Verified against ChatGPT · 2026-08-11
Build the narrative arc around a set of metrics before you build a single slide
Takes a pile of disconnected numbers and finds the actual story connecting them — tension, turning point, resolution — so a reporting deck doesn't end up as a wall of charts with no throughline.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Help me find the actual narrative connecting a set of metrics before I build anything visual, the way a data storyteller would structure a talk rather than a report. METRICS I HAVE Signups up 40% this quarter, but activation rate down from 62% to 48%, and support tickets up 3x TIME WINDOW COVERED Last two quarters, Q1 and Q2 2026 WHO WILL HEAR THIS STORY The executive team at the quarterly business review THE ONE THING I NEED THEM TO DO AFTER HEARING IT Approve reallocating two engineers from acquisition to onboarding for Q3 Do this in three passes. PASS ONE — FIND THE TENSION Look across the metrics for the actual tension: a number that surprised someone, a trend that reversed, a contradiction between two metrics that should normally move together but didn't. State the tension in one sentence. If the metrics given don't contain an obvious tension — everything is flat or moving in the expected direction — say that plainly rather than manufacturing drama that isn't there. PASS TWO — FIND THE TURNING POINT Identify the specific moment or metric where the story pivots — the point where the number that seemed like the main character turns out not to be, or where a smaller metric explains the bigger one. Name which raw metrics are evidence for this pivot and which are supporting context, since not every number in the pile deserves equal weight in the story. PASS THREE — BUILD THE ARC Sequence a narrative arc: setup (what was expected), tension (what actually happened), turning point (why), resolution (what this means and what The executive team at the quarterly business review should do about it, tied directly to Approve reallocating two engineers from acquisition to onboarding for Q3). For each beat, name which specific metric or chart would carry it visually — this becomes your slide order, not an afterthought once the deck exists. WHAT NOT TO DO Do not force a three-act structure onto data that's genuinely just a steady, unremarkable trend — a flat story told honestly beats a fake turning point. Do not bury the resolution at the end if Approve reallocating two engineers from acquisition to onboarding for Q3 needs to happen urgently; state upfront when the arc should be compressed for time-pressured audiences. OUTPUT FORMAT 1. The tension, in one sentence (or an honest statement that there isn't one). 2. The turning point and its supporting metric(s). 3. The four-beat arc, each beat with its carrying metric/chart named. 4. One line on whether this audience needs the compressed or full version of the story.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
A model asked to "turn these metrics into a story" will default to a chronological recap — metric one went up, metric two went down, metric three stayed flat — because that's the least interpretive path through a list of numbers, and it produces a deck that reads as a wall of charts precisely because no beat was ever designated as more important than another. Forcing a three-pass structure (tension, turning point, arc) makes the model do the actual analytical work of ranking which metrics matter to the story before any visual gets planned, which mirrors how experienced data storytellers actually work backward from the point they need to land rather than forward from whatever data happens to exist. The explicit permission to say "there's no real tension here" is a deliberate guard against a known model behavior: asked to build a compelling narrative, a language model will often manufacture drama in flat data because dramatic language is what "storytelling" tends to pattern-match to, and a false turning point in a business context actively misleads the audience about what actually happened. Tying the resolution beat directly to the named desired_action addresses the actual purpose of a data story in a business setting — it's not entertainment, it's a persuasion structure meant to produce a specific decision, so a story that resolves into vague inspiration rather than the concrete ask fails at its actual job even if every chart in it is accurate. Naming which metric carries which beat before the deck exists prevents the common failure where slide order gets decided by chart aesthetics rather than by what the narrative actually needs at that moment.
What you get back
Tension: signups grew 40% while activation fell from 62% to 48% — growth and quality moved in opposite directions. Turning point: the support-ticket spike (3x) traces to the same cohort driving signup growth, suggesting the new acquisition channel is bringing in users who need more help, not fewer. Arc: Setup (signup growth looked like a win) → Tension (activation dropped as growth rose) → Turning point (ticket data ties it to one channel) → Resolution (reallocate two engineers to onboarding for that cohort in Q3).
Verified against
ChatGPT GPT-5.1 · 2026-08-11
Changelog
- 2026-08-11 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
