We've been working with Pragmatic Institute at SAP — the training organization that's been teaching product managers and product marketers how to do their jobs with rigor and discipline for over three decades — and one of their Certified Instructors, Will Scott, has been pushing my thinking in a direction I wasn't expecting.

Will and I have been talking about how AI changes the PMM role, and last week he showed me something that reframed the way I think about prompt design entirely. Not a tool. Not a plugin. A prompt.

But not the kind of prompt you're imagining right now, which is probably something like "write me a positioning statement for our new analytics feature."

That's a question. What Will built was something fundamentally different — a workflow disguised as a prompt — and our conversation about the thinking behind it is worth sharing.

The Methodology as Prompt

Will had been working on a methodology for producing product explainer videos — the kind of two-minute pieces that live on a landing page or get embedded in a sales deck — and he was frustrated by the usual production cycle where someone writes a creative brief on a Google Doc, hands it to an agency, gets back something that looks cinematic but says nothing specific about the customer's actual problem.

So instead of prompting the AI to write the video script, he built a prompt that turns the AI into the strategist running the discovery session.

The model walks a human through a structured conversation, one question at a time, collecting every input needed to produce the brief and the script at the end. The AI isn't the author. It's the methodology.

"The prompt is the process."

— Will Scott, Pragmatic Institute

When Will walked me through the architecture, I immediately saw the Pragmatic DNA in it — that insistence on understanding the market problem before you build anything, on talking to real humans before you make assumptions, on following a framework instead of freelancing.

The Six-Stage Story Arc

The entire intake is organized around a six-stage story arc that maps the customer journey:

The Customer Journey Framework

1
Context and Trigger The moment a change, challenge, or opportunity appears and someone realizes something needs to be done.
2
Current State Assessment Where they study the situation and uncover what's missing, broken, or incomplete.
3
Exploration and Discovery The stage where they search for information, gather external insights, and start to understand the problem in context.
4
Decision Making The turning point where they evaluate what they've learned and choose a path.
5
Execution The action phase where the plan becomes real.
6
Outcome Where the results land and the impact becomes visible.

Every question in Will's intake maps to one of those stages. By the time you've answered all of them, you haven't just filled out a form — you've told the complete story of your customer, from the moment they felt the pain to the moment they saw the result.

And the model, having listened to all of it, can now produce a brief and a script that actually reflects the customer's world instead of regurgitating feature copy from a product page.

The Structure Holds You Accountable

I told Will this is exactly the pattern I've been trying to articulate — the idea that AI doesn't make you faster at producing deliverables, it makes you faster at doing the thinking that makes deliverables worth producing.

He pushed back, in the way that good instructors do.

"It's not just about thinking faster. It's about thinking completely. The prompt forces you through every stage of the story. You can't skip the current state assessment because you're excited about the features. You can't jump to the outcome because someone on the leadership team wants a demo video by Friday. The structure holds you accountable to the customer's journey, not your deadline."

— Will Scott

That landed hard, because he's describing a problem I see constantly — in my own team, in every organization I've worked with, and honestly in my own work when I'm moving fast and cutting corners.

We treat the model like a vending machine. Insert quarter, receive candy bar.

And the output is... fine. It's grammatically correct, it hits the right buzzwords, it has the shape of the thing you asked for. But it doesn't have the substance, because you never gave the model the substance. You skipped the thinking and went straight to the typing.

Beyond Video: The Pattern Scales

Will and I talked about how this principle extends far beyond video. Take competitive intelligence — most of us have asked an LLM to "write a battlecard for [competitor]" at some point, and most of us have gotten back a generic two-pager that sounds like it was scraped from the competitor's About page.

Now imagine a prompt that walks you through the competitive analysis first — asking you to specify the deal scenario, the buyer persona, the specific objection you keep hearing, the feature comparison that matters in this segment, the proof point that would swing the deal.

By the time the model writes the battlecard, it's writing your battlecard for your deal.

"That's the difference between a prompt and a process. A prompt gives you an output. A process gives you an output that's grounded in reality."

— Will Scott

Separating Intake from Output

There's a design principle in Will's approach that's worth calling out specifically: the separation of intake from output.

Phase One

Discovery

The model is asking, listening, confirming, and building understanding. Pure intake — no deliverable yet.

Phase Two

Synthesis

The model takes everything it learned and produces the deliverable. Output built on a foundation.

This is not how most people use AI. Most people combine intake and output into a single shot: "Here's some context, now write the thing."

And the result is predictably mediocre, because you're asking the model to simultaneously figure out what it needs to know and produce a polished artifact.

Will's architecture says: first, let's make sure we understand the problem completely. Then, and only then, let's build the solution.

It's the same principle behind every good consulting engagement, every good product spec, every good brief — except now it's automated and repeatable.

The Story Is the Real Asset

The other thing I took away from our conversation is that the story arc Will uses isn't just a video framework. It's a customer empathy framework that happens to produce great videos as a side effect.

When you force yourself to answer "what situation or change creates the need for action" and "what does the current state look like before using the product" and "where does the buyer go to find the information they need," you're not doing video pre-production.

You're doing the foundational customer understanding work that should underpin every single thing you produce — the positioning, the messaging, the competitive plays, the sales enablement, the analyst briefings, all of it.

The video is just one output format. The story is the real asset.

"If you get the story right, you can generate ten different deliverables from the same intake. The thinking is the leverage point, not the format."

— Will Scott

That's the key insight in a sentence. The thinking is the leverage point, not the format.

🎯 Your Challenge This Week

Take your most common deliverable — whatever it is, the thing you produce more than anything else — and try this:

  1. Build a prompt that interviews you about it first. Structure it around a story or a methodology you believe in.
  2. Make it ask one question at a time. Make it confirm each answer before moving on.
  3. Make it do the synthesis only after the discovery is complete.
  4. Then run it.

I think you'll find that the output is dramatically better — not because the AI got smarter, but because you did the thinking you were always supposed to do, and the AI made sure you didn't skip any steps.

The prompt isn't the question you ask the machine.

The prompt is the process you refuse to shortcut.

Will Scott, Pragmatic Institute
Will Scott Certified Instructor at Pragmatic Institute, the premier training organization for product management and product marketing professionals. Pragmatic works with organizations to build market-driven teams that ship products people actually want. Learn more at pragmaticinstitute.com.