Medium: Using Claude as a strategic thinking partner rather than a productivity tool

What the article is about

Mohit Aggarwal, a product manager with eight years of experience, describes how he restructured his relationship with Claude — moving it from occasional writing assistance to a persistent strategic thinking partner. The article is part personal case study, part argument: most PMs who use AI are deploying it on low-value documentation tasks while continuing to spend their own time on the administrative work that surrounds those tasks.

Context

Aggarwal identifies a familiar problem in modern PM roles: the job has expanded to include strategy, analytics, writing, project coordination, and stakeholder management, often simultaneously. The result is chronic context-switching that fragments the time available for actual product thinking. He describes a specific scenario — managing PRDs, reconciling stakeholder feedback, and navigating sprint priorities at the same time — as the kind of situation where most PMs feel perpetually behind despite being fully occupied.

Key takeaway and method

The article’s central argument is that the shift from “productivity tool” to “thinking partner” requires a deliberate change in how you structure inputs to the model, not just a change in what you ask it to do. Aggarwal describes giving Claude ongoing context about his product, stakeholders, and priorities — effectively treating it as a team member with a working model of the situation — rather than starting each session from scratch with a minimal prompt. This persistent context changes the quality of outputs on strategy documents and decision frameworks, since the model can test reasoning against existing constraints rather than respond to isolated requests.

The practical outcome is that documentation, stakeholder summaries, and internal communication can be delegated without sacrificing quality, freeing time for the work that requires judgment the model cannot substitute for.

Who it is useful for

Product managers who already use AI tools but find the results inconsistent or feel they are spending more time editing AI outputs than thinking through problems. Also relevant for senior PMs managing multiple product areas simultaneously, where the administrative overhead of coordination is highest.