How Substack’s 2025 Algorithm Actually Works
Why it suddenly feels so “cozy” — and the restack/recommendation trick small creators are using to explode
Limited time: Get the Top 10 Substack Take-off Secrets free (paid members only after December 31st)
The ten foundational steps to turn your writing into a sustainable business. Stop guessing what to write, who to write for, and how to make money. These ten steps give you the complete business foundation successful creators built before they “made it.”
I discovered Substack a few years ago through a crypto YouTuber who was using it as a platform to publish his newsletter—another newsletter platform, okey. But it was only recently that I discovered the Notes feed, and I noticed something weird.
It felt different. Clean. Genuine. Nothing like the other platforms I came from.
While other social media feeds feel like drinking dirty water from a sewer, Substack’s Notes feed tastes like fresh water from a mountain spring.
That got me thinking. Why?
It was only last month when the senior members of the Substack team gathered a group of publishers in New York City for a behind-the-scenes look at their newly redesigned Notes algorithm. They answered the questions I had—and probably the ones you have too:
How is the algorithm actually built? What is it optimized for? How does it decide what to show? Why does it feel so different from everything else?
Mike Cohen (the architect behind it) and Hamish McKenzie (the co-founder) explained how the algorithm is built and why they built it this way in the first place. I’ve distilled everything they explained into simple terms. Because the first step to master something is to understand it.
What I found has the potential to change how we think about social media forever. We’ve been waiting for a better way. Turns out, Substack built it.
Algorithms Are Necessary
“Algorithms, for lack of a better word, are good.”
I’m paraphrasing Gordon Gekko in the 87’ Wall Street movie.
Now seriously. Social media platforms need algorithms to function. The internet is drowning in content. Without something to filter and rank it, we wouldn’t have time to go through all the garbage there is online.
That’s why algorithms are necessary. The problem are not the algorithms. It’s how they’ve been designed for the past decade. Traditional social media platforms have weaponized them using psychological tactics borrowed from Vegas casinos and coded by the sharpest nerds in Silicon Valley, all to maximize one thing: How long you stay on their platform.
Here’s How the Typical Social Media Algorithm Work:
Time-Based You see what was just posted, in chronological order. Simple and fair—everyone’s content gets equal visibility based on when it drops.
Creators: Post whenever. No need to optimize for the algorithm. Equal shot at visibility.
Consumers: See random content. Miss stuff you’d love, but also discover things organically. Less addictive.
Platforms: Lower time-on-platform. Less data to monetize. Revenue suffers.
You-Based (Personalization) The algorithm studies your behavior—what you like, interact with, follow—and shows you content tailored to your interests.
Creators: Have to understand their audience deeply and create content that matches what the algo learns triggers engagement. Constant optimization needed.
Consumers: See incredibly relevant content. Feels like the feed knows you. Addictive. Easy to lose hours without noticing.
Platforms: Maximum time-on-platform = maximum ad impressions = maximum revenue. Users stay hooked.
Popularity-Based The algorithm shows what’s getting the most engagement from everyone—trending posts, viral content, what’s blowing up.
Creators: Have to create polarizing, emotional, or sensational content to cut through the noise. Playing the viral game.
Consumers: See what’s trending and big. Great for discovery, but sensationalism dominates. Outrage and FOMO drive your emotions.
Platforms: Sensationalism = engagement = time spent = revenue. Users keep coming back to see what’s viral next.
Most platforms blend all three. They’re hybrids, mixing these signals into their unique recipe.
Each one asking: How do we keep people here as long as possible while making it feel natural?
Substack Has Changed the Game
Traditional social media and Substack are optimized for completely different outcomes.
That’s the paradigm shift that changes the way we consume social media forever.
Traditional platforms make money from ads. So their algorithm asks: How do we keep users scrolling forever? Longer scrolls = more ad impressions = more revenue.
Substack makes money when creators make money. So their algorithm asks: How do we connect readers with creators they’ll actually pay to support?
The Core Difference
Traditional Social Media: Optimization for Time Spent
Core Goal: Keep users stuck in an endless scroll. More time on platform = more ads seen = more money.
Practical Goal: Users must never want to leave the app or seek out a long-form post or deeper experience. Leaving = lost revenue.
How It Works: Algorithms are largely based around time spent. Every design choice is built to make you stay longer.
Substack Notes: Optimization for Conversion (Value)
Core Goal: The platform doesn’t optimize for time spent. It optimizes for discovery, sign-ups, and subscriptions.
