A newsletter with four hundred readers can produce more authority than one with forty thousand. That sounds like consolation for people with small lists. It is not. It is the whole argument, and it inverts almost everything written about newsletters, which treats subscriber count as the scoreboard.
Here is the mechanism. Authority is a verification asset. It is built when a specific group of people who matter in your category come to associate your name with a specific claim, and then repeat that association to other people. Four hundred readers who buy, hire, invest, write, or refer in your market will do that. Forty thousand people who subscribed for a free template will not.
Which means the decision to start a newsletter is not a distribution decision at heart. It is a positioning decision wearing a distribution costume. And the founders who get this wrong fail in a predictable place: issue nine, when the initial enthusiasm has burned off and there is no structural reason for the newsletter to exist.
Seven questions, answered before you send issue one.
Why do most authority newsletters die at issue nine?
Issue one is easy. You have been saving that idea for months. Issue two is fine. Issue three still has momentum. Somewhere around issue nine, you sit down on a Tuesday morning with nothing queued, a full calendar, and the dawning sense that you signed up for a weekly obligation with no defined subject.
That is not a discipline problem. It is a design problem. The newsletter was never given a job, so every issue requires you to invent one from scratch, which is exhausting in a way that writing itself is not.
The second failure mode is scope creep in the other direction. The founder decides the newsletter will cover “industry news and insights,” which sounds safe and turns out to be the hardest possible assignment, because it means competing with actual trade publications who have staff. You will lose that fight, and worse, you will lose it while producing content that says nothing about what you believe.
The third failure mode is the one that stings. The newsletter succeeds on engagement metrics and produces nothing. Good open rates, kind replies, zero effect on your standing in the market, because the writing never took a position anyone could attach to your name. Pleasant is not the same as memorable.
There is a fourth failure that hides inside apparent success: the newsletter becomes a company update. Product releases, hiring news, a conference you attended. This happens because company news is the easiest thing to write and requires no thinking. It also guarantees the newsletter is about you rather than about your reader’s problem, and nobody outside your payroll has ever wanted a monthly update on a company they do not work for. If you need a channel for company news, build one. Do not let it eat the newsletter that was supposed to carry your position.
All four failures have the same root. There was no standing question.
What is the standing question your newsletter answers?
The standing question is the single question your newsletter answers again, in a different way, in every issue, for years. It is the most useful concept I know for anyone who wants to start a newsletter that builds authority rather than filling an obligation.
Not a topic. A question. Topics are nouns and they sprawl. Questions have shape, and a good one generates issues without end because reality keeps supplying new instances of it.
Consider what Ben Thompson has done with Stratechery since 2013. The standing question is roughly this: given how the internet changes the economics of distribution, what is the strategy behind this company’s move? Every issue is a fresh instance. A new acquisition, a new platform policy, a new earnings call, and the same question gets applied. He built Aggregation Theory out of answering it enough times that a pattern emerged, published the framework in 2015, and it became a piece of vocabulary the industry uses. Stratechery also proved something structural: a single writer with a paid daily update could support a business, years before Substack made that a category.
That is what a standing question buys you. Compounding. Each issue makes the next one stronger because you are building an argument rather than producing entries.
Your version should be narrow enough that you could be wrong. “How do B2B companies price themselves into a corner, and how do they get out?” is a standing question. “Insights on B2B pricing” is a topic. The first one has a claim buried in it. The second one is a folder.
Write your standing question in one sentence. If you cannot, you are not ready to send issue one, and no amount of platform research will fix that.
Who is it for, and who is it not for?
The second half of that question is the one that does the work. Naming who the newsletter is not for is what makes it feel like it was written for somebody in particular.
Write your reader down as a person, not a segment. Not “marketing leaders at mid-market SaaS companies.” Try: the VP of marketing at a company between fifteen and eighty million in revenue who has been told to fix pipeline, has a team of six, has already tried the obvious things, and is losing an argument with a CFO about attribution. Now you know what they already believe, what they are tired of hearing, and what they would forward to a peer.
The forwarding behavior matters more than it sounds. Authority spreads through forwards, not through subscriptions. A newsletter that gets forwarded inside companies reaches the people you would never be able to reach with an ad, and it arrives with an implicit endorsement from a colleague. Write every issue with a specific forward in mind: who is going to send this to their boss, and what line will they paste above it?
Exclusion also protects you from the slow drift toward the generic. When you know who the newsletter is not for, you can turn down the subject that would broaden your audience and dilute your claim. That trade looks bad on a growth chart and good on a two-year view.
What goes in an issue that a blog post cannot do?
If your newsletter is a blog post delivered by email, you have built a distribution redundancy rather than an asset. There has to be something in the inbox version that could not exist on a public page.
Three things qualify. The first is candor. Email feels like a room with a door on it, and readers respond to writing that acknowledges that. You can say the thing you would not put on a landing page: the mistake you made last quarter, the tactic you now think is wrong, the client engagement that went badly and what it taught you. That candor is the fastest trust-building mechanism available in business writing, and it does not survive being turned into marketing copy.
The second is work in progress. Show the argument before it is finished. Publish the half-formed pattern and ask readers whether they have seen it too. A newsletter is the only format where an unfinished idea is a feature, because the reader is being let in early rather than sold to.
The third is the reply loop. Every issue should make a reply easy, and every reply should get an answer from you. Not a thank-you. A real answer. Do this for a year and you will have a private research asset about your market that no competitor has, and a hundred people who feel a personal connection to your work. Those hundred people are the referral engine that shows up in year two.
There is a format point buried in all three. Write in the first person and keep the voice you use in a meeting. Corporate newsletter voice, the passive institutional register where nobody is responsible for any sentence, destroys the one advantage email has. A reader opening your issue in an inbox full of automated marketing is looking for evidence that a person wrote it. Give them that in the first paragraph, and skip the branded header image that pushes your opening line below the fold on a phone.
What does not qualify: news roundups, recycled social posts, and anything that reads like it was assembled rather than written.
How often should it ship?
Weekly if you can hold it. Every other week if weekly would degrade the writing. Monthly is a mistake for authority purposes, because a month is long enough for a reader to forget who you are between issues, and the newsletter never becomes a habit.
The honest test is not what you can do in a good month. It is what you can do in the worst month of your year. Pick the cadence that survives a fundraise, a product outage, and a family emergency, then hold it without negotiating with yourself weekly.
Consistency matters more than volume for a reason that has nothing to do with algorithms. Predictable arrival is what converts a newsletter from a thing you send into a thing your reader expects. Expectation is the beginning of habit, and habit is what makes an audience durable when your posting slows down or a platform changes its rules.
Length should follow from the standing question, not from a template. Some weeks the honest answer takes six hundred words. Some weeks it takes two thousand. Padding to hit a word count is visible to readers and it teaches them to skim.
One practical mechanic: keep a running file of half-ideas that relate to your standing question, and add to it whenever something irritates you during the workweek. Irritation is the most reliable source of material a practitioner has. The blank page on a Tuesday morning is what kills newsletters, and the file is the cure.
Where should it live, and who owns the list?
Three requirements, and the platform choice falls out of them. You must be able to export your subscriber list in full. You must be able to send from a domain you control. You must be able to publish a public web archive of every issue.
The first two are about ownership. Any platform can change its terms, its pricing, or its posture toward your category. A list you can export is portable. A list you cannot is somebody else’s asset that you are renting.
The third requirement is the one founders skip, and it is the one that connects the newsletter to everything else you are building. A public archive means every issue becomes a crawlable page. Search engines index it. AI answer engines retrieve from it. Over two years, a weekly newsletter with a public archive becomes a hundred pages of text arguing one coherent position under your name, which is a stronger signal of category authority than almost anything else you could build in the same time.
This is why I push against publishing only inside a walled platform. The email does the relationship work. The archive does the retrieval work. You want both, and the cost of having both is a checkbox.
Growth deserves one paragraph, not a chapter, because for an authority newsletter the acquisition problem is smaller than people think. The list grows from three places: your existing network, the public artifacts that point at it, and other people’s audiences. Import the first. Put a real pitch for the newsletter at the end of everything you publish elsewhere, describing the standing question rather than saying “subscribe to my newsletter.” And write guest pieces or appear on shows where your reader already is. Paid acquisition works for media businesses selling ads and works badly here, because it buys volume in a game scored on who reads rather than how many.
One thing the platform choice does not determine: whether anyone reads it. Founders spend weeks comparing feature matrices and then send issues nobody opens. When you start a newsletter, the tool is the least consequential decision on the list.
How do you know it is working?
Open rate is a hygiene metric. It tells you your emails are arriving and your subject lines are not broken. It tells you nothing about authority.
Watch four things instead. Reply quality, meaning how many readers write back with something substantive rather than a compliment. Forward evidence, meaning subscribers who joined because a colleague sent them an issue, which you find by asking on the signup form. Citation, meaning your ideas or phrases showing up in other people’s writing, decks, and conversations. And inbound specificity, meaning prospects who arrive already using your framing, which is the clearest sign your position has traveled.
Unsubscribes deserve a specific note, because founders read them as failure and they are more often the opposite. When you take a real position, some people leave. That is the position working. The unsubscribe you should worry about is the slow kind, where nobody leaves and nobody opens, which means the newsletter has become background noise that people tolerate. A spike in unsubscribes after a sharp issue, paired with a spike in replies from the people who stayed, is the healthiest pattern this format produces.
Those four move on a slower clock than open rate. Expect two quarters before the pattern shows. The founders who quit at issue nine quit before any of the real signals had a chance to appear.
The compressed version: pick a standing question you could be wrong about, write to one named reader and exclude everyone else, put something in the email that could not exist on a public page, ship on a cadence you can hold in your worst month, keep the list and the archive under your own control, and measure replies and citations rather than opens. Do that for a year and the newsletter stops being a marketing channel. It becomes the record of what you think, which is what authority is made of.