engagement dropped 18% in October.

What goes wrong: You get an elegant slide about an 18% drop. The room's first question is “so what should we do?” and you have nothing, because the prompt never asked for the rungs above the number.

The pattern, filled in

Worked example

Finding: weekly practice minutes dropped 18% in October compared with September, across all four classes.
Context the audience already has: they know October had two weeks of school exams. They will not be surprised that activity dipped.
Audience: the principal and the two teachers who set the exam calendar and the club timetable.

Climb the ladder, one rung per line:
1. What it says — restate precisely, with limits.
2. What it means — the interpretation, and what would have to be true for it to hold.
3. Why it matters to THIS audience — in terms of something they control.
4. What to do about it — one specific action, with who and when.

Then: name the strongest reason a smart person here would reject rung 2, and how I should answer it.

If the finding cannot honestly support rung 4, say so and stop at the highest rung it supports.

What that buys you: The exam-calendar context defuses the boring interpretation, so rung 2 becomes the interesting one: the drop persisted for three weeks after exams ended, which points at habit loss rather than time pressure. Rung 4 is now a timetable change someone in the room can actually make.

Now do it on your own work

Take a real finding of your own — a marks trend, an experiment result, something you measured. Climb all four rungs, then ask someone to attack rung 2. See whether your prepared answer holds.

Check yourself — every box should be true

Rung 1 states a limit on the finding, not just the finding.
Rung 2 names an assumption that could fail.
Rung 3 refers to something your specific audience can change.
You tested the objection on a real person, and can say whether your answer held.

Then explain it back

Which rung was hardest to climb honestly, and what does that reveal about how strong your finding actually is?

Write two or three sentences in your own words. If you cannot, you copied a prompt rather than learning a pattern — which is the whole difference this library is trying to make.

Does this depend on which AI tool I use?

Works in any assistant. One caution that is about tuning rather than capability: models are agreeable, so if you state your preferred conclusion anywhere in the prompt, rung 4 will tend to arrive at it. Give the finding and the audience, and leave your own conclusion out.

Why it actually works

Moving from evidence to a justified conclusion — and knowing when the evidence does not license it — is Bayesian reasoning in plain clothes.

See evidence update a conclusion →

Prompting well is a skill you can pick up in a week. Understanding what is happening underneath is the 18-month curriculum — and it is what stays useful when the tools change.

FAQ

What if I only get to rung 2?: Then present rung 2 and say what you would need to reach rung 3. A presenter who names the limit of their own evidence is more trusted than one who leaps to a recommendation.
Is this not just adding spin?: The opposite. Spin skips rung 2's assumption check and the objection step. This pattern makes both explicit, which is what stops a finding from being over-claimed.

Next patterns

The Prompt Patterns Library is a separate, optional track — it does not change the 78-week curriculum or your roadmap. See what the curriculum actually covers →