The reading list
Essays, tech notes, and the in-between.
- Product·3 min read
The Press Release That Kills Products: Amazon's PR/FAQ, As Amazon Tells It
Prime launched February 2, 2005 at $79 a year to a reception Amazon itself remembers as 'crazy and stupid.' The Kindle launched November 19, 2007 at $399 and sold out in under six hours. Amazon's insiders credit both to a document — the PR/FAQ — and the honest version of that story is more interesting than the legend: the mechanism's real product isn't the hits, it's the products that died on paper before anyone built them.
- Product·3 min read
The Product Owner's Stakeholder Craft: Run It as a System
The worst stakeholder blowup I've been part of wasn't caused by a bad decision — it was caused by a good decision that a VP learned about at release. Stakeholder management isn't a soft skill you improvise in the hallway; it's a system you run: a living map, a pre-wiring cadence, a decision log, an escalation contract, and a trust ledger you're always either funding or draining.
- Product·3 min read
Watch What They Do: Segment, Superhuman, and Discovery That Worked
Segment's founders built a classroom tool that professors loved — until Peter Reinhardt watched students open Facebook the moment lectures got boring, and offered all eight investors their money back. Superhuman's Rahul Vohra took the opposite route: he turned the fuzzy question 'do people want this?' into a single segmentable number and moved it from 22% to 58% in a year. Two very different shapes of discovery, one shared principle: behavior, not opinion, is the input to strategy.
- Product·3 min read
Pricing Strategy for New Ventures: Price Before You Build
Roughly 72% of new products miss their revenue targets, and the cause usually isn't the product — it's that nobody asked what anyone would pay until the thing was already built. Simon-Kucher's Madhavan Ramanujam calls this designing the plane and then discovering physics. Here's the discipline new ventures skip: testing willingness to pay before the roadmap exists, choosing the pricing metric before the price, and treating monetization as an experiment stream instead of a launch-week scramble.
- Product·3 min read
Running Quarterly Planning Without Losing the Quarter
The worst quarterly planning I ever ran took three weeks, produced a beautiful deck, and changed nothing. The best took two, ended before the quarter started, and the plan survived to week twelve. Here's the runbook: the four input docs, the three sessions, the 70/30 commitment split, and the in-quarter cadence that keeps the plan alive without re-litigating it every sprint.
- Product·3 min read
$1.75 Billion of False Validation: Quibi, Juicero, and the Humane Pin
Quibi raised $1.75 billion and lasted about six months. Juicero raised $120 million for a value proposition that a reporter falsified with her bare hands in thirty seconds. Humane sold roughly a tenth of its Pin target and watched HP buy the wreckage for $116 million. Three eras, three categories, one identical discovery failure: each had 'validation' — surveys, investor enthusiasm, demo applause — and none had evidence of the kind a good interview or a behavioral test produces. The uncomfortable through-line is that money made it worse.
- Product·3 min read
How to Back the Right Startup: Picking in a Power-Law World
Venture returns follow a power law: a handful of deals produce most of the money, and half of everything you back goes roughly to zero. That single fact inverts every risk instinct a corporation has spent decades perfecting — because in venture, the expensive mistake isn't the startup that failed, it's the outlier you screened out for looking weird. Here's what the evidence actually says about picking: timing, teams, markets, and the anti-portfolio discipline of studying your misses.
- Product·3 min read
Agile in the AI Agent Era: The Ceremonies Die, the Principles Don't
Every agile ceremony was derived for a world where writing code was the bottleneck. AI coding agents are ending that world — implementation collapses toward cheap, and review becomes the constraint. Sprints, estimation, and standups start running on fumes; the four values come through fine, because they were never about typing speed. They were about uncertainty and feedback, and those don't move.
- Product·3 min read
When the Metric Becomes the Target: Wells Fargo, Zynga, Builder.ai
Wells Fargo's 'eight is great' cross-sell target produced up to 3.5 million unauthorized accounts, roughly 5,300 firings, and a $3 billion resolution — $500 million of it from the SEC, because the ratio itself was the number sold to Wall Street. Builder.ai claimed roughly $220 million in revenue against about $55 million of reality and was insolvent within weeks of the discovery. Between them sits Zynga, an honestly metrics-driven company that lost 40% of its value in a day. Three escalating answers to the same question: what happens when the number stops describing the business and starts being the business?
- Product·3 min read
The Planning Stack: Vision to Quarter Without Repeating Yourself
Most companies' 1/3/5-year plans are the same slide deck at three zoom levels — the annual plan is the quarterly plan times four, the 3-year plan is the annual plan with bigger numbers. That repetition is the tell that no planning happened at the longer horizons. Here's the fix: five documents, five different questions, and the smell test that reveals when each one is fake.