The Diff

The Diff went paid in early 2020, timed to a fear and a moment. It started as writing for fun, then became a way to stay in the loop for Hobart's next job. Fearing COVID would block job interviews, he made it paid in early 2020.

Spring 2020 suited online content: white-collar workers had time freed from commuting and spent it on work-adjacent reading, which industry newsletters fit well.

How I Write

The newsletter grew like a bundle: break even on one thing, profit on another. Hobart compares it to gas stations, where gas brings foot traffic and the money comes from lotto tickets, cigarettes and snacks. The newsletter is the traffic; a reader network built around it is the second business.

The gas station post was itself a surprise hit, which he says worked because it filled a real gap in his own knowledge. His reaction was mostly "huh", since it is hard to predict what readers don't know but will find interesting.

How I Write

Smart readers form a network worth more than the subscriptions. Sharp comments from founders and curious readers led Hobart to connect them, and that side is growing faster than subscriptions. Doubling subscribers roughly doubles the value of the writing, while doubling a network more than doubles its connections.

How I Write

Compete on timing plus quality by writing what an event meant. Being first is a random, one-dimensional race, like Joe Weisenthal pre-loading two tweets for jobs day and hitting enter on the right one. Hobart writes the definitive explanation, since readers always want a piece they can cite when explaining a topic to someone else.

Longer pieces do slightly better for this reason. He says the success rate beats racing to be first but will never approach 50%. Quote: "there's always demand for the last important piece on it."

How I Write▶ 8:03

Pick a narrow topic so you can be the one who worked hardest on it. Hobart cites Robert Caro quoting LBJ's line "if you do everything, you will win" and then doing everything himself. When no one owns a single topic, pair two of your interests; most pairings have no good analogy, but one may have no competition.

Effort only weakly predicts which pieces take off, and many newsletter writers complain their favorite pieces do poorly. Experience raises the odds: going from a 50% to a 90% shot at a piece he is proud of raises output, and output brings the next ideas.

How I Write

Writing the newsletter he would open first draws like-minded readers. Curious people who want to go deep on odd business and tech topics self-select in, so reader quality needs little direct management. He doesn't worry about smart readers running out.

A narrower domain focus draws smarter people inside that domain; his approach suits a smart general audience or specialists with general interests. In any niche, assume everyone will hear about your work, since its talented people fit in a group chat.

How I Write

Ask readers and experts for the case against your ideas. Hobart seeks people less likely to say everything is great and asks for the counterargument. He also covers topics he is less sure about so experts write back with context he lacks, which feeds the next piece.

Quote: "what is the case against this is really effective."

How I Write▶ 1:03:06

One big notes file grows one-line ideas into finished posts. Entries range from "look into company X" to paragraphs with points, links and quotes. Some become ready posts, so writing is just copy and paste; others are a starting point for more research.

Letting notes accumulate lets his viewpoint shift halfway through, which he calls annoying but fun. Books in his weekend links email often signal a coming post, since he reads the book on a company or industry first.

How I Write▶ 0:51

Write by day, sleep on it, send the next morning. A draft is far easier to edit after sleep, and it leaves room for overnight news. Once he had written that COVID would matter less if a vaccine ever arrived, and the next headline said vaccine tests succeeded, so he rewrote it.

He wakes around 5 AM, and by 4 PM his calendar is spoken for until 8. While employed, he wrote for a half hour at lunch with a timer and no internet, adding links afterward. The distractions keep improving, so his defenses must keep evolving.

How I Write▶ 15:50

Writing as you read makes knowledge stick and compound. Linking what you read to other reading shows which patterns hold and which break, helping retention and comprehension. His bookshelf test: point to a random book and ask how long you could talk about it.

He argues memorized facts make connecting ideas far faster than searching, though he doesn't memorize systematically; retention comes from the volume he reads.

How I Write

Add commentary so the link roundup beats Hacker News or Reddit. Plain link lists are available elsewhere, so Hobart builds in a take, often a meta point about why a piece was written rather than its content.

He says you can understand a story better than its writer by asking who is talking and why, and why a source leaked to this reporter now.

How I Write▶ 3:13

Assume some readers find your point new, and elaborate. An industry cliche can be unknown outside it, so his bias is to explain. In tech and finance he can skip AWS, but in liquid natural gas he explains how it works, weighed against an angry email from a 30 year veteran.

How I Write▶ 1:01:20

Split an overgrown draft into Part One and publish the rest tomorrow. When a topic needs more than he can finish today, Hobart adds "Part One" to the title and publishes the rest the next day.

He calls deciding how much to do one of his weakest areas, and asks whether a piece is getting far more or less attention than it deserves.

How I Write▶ 58:10

Recheck the title against the finished piece. His titles are often plain, like "Understanding Jane Street". His mistake is titling early, then finding the theme belonged to a footnote he deleted.

He says that with Twitter and screenshots, quotable sentences now matter more than headlines for newsletters.

How I Write▶ 56:55