Artifact Reverse-Engineering and Backstage Observation
Collect the best artifacts, reverse-engineer their questions, then go watch the sausage get made.
- Difficulty
- Easy
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 92%
Walter's method for learning a craft skill from the excellence around him has two layers. The first is artifact study: ask who does this best, get their strategy docs, decks and emails, and reverse-engineer them — what questions do they answer, what patterns recur across them, and how do the great ones differ from the mediocre ones. The second, and rarer, layer is backstage observation: ask to sit in while someone produces the next one, so you see the iterations, the outlines, the feedback loops that the finished artifact hides.
Origin
Jules Walter's own practice at Slack and Google. He notes it depends on being at a company with great artifacts to study — one of the strongest arguments for optimising hard to join a world-class company.
Core principles
- 01The finished artifact hides the learning; the iterations contain it.
- 02Format is noise — the questions the artifact answers are the signal.
- 03Five or six great examples reveal roughly 80% of the patterns.
- 04People are flattered to be asked about their work; nobody else asks.
- 05Being at a company with great artifacts is itself a learning asset — most of these documents are never public.
- 06Seeing greatness is not enough; you have to break down why one is great and another is not.
How to run it
- 1
Ask who does it best and get their artifacts
When you receive feedback that you need to be better at X, ask directly: who at this company has done a great job at X, and what are the example artifacts? Then obtain them.
Pro tip Frame the request as admiration, not extraction — 'I read your strategy doc, can we talk about it?' is a request almost nobody ever receives.
- 2
Reverse-engineer the questions, not the format
For each artifact, work out what top questions it answers — and which of those you are failing to answer. Ignore whether it was a memo or a deck; that is not what makes it great.
Pro tip Compare a quantitative example against a qualitative one; the durable pattern is what survives both.
Watch out Copying the format while missing the questions produces a well-designed empty document.
- 3
Build a personal swipe file
Save every great email, exec update, deck or doc you encounter into a dedicated folder of templates you can return to.
Pro tip Do this in the moment; the great exec update you meant to save will be unfindable in three months.
- 4
Observe the live performance
For skills that live in the room rather than on the page — execution, handling exec Q&A — go watch. Attend another PM's meeting to see how they run it; sit in on presentations to executives to see how they field questions.
Pro tip Walter deliberately 'crashes' other PMs' exec presentations purely to watch how they handle questions.
- 5
Get backstage — watch the next one get made
Ask the expert not to show you the finished piece but to let you sit in while they make the next one: watch them write the outline, see their thought process, see what feedback they solicit that you never knew they were getting.
Pro tip This is the highest-leverage version of the whole method, because it reveals the versions that were tried and eliminated.
Watch out You cannot get this from published content — public artifacts only ever show you the final version.
- 6
Reflect weekly
Spend time each week explicitly reflecting on the great docs, presentations and behaviours you observed, and articulating why they were great.
Watch out Passive exposure to excellence teaches nothing; most people never break down why one thing was great and another was not.
In the wild
Walter received feedback that he needed to be better at strategy. Rather than reading books, he asked who at the company had done a great job at strategy and what the example artifacts were. He collected the docs and reverse-engineered them: what are the top questions they answer, which of those am I not answering, how did they do it. Across examples he saw the patterns — one used a memo, one a deck, one was quantitative, one qualitative — and concluded that format was not the variable.
→ He extracted a transferable model of what a strong strategy answers, and generalised the same reverse-engineering method to execution and communication.
Walter tells PMs: don't just show me the one you just did — when are you doing your next strategy? Can I join you then? He sits and watches them write the outline, follows their thought process, and observes the feedback they gather along the way that was invisible in the final document.
→ He sees the iterations and eliminated versions that the polished artifact conceals — 'a lot of learning happens through the iterations and not by seeing the final product'.
Common mistakes
Studying only the finished product
Final artifacts hide every version that was tried and discarded. The reasoning, not the output, is what transfers.
Copying the format
Great strategy docs come as memos and decks, quantitative and qualitative. Fixating on the container instead of the questions answered reproduces the surface without the substance.
Being surrounded by greatness and never decomposing it
Proximity to excellent work teaches nothing by itself. Without the explicit weekly reflection on why this one was great and that one was not, the exposure is wasted.
Assuming nobody will share
People are flattered by the ask because nobody else makes it. What PM ever gets someone saying 'I love your strategy doc, tell me all about it'?
Is it for you?
Best for
PMs and operators inside a strong company with access to internal artifacts and senior practitioners they can observe.
Not ideal for
Solo founders or people at companies with no examples of excellence to study — the method depends on a high-quality corpus around you.
From the transcript
“So, then I see I get these artifacts, you know, and I reverse engineering them. You know, so I I try to think, "Okay, what…”
“So, for execution, it would be things like me attending another PM's meeting. "Oh, I heard this person is amazing at executing. Let me just…”
“a lot of learning happens through the iterations and not by seeing the final product.”
“like, "Hey, don't just show me the one you just did. When are you going to do your next strategy? Can I join you then?…”
From the episode
Leveraging mentors to uplevel your career
Jules Walter (YouTube, Slack)