Practical Goal: The platform does nothing to stop a user from leaving Substack except providing enough value to make them want to stay.
How It Works: The platform succeeds when a reader receives enough value to subscribe and pay. That’s their whole game.
The Practical Difference
Traditional algorithms prioritize engagement at any cost—viral outrage, engagement bait, controversial takes. Whatever keeps you scrolling.
Substack’s algorithm optimizes for genuine connection. It asks: Does this reader actually care about this creator’s work? Will they subscribe? Does this add real value?
One platform succeeds by making you addicted. The other succeeds by making you a paying fan.
How Notes Actually Works: The Engine of Discovery
Substack Notes has become the beating heart of the platform and the main traffic source to the platform’s core product: paid newsletters.
Here’s how it works:
Notes Feed as Discovery Engine
Notes is the main driver of subscriptions and revenue growth on Substack. In just three months, the app drove nearly half a million paid subscriptions and over 32 million free ones.
Gateway to Your Work: Notes helps creators reach beyond their existing audience and introduce new people to their full Substack. It’s a short-form entry point to long-form work.
Articulating Your Value: Creators use Notes to show who they are and what they do. Short videos, illustrations, casual observations—a sneak peek of what subscribers get when they sign up.
The Algorithm Notices: The Notes feed matches readers with content based on their language, location, interests, subscriptions, and follows. It’s asking: Will this reader eventually pay for this?
Community Over Individualism
If you are only going to remember one thing from this article let it be this one: Substack’s algorithm actively rewards community behavior.
Restacking and Replying Matter: When you restack someone’s note (share it with your own commentary) or reply, you’re signaling to the algorithm that your audience might like that writer. The algorithm notices these overlaps and shows similar readers to similar creators.
Generosity Gets Amplified: Notes that explicitly recommend subscribing to another writer get massive algorithmic boost. The platform is literally rewarding creators who help other creators.
Build communities, Not Just Feeds: If the algorithm sees audience overlap between creators, it creates a “virtuous cycle”—both writers reach new readers. It’s collaborative growth, not zero-sum competition.
Consistency Counts: Publish Notes regularly. It keeps your work visible, helps the algorithm understand your voice, and keeps the feed dynamic for your readers.
The entire system is designed around one idea: creators helping creators reach the right readers. Sounds different, right?
How to Actually Grow on Substack Notes
The best way to use the Substack algorithm to grow your audience is to align your activity with the algorithm’s core function: drive discovery, sign-ups, subscriptions, and paid conversions.
This is the opposite of chasing viral moments or engagement metrics. It’s about building a real business.
Here’s how to leverage it:
1. Be Consistent and Clear
The algorithm needs to understand who you are and what you offer. Make its job easy.
Publish Regularly: Consistency beats virality every time. Show up on a schedule. Make your job to at least publish a note once a day. Your audience recognizes you through repetition.
State Your Value Proposition: Don’t make people guess. Use Notes to show who you are and what subscribers get. Share teaser ideas. Point back to long-form posts. Share a personal story. Whatever communicates what you offer.
Treat Every Note as Discovery: Someone’s seeing your work for the first time. Don’t be afraid to be repetitive.
Mix All Content Types: Short-form, long-form, memes—the algorithm has no preference. It just cares if people engage and eventually pay.
2. Engage with others
Growth happens through community, not isolation. The algorithm rewards creators who lift others up.
Restack and Reply: Restacking (sharing with commentary) and replying to others helps the algorithm understand audience overlaps. Show your community what you think about other creators’ work.
Share Long-Form Posts: Restack others’ posts and quotes. It introduces their audience to you while helping them reach new readers. You’re building bridges between audiences.
Real Connection Over Engagement Bait: For small audiences especially, focus on genuine interaction. Thoughtful replies, authentic likes, meaningful restacks beats engagement bait every time. The algorithm notices authenticity and helps you be visible to their audiences.
3. Recommend others (Your Secret Weapon)
This is where the algorithm really takes notice.
Explicit Recommendations Get Amplified: Notes that directly recommend subscribing to another writer get massively amplified. Phrases like “I encourage you to subscribe” or “I recommend reading this” often reach thousands of people outside your current network.
Build Reciprocal Partnerships: Instead of one-way recommendations, find a few creators doing complementary work and commit to mutually elevating each other. You recommend them. They recommend you and both audiences grow.
If You’re Just Starting focus on this:
Substack is built for you to own your audience and build an independent business. Every action should ladder back to one goal: driