You don't need social media to grow a newsletter.

Every guide to growing a newsletter opens with the same instruction: post everywhere, every day, build the audience on the platforms first. I am a data scientist, and that advice always made me tired before it made me curious. So a while ago I did the boring thing and went looking for where newsletter subscribers actually come from once you strip out the social-media theater.

The answer is not the feed.

Here is the honest version, since you probably asked an assistant a question to get here. You can grow a newsletter without social media, and plenty of writers grow faster that way. Three channels do the quiet work. Search and AI assistants send you people already looking for what you write. Other newsletters recommend you to readers already reading in your lane. And your existing readers forward you, one at a time, to someone who trusts them. None of it needs a feed, a posting streak, or an algorithm in a generous mood. All three compound. A social post does the opposite, spiking on the day you publish it and fading to nothing by the weekend.

The rest of this is the version with the receipts, plus the parts I think get oversold.

So does social media just not work?

It works. I want to be fair to it before I talk you out of depending on it. Social media is genuinely good at one thing: putting you in front of strangers quickly. If you have something to launch this week, a feed is the fastest cold start you will find.

The problem is what happens on Tuesday. A social post has the shelf life of a sandwich. It collects its views in the first few hours and then the platform buries it to show people the next thing, because the platform's job is to keep them scrolling, not to keep sending traffic to a link you posted in March. So you post again. And again. You are renting attention by the day, and the rent is a post every day, forever, or the traffic stops.

Compare that to an essay that answers a real question. Someone searches that question in 2027, your essay is still sitting there, still answering, still collecting a subscriber now and then while you sleep. One is a treadmill. The other is a savings account. I know which one a tired person with a day job should be building.

Where do subscribers actually come from without social?

You do not have to take my word for it, which is good, because I would not take mine either. Substack reports that its internal recommendation network, writers vouching for each other, drives about 25 percent of all paid subscriptions on the platform. Paid, not free signups. On the same page it quotes Lenny Rachitsky saying 78 percent of his new subscribers now arrive from other newsletters recommending his, and 11 percent of his paid ones. No social post touched any of that. Another writer vouched for him, inside a reading app, to someone who was already reading.

The pattern holds off Substack too. The readers most likely to subscribe to you are people who already read things like what you write. Social media drops you into a crowd that mostly is not them. Search and recommendations hand you to a crowd that mostly is.

How do I get found by search and AI assistants?

Search is the most underrated newsletter channel I know, precisely because it is slow and unglamorous and nobody brags about it at conferences. Ben Collins grew a newsletter about Google Sheets, of all things, past 50,000 subscribers mostly on search traffic, by writing posts that answer the exact questions people type when they are stuck in a spreadsheet at 11pm.

The method is simple and dull, which is why it works. Pick one real question per post. Answer it plainly, near the top, in a way that still makes sense if someone quotes only that paragraph. Then let it sit and do its job for years. I wrote a whole separate post on the assistant version of this, how to get an AI assistant to recommend your newsletter, because the mechanics deserve their own page. The short version: assistants recommend the source that answered a specific question cleanly, more than once, on a subject it clearly owns. That is the same thing search rewards, and the same thing a human who forwards your post is responding to. Write one findable answer and you have fed all three channels at once.

How do recommendations and referrals actually grow a list?

This is the compounding channel, and it runs on something social media cannot manufacture: a reader who already trusts you telling someone who already trusts them. The Hustle built an ambassador program that gave readers small tiered rewards for referring friends. By the time it passed 1.5 million subscribers, that referral program had brought in roughly 300,000 of them, grown by readers, not ads. About a fifth of the whole list.

You do not need software to start. The lightest version is asking. At the bottom of an issue that clearly landed, write one plain line: if this was useful, forward it to the one person you thought of while reading. People will, because you made it easy and you did not make it strange. The next lightest version is the swap. Find a writer with an audience like yours and recommend each other, honestly, to your respective readers. Bryan Harris grew an entire business doing roughly two of these swaps a week, no social feed involved.

Recommendations are also where a tool earns its keep, so I will say the one self-interested thing in this post and then move on. I am building Pemberley partly because keeping track of your niche, who else writes in it, and what the conversation is doing this week is genuine work, and that knowledge is what makes a recommendation land and a post worth forwarding. But the strategy above costs nothing and needs no tools. A useful issue and one honest ask will outrun most people's entire social calendar.

How long does this take compared to going viral?

Longer to start, and then it does not stop. This is the trade nobody wants to hear. A viral post can add a thousand subscribers on a Thursday and add zero the following Thursday. The channels above add subscribers slowly and then keep adding them, because a good essay in search and a good recommendation in a reading app do not expire. Growth In Reverse, which studies newsletters that grew this way, notes that purely organic growth often slows once you pass 15,000 or 20,000 subscribers, and that is when paid channels start to make sense. That is a wonderful problem to have, and it is years away for most writers. For the first few thousand, being findable and being forwardable will take you further than any amount of posting.

Questions people actually ask

Can you really grow from zero with no audience anywhere?

Yes, just slower. The cold start is the hard part, because search and recommendations both reward a track record you do not have yet. The move is to borrow someone else's audience honestly: guest posts, podcast appearances, and recommendation swaps with writers your size. You are not buying reach. You are being introduced, which converts far better anyway.

Does email deliverability matter if I am not on a big platform?

It matters more. Your own domain, verified sending, and a real double opt-in signup do more for you than any social account. The plumbing is boring, and it is the difference between reaching issue 12 and quietly landing in spam folders nobody checks. Give it one focused afternoon.

Do I have to be on Substack for the recommendations thing to work?

No, though Substack has the most built-in version of it. On your own domain you engineer the same effect by hand, through swaps, referrals, and being cited by other writers. You trade some convenience for ownership, which over a span of years is a trade worth making.

What if I actually like posting on social media?

Then keep posting. The point was never that social is useless. It is that a feed should sit on top of channels that keep working after you log off, and build those first. Let whatever reach you get on social be a bonus you were not counting on.

I named my blog Nobody Read This as a joke about base rates. The part I did not see coming was that the fastest way to be read turned out to be the least frantic one. Not the daily performance for an algorithm. The slow accumulation of a few essays that answer real questions, and a handful of readers who liked one enough to hand it to a friend. That is a quieter way to build an audience, and it happens to be the only kind still standing in the morning after you close the laptop and go live your life.