Why I Hate Substack Notes
And you should too
Substack Used to Be Different
When I first started using Substack, it was for the long-form content. Quality writers were moving to the platform to dive deep into topics that couldn’t or wouldn’t be published in traditional outlets. I came here for Paul Krugman, Robert Reich. I enjoyed reading their in-depth analysis.
I started using Substack for my own writing, because it gave me a place to write without setting up domains, mailing lists, and websites. I liked the low barrier to entry.
And then came notes.
Today—even though I’m a paid subscriber—I have to search for each Paul Krugman article. The writers I read regularly don’t show up at the top of my subscription list. My home page is littered with notes. Most of them are notes by gurus telling me to write more notes. Ninety percent of the notes are fluff and filler people write because the gurus tell them to, and they want to be seen.
I didn’t come to Substack to replace my Facebook feed. I don’t even use Facebook. I don’t use Twitter (or X, if you want to call it that). I’m not looking for more sound bites or performative notes. I want deep, relevant content that makes me think. That takes the time to go deeper, analyze facts, apply rational arguments. That was what Substack brought to the table—quality content.
Here’s what bothers me most. It’s not that Notes exist. It’s that Substack chose to let Notes drive discovery instead of building something better.
Substack sits on top of an enormous catalog of long-form writing. Years of serious, substantive work across thousands of publications. Political analysis, literary criticism, science journalism, craft essays, fiction. The kind of content people used to pay magazine subscriptions to access. It’s all there, indexed, searchable, full of distinctive voices saying specific things about specific ideas.
And instead of using that catalog to connect readers to writing they’d actually want to read, the algorithm pushes whatever generated engagement in the last 24 hours. Notes. Restacks. Reactions to other people’s notes. The feed optimizes for activity, not depth.
Spotify doesn’t show you a feed of musicians talking about music. It plays you music.
When Spotify wants to surface a new artist, it drops a track into your Discover Weekly playlist. Thirty seconds of audio that either grabs you or doesn’t. The recommendation is made through the work itself. You hear something that sounds like something you already love, and you follow that thread.
Substack could do the same thing with writing.
If I’ve read three essays by the same author about the intersection of craft and technology, the algorithm knows something about what I’m looking for. It knows the sentence length I tolerate, the argument structures I follow to the end, the topics that keep me scrolling past the point where I should have stopped reading. That’s not vague data. That’s a detailed reader profile built from actual reading behavior.
Use it. Pull a paragraph from a post I haven’t seen yet — one that matches the texture of what I’ve already engaged with — and show it to me. Not a headline. Not a social post about the post. A paragraph. Let the writing do the work of recommendation.
If the paragraph catches me, I’ll click through to the full piece. If it doesn’t, move on. That’s how you surface long-form content. Not by building a social layer on top of it and hoping the social layer points people back toward the writing.
The Feed Ate the Platform
The technical capability for this isn’t far-fetched. Substack has the reading data. They know which posts I open, how long I spend on each one, where I stop reading. They know which writers I’ve subscribed to and which posts I’ve paid to access. That behavioral signal is far more precise than a like on a note.
Pair that with the content itself. The platform could analyze thematic overlap between posts — not just keyword matching, but the kind of semantic similarity that captures when two writers are working on the same intellectual problem from different angles. A reader interested in AI and writing craft should be surfaced work from writers they’ve never encountered who are wrestling with the same questions, even if those writers never use the same terminology.
That’s a meaningful discovery engine. That’s the kind of thing that would make Substack genuinely different from every other platform competing for writers’ attention.
Instead, the most visible path to discovery is posting more notes.
Substack Made the Same Mistake as Everyone Else
I understand why Substack built what they built. Social features drive engagement metrics that investors understand. Daily active users go up when there’s a feed to scroll. Notes create network effects that pure email newsletters don’t have. The business logic is there.
But the platform built its reputation on being the place where serious writers do serious work. That’s the promise that attracted the writers, which attracted the readers, which made the whole thing worth paying for. Eroding that promise to optimize for engagement metrics is the same trade-off every social media company has made, and it’s produced the same result: a feed that rewards performance over substance.
The writers doing the deep work are still here. They’re still publishing. They’re just harder to find.
Substack has the tools to fix that. Snippets, semantic matching, behavioral reading data — use the catalog. Surface the writing through the writing. That’s the discovery engine this platform should have built from the start, and it’s still not too late to build it.
I Won't Feed the Notes Machine
In the meantime, I’ve made a decision about how I use this platform.
I’m not going to feed the notes machine. I won’t write notes about writing notes. I won’t post three times a day to stay visible. I won’t perform engagement to please an algorithm that’s pulling the platform away from the thing that made it worth being on.
The only notes I’ll post are restacks of my own work — pointing people toward the essays I’ve actually written, with the full piece one click away. That’s it. If the algorithm doesn’t reward that, I’ll accept the tradeoff. My time is better spent writing something worth reading than generating social content to promote the idea that I write things worth reading.
There are other platforms. Ghost gives writers full control over their publication without the social layer. Beehiiv is building serious infrastructure for writers who want to grow an email list without becoming content creators. If Substack keeps optimizing for engagement over depth, those platforms become more attractive — not just for writers, but for readers who came here for the same reasons I did.
I’m not leaving yet. The catalog is still here, and so are the writers I came for. But I’m not going to be complicit in turning this place into something it was never supposed to be.
Write the work. Publish it. Let it stand on its own. That’s the only strategy I’m interested in.
Are you still here for the long-form content, or has the feed changed how you use the platform? I'd genuinely like to know whether other readers and writers are drawing the same lines.
Good points. I'll make a note of it.
Really appreciated how clearly you articulated something many readers and writers have probably felt but not fully named: the frustration is not simply that Notes exist, but that they increasingly mediate discovery on a platform people originally came to for substantive long-form writing. Your argument was especially strong when you contrasted engagement-driven visibility with writing-driven discovery. The Spotify analogy worked well, and your idea that Substack should surface the work through the work itself via excerpts, semantic matching, and actual reading behavior felt both persuasive and practical. 
What stood out to me most was that it was really a defense of medium integrity. You are not arguing against conversation, but against a recommendation architecture that risks rewarding performance over depth. I also thought the essay captured an important tension in the current creator economy: platforms often attract serious work by promising substance, then gradually incentivize the habits of social media once scale and engagement become the priority. That is a real structural problem, and you described it with clarity and restraint!
If I had one quick suggestion, it might be to push even further on what a healthier middle ground could look like for writers who do want discovery without feeling conscripted into feed behavior. For example, a few concrete product ideas around reader-controlled filters, long-form-only discovery modes, or “show me posts, not notes” settings could make an already strong critique even more constructive. But even as it stands, the piece succeeds because it names the tradeoff so directly.
Overall, this was an excellent publication. Thank you for defending the value of depth in an ecosystem that too often confuses activity with substance. It reminds both writers and platforms what made long-form publishing worth protecting in the first place!