Five frameworks for high-impact product management from Shreyas Doshi

Executive overview

Most product teams waste leverage by treating all tasks equally, misreading execution failures as execution problems, and optimising for ROI instead of opportunity cost. Shreyas Doshi, former PM leader at Yahoo, Google, Twitter, and Stripe, offers five interlocking frameworks that address where effort actually compounds.

The core insight: the biggest gains come not from working harder on more tasks, but from correctly categorising work, diagnosing root causes, and refusing to mistake busyness for leverage.

Pre-mortems: surfacing risks before they become failures

• Start the meeting with: "Imagine this project has miserably failed. What went wrong?"
• This framing grants psychological safety — people share concerns they'd suppress in optimistic planning meetings
• Each participant contributes three items: a tiger (real threat), a paper tiger (seeming threat you're not actually worried about), and an elephant (unspoken assumption nobody is addressing)
• Ask participants to vote on the tiger someone else raised that scares them most
• Output is a prioritised action plan owned by the pre-mortem leader
• The shared vocabulary persists: teams begin flagging tigers and elephants in future meetings without being prompted
• Split large launches into two sessions — one for engineering risks, one for go-to-market risks

The LNO framework: matching effort to leverage

• Every task is one of three types: leverage (L), neutral (N), or overhead (O)
• L tasks return 10–100x the effort invested; O tasks return less than was put in
• Ambitious people default to treating all tasks equally — this is the core error
• The same activity (bug report, meeting notes) can be L, N, or O depending on context
• Reserve perfectionism for L tasks; on O tasks, do the minimum required
• Two tactics to stop procrastinating on L tasks: (1) spend the preceding days on intentional N/O work ("placebo productivity") to create contrast; (2) change your physical location on the day you've scheduled the L task

The three levels of product work

• All product work operates at one of three levels: impact, execution, or optics
• Most team conflict comes from people defaulting to different levels without realising it — a CEO thinking about brand impact and a PM defending execution trade-offs are talking past each other
• Optics (internal communication, status updates) is legitimate and necessary — the problem is when it becomes the implicit goal rather than a means to enable impact
• The right default level depends on context: early-stage teams should optimise for execution; platform teams with visibility problems may need a temporary shift toward optics
• Use the vocabulary explicitly with your team so mismatches can be named and resolved without personal friction

Most execution problems are strategy or culture problems

• In high-performing organisations, persistent execution problems are rarely execution problems
• Strategy problem: teams are misaligned because no one has clarified priorities, and OKRs drafted in a vacuum naturally diverge
• Culture problem: teams are incentivised to protect their own OKRs rather than collaborate, so no amount of process or syncs fixes the underlying behaviour
• Interpersonal problem: two key people simply don't get along; the friction will re-emerge regardless of how the process is restructured
• True execution problems do exist: infrastructure debt, skills gaps, wrong person in a role — but these are diagnosable and bounded
• The diagnostic signal: if you've put on a bandaid and the same problem resurfaces in a different form, it isn't an execution problem

Opportunity cost thinking vs. ROI thinking

• ROI formula puts time in the denominator, which incentivises chasing quick wins and low-hanging fruit — filling the plate with positive-ROI tasks that collectively miss the upside
• Opportunity cost = value of optimal option minus value of chosen option; minimising it means asking "is this the best use of my time?" not just "is this a good use?"
• Don't try to quantify opportunity costs in a spreadsheet — the value is in the mindset shift during planning, not the arithmetic
• Practical implementation: set explicit allocation targets for the team (e.g. 60% incremental features, 30% big new initiatives, 10% stability) so space for high-opportunity work is protected by design, not competed away by obvious quick wins
• The 30% "big new initiatives" bucket should hold one or two bets, not five — scarcity forces genuine prioritisation

High agency as a PM differentiator

• High agency: finding a way forward without waiting for perfect conditions or attributing failure to circumstances
• Three components: ownership mindset, creative execution through constraints, and resilience
• It predicts PM impact better than credentials or apparent potential on paper
• PMs who underperform relative to their background almost always lack this trait; those who overperform almost always have it