Constant, Gentle Pressure

Over half the price of a median Manhattan condo is a zoning tax. America's affordability crisis is policy-made — ranked by impact: zoning, permitting delays, rent control, then compliance.

The telephone and the light bulb were invented in several places at once, which makes progress feel inevitable. Some things aren't — electric cars had no reason to exist, and neither does what we're building.

Reliable and predictable aren't the same thing. Reliability adapts to the situation; predictability just repeats. The best people and products manage both — stable enough to trust, surprising enough to stay interesting.

In Starcraft, by the time you scout what your opponent is doing, it's too late to change plan. Market trends work the same way — scouting tells you where your existing strategy can win, not which strategy to switch to.

Management can't fix problems it doesn't know exist, or reward work it never saw. Surface the problem, propose three solutions, then recommend one — that's the whole method.

Half the managing partners at one firm don't use email daily. Their number one question is still "what can we automate?" — which tells you how painful the manual processes have become.

Founders say they expect their values to degrade at scale, as if it were physics. Sam Walton wore a hula skirt on Wall Street and kept shrinkage at half the industry average — big going bad is a choice.

Selling Manhattan for $24 is the standard example of a terrible trade. At a 6.5% real return over 398 years, that $24 becomes $1.8 trillion — enough to buy the island back with billions to spare.

Min/max the question: what does a country look like at 80% institutional ownership, and what does it look like when everyone owns? Both extremes have upsides worth naming and costs worth admitting.

Homeownership rates fall as countries get richer, and US cities with the highest rates have the weakest economies. Detroit and Pittsburgh sit near 75%; LA and Boulder near 55%.

$100K in the S&P in 1928 is worth $55 million today. The same $100K in the average US house is worth $1.1 million. Housing became an "investment" only in the 1980s — and the holding costs rarely make the story.

The claim that institutions bought 44% of homes in 2023 got repeated by governors. The actual figure was 0.4% — wrong by two orders of magnitude, and it keeps circulating anyway.

Institutions own 3% of single-family rentals, and the backlash is already loud. But scale buys 1,000 HVACs cheaper than one — and homeownership carries costs nobody counts.

Homeownership joined the American dream after World War II, pushed there by government programs. But between zoning, permitting, and HOAs, what most people call ownership today looks a lot more like an investment.

We say we want innovation, then mock anything genuinely new. What we're actually addicted to is predictable novelty — the annual launch, the weekly match, the sequel. Different every time, inside a familiar frame.

Wearables promised data-driven health and we got more sedentary anyway. The same gap exists in real estate — almost no institution knows what each function actually costs in time and spend.

How many people in your life will tell you something that upsets you? Everyone has blind spots, and nobody can advise you on a game they've never watched you play.

Average StarCraft players watch the battle. Top players keep building while it happens. Exceptional macro with good-enough micro wins — in the game, and in running a company.

Zoning is the fastest way to artificially limit supply, and limited supply is what drives real estate prices up. Tokyo's street-by-street model shows what the alternative looks like — rising population, flat prices.

Functional, Efficient, Best in Class, Best Imaginable. Four stages that decide how much a given feature deserves — and knowing which one you're aiming at is what keeps prioritisation honest.

Most proptech is built by people who have only experienced the resident side, so it solves problems residents can see. In proptech you can go high-end to low-end, but not the reverse.

Saying you don't prioritise just means you prioritise whatever is in front of you. Larry Page circled the ads and wrote "THESE ADS SUCK" — he set the priority, not the solution.

Grace Hopper taped the original bug — a moth — into her logbook in 1946. Plenty of things filed as bugs since then, in software and in organisations, turn out to be deliberate design.

Spielberg, Phil Knight, and Steve Jobs managed the same way. Directing, shoes, and technology turn out to run on shared frameworks — which is why going deep in one field makes you faster in the next.

Asimov wrote or edited over 500 books. The people we admire are almost always prolific, and volume is what produces quality — which makes discipline the thing worth building, not talent.

You want a bow and arrow, so you build a sharp stick first. Refactoring too early sharpens a stick that's about to be thrown away; too late and you're rewriting the product from scratch.

Most people manage a startup like architecture — strategy, planning, execution. It behaves more like a garden: plant, water, prune, pull weeds, and keep experimenting with what makes things grow.

Mudita is the opposite of schadenfreude — joy at someone else's good fortune. Since you can't undo another person's success anyway, being glad about it is the cheapest upgrade available.

Three men laying bricks: one has a job, one a career, one is building a cathedral. Real estate's costs haven't fallen despite a decade of proptech because nobody changed the underlying system.

Should the janitor be fearless with new cleaning agents? Being 100% of any core value is as ill-conceived as being 100% of any virtue — Ned Stark's honesty is the cautionary tale.

You can judge an organisation by how it handles edge cases. A blocked domain nobody could unblock, and a venue that permits mini horses but not a caged rabbit — both are the same failure.

You can be competent, disciplined, and prioritising correctly, and none of it matters if you're inefficient. Functional, efficient, best in class, best imaginable — knowing which one a task deserves is the whole skill.

Manage too loosely and the work drifts; push too hard and people scatter. Steady, measured guidance is what keeps small standards from quietly collapsing into big ones.

Our minds have limited RAM, and how we spend it is how we live. Putting things in a box and closing them is a learnable skill — and good software should help you do it, not add to what you're carrying.

QWERTY isn't the best keyboard layout, and the fastest typists in the world still use it. Infrastructure lock-in means good enough usually wins — until the gap between what's done and what's possible gets too wide.

A casual player can't tell the 800th best tennis player from the 10th — both beat them easily. Below a certain competence you can't judge performance at all, which makes knowing your own gaps the more useful skill.

Two opposing things can both be true at once — in life and inside a company. Holding both in mind at the same time is how teams weigh real trade-offs instead of letting one function win by default.

The most common reason people are wrong is that they started from the wrong axioms. The second is applying the wrong framework on top. Neither is something we were trained to examine.

Enron's core values were integrity, communication, respect, and excellence. If the opposite is absurd, the value isn't a value — it's a platitude. Real ones carry a trade-off.

Businesses can succeed without compassion — plenty have. But Carnegie found that less than half of a labour dispute was ever about wages, and a united team still beats a talented one that can't stand each other.

Startups trade market risk against execution risk. We deliberately took the second — a finite set of workflows, hundreds of them, all of which have to work in one system on a foundation nobody wanted to rebuild.

A birthday in a note field. A licence plate buried in a tenant profile. Unstructured data doesn't look expensive until someone has to click through every record to find one thing.

Debt isn't only financial. A missing feature is product debt; bad UX is design debt. Used deliberately it's how a company moves fast — as long as you know in advance what will trigger paying it back.

A leaking sink, a call to a personal line, a promise forgotten — and a tenant who thinks nobody listened. Scattered communication doesn't stay a small problem at scale.

Founders build for tenants because they were once tenants. But owners are the ones paying for the software and carrying the operational complexity — so when the two experiences conflict, we design for the owner.

Someone proposed a $100 weekly lunch cap. The cap itself was a rounding error — the dozens of processes needed to enforce it were the real cost. Wait for things to break before adding structure.

Lessons from the first three years of building Ender — why pilot customers make or break a company, why the team stayed technical, and what it means to become the operating system for real estate.

Listing sites are the tip of the iceberg. The real operating system of real estate is the accounting system — and imagine if every accounting system, Quickbooks included, had a closed API.

There's technical debt, and there's managerial debt. Transactional workplaces print the second kind — and it comes due exactly when teams need to trust each other to weigh a hard trade-off.

Ideas that hijack our identity make us seek confirmation and reject dissent. Being open minded means no idea controls us — which is what turns conviction into something earned rather than assumed.

Existing property management software is a decrepit skyscraper where every floor had a different contractor and the elevator has been out for years. Rebuilding the foundation is the interesting problem.

Startups have inertia like everything else, and nothing outside pushes them to be great. Fearlessness is the force of will that replaces that push — and it starts with asking the questions everyone else is too embarrassed to ask.

Startups lack checks and balances — that's a feature, not a bug. Autonomy is a double-edged sword, and it only works when the people holding it care about results rather than authority.

Mario Kart's time trial has you racing the ghost of your own best lap. Companies that watch competitors instead copy their way to mediocrity — the harder question is what your own ghost would do.

A map of the proptech landscape, from listing sites to end-to-end platforms. Ten categories, what each one actually does, and why the ERP layer underneath determines what the rest can achieve.

A non-profit offering free telemedicine to hundreds of thousands of tenants couldn't get names and phone numbers out of their own software. That's the state of the systems running the largest asset class in the world.