Asking Claude for brutal feedback on my article
Here’s the prompt I used against an upcoming article:
Review this article as an editor with a brutal lense - I don't want to hear about how it's good, I'd like you to challenge the premise, key points, and call out opportunities I missed to make a compelling argument for teams to lead with a diagnosis.
And, here’s what Claude Opus 4 delivered. I think it did pretty good.
Looking at this article with a critical eye, here are the key issues and missed opportunities:
Fundamental Challenges to Your Premise
The opening example is too vague to be compelling. You tell us the onboarding project failed because they "skipped the diagnosis," but you don't actually show us what diagnosis would have revealed. What were the actual underlying issues? Without specifics, readers might think "well, maybe they just executed poorly" rather than buying into your diagnosis thesis.
You're preaching to the choir without addressing the real barriers. Most readers probably already know they should understand problems before solving them. The question is: why don't they? You need to tackle the organizational forces that create this behavior - quarterly targets, promotion cycles that reward visible action, the career risk of "analysis paralysis," etc.
Critical Gaps in Your Argument
Where's the counter-argument? You never address when diagnosis might be wrong or harmful. What about fast-moving situations where the cost of diagnosis exceeds the cost of trying something and failing fast? Tech companies often succeed through rapid experimentation rather than careful diagnosis.
Your diagnosis examples are superficial. They read like problem statements, not diagnoses. A real diagnosis should reveal non-obvious insights. For instance, your PM example states they need more time for discovery - but WHY are they stuck in support? Is it a skills gap? A process failure? A cultural expectation? Push deeper.
You don't address diagnosis conflicts. What happens when different stakeholders have contradictory diagnoses? How do you reconcile when engineering thinks it's a process problem but product thinks it's a tooling problem?
Missed Opportunities to Strengthen Your Case
No proof of impact. Where are the before/after stories? "Team X spent 2 weeks on diagnosis and delivered in 6 weeks what Team Y couldn't deliver in 6 months." Give us the ROI argument that leaders need.
You don't teach the skill. Mentioning "Five Whys" in passing isn't enough. Show us what good diagnosis looks like in practice. Walk through a real example of moving from surface symptoms to root causes.
Missing the "how to sell it" angle. Your readers will face pressure to "just fix it now." How do they push back? What's the script for getting buy-in on diagnosis time?
No discussion of diagnosis pitfalls:
Analysis paralysis
Confirmation bias in diagnosis
When diagnosis becomes an excuse for inaction
How to know when you've diagnosed "enough"
Weak connection to your newsletter's theme. "Lost is the Way" is about navigating uncertainty, but you don't explicitly connect diagnosis to uncertainty navigation. How does diagnosis help when the landscape keeps shifting?
Structural Issues
The Rumelt framework feels shoehorned in - you mention all three parts but only discuss one, making readers wonder why you brought up the full framework at all
Your "bringing the discipline" section has good questions but no framework for when/how to deploy them
The collaborative diagnosis section is underdeveloped - you defer to Will Larson rather than providing your own insights
Suggested Related Posts to Link
Based on your newsletter's theme, you should reference posts about:
Decision-making under uncertainty
Building psychological safety (for honest diagnosis)
Managing up when leadership wants quick fixes
Creating space for reflection in action-oriented cultures
The article has a solid foundation but needs more depth, more edge, and more practical application to truly convince skeptics that diagnosis is worth the investment.