Two-way doors, the local maximum trap, Goodhart's Law, the strategy kernel, trust debt — fifteen mental models that separate senior product managers from feature-shippers, with the senior move for each.
Junior PMs ship features. Senior PMs make decisions — about which problems are worth solving, which metrics to trust, which bets compound, and which doors you can walk back through. The difference usually isn't intelligence or effort. It's a vocabulary of mental models that turn fuzzy judgement calls into sharp, repeatable reasoning.
What follows are fifteen of the most useful, grouped into the four jobs they actually help with: making decisions, setting strategy, measuring what matters, and understanding users. For each, the concept in a sentence — and the senior move that turns the concept into behaviour.
The first sign of seniority is calibrating how much rigour a decision deserves. Most teams over-deliberate the cheap calls and under-think the expensive ones.
Decisions that can be reversed easily. Amazon's Jeff Bezos popularised the frame: when the cost of being wrong is low, move fast. Most decisions are two-way doors that teams mistakenly treat as one-way.
The general principle behind two-way doors: not every decision deserves the same process. Match the speed of your decision to the level of risk it carries.
Endlessly optimising today's solution can prevent you from discovering a far better one. A/B tests are hill-climbers — they find the nearest peak, not the highest. The next breakthrough may require abandoning today's success.
A backlog is not a strategy. These models force the difference between a plan and a pile of tasks.
From Richard Rumelt's Good Strategy / Bad Strategy: every real strategy has three parts — a diagnosis of the core challenge, a guiding policy for addressing it, and a set of coherent actions that carry it out.
Teresa Torres's framework connecting Outcome → Opportunities → Solutions → Experiments. It keeps teams focused on solving the right problem rather than shipping random features.
Small advantages that strengthen over time — network effects, proprietary data, brand, switching costs. Great products become harder to compete against every year, not easier.
Bain's term (Mankins & Garton, Time, Talent, Energy) for the internal friction that quietly slows execution. Many "product problems" are actually organizational problems in disguise.
Metrics are how products lie to you. These four models keep your numbers honest.
"When a measure becomes a target, it ceases to be a good measure." Optimise a number hard enough and people will hit it in ways that destroy the thing it was meant to represent. Metrics are signals, not goals.
Leading metrics predict future outcomes; lagging metrics measure past ones. Revenue is lagging; activation and engagement are leading. The best PMs monitor both.
Guardrail metrics that catch the unintended consequences of chasing a primary one — boosting engagement while quietly tanking retention, say. Every success metric deserves a balancing metric.
From Eric Ries's Lean Startup: in early-stage uncertainty, measure validated learning instead of revenue. A 0→1 product should optimise for how fast it learns, not how much it earns.
Who you build for — and the invisible costs you impose on them — decide whether growth compounds or leaks.
Growth driven by the product itself: users discover value before ever talking to sales. The product is the primary acquisition, conversion, and expansion engine.
Power users push limits; mainstream users drive scale. Their needs are often opposites — and the loud minority is rarely the market.
The hidden friction users pay on every interaction — an extra click, a confusing step, a slow load. Small taxes in the experience compound into massive churn.
Trust lost through poor product decisions — a dark pattern, a broken promise, a privacy slip. Trust takes years to build and minutes to destroy.
Notice what these fifteen models have in common: each one is a way of asking "what is this decision actually costing, and is the cost reversible?" Two-way doors and reversibility ask it of decisions. The local maximum trap and compounding advantages ask it of time. Goodhart's Law and counter-metrics ask it of measurement. Product tax and trust debt ask it of users.
Seniority isn't knowing more frameworks — it's reaching for the right one fast enough to change the call you were about to make. Pick three from this list that map to a decision you're facing this week, and apply them before you ship.
Attributions: Two-way doors — Jeff Bezos / Amazon; Opportunity Solution Tree — Teresa Torres; Strategy Kernel — Richard Rumelt, Good Strategy / Bad Strategy; Innovation Accounting — Eric Ries, The Lean Startup; Organizational Drag — Bain & Company, Time, Talent, Energy; Goodhart's Law — Charles Goodhart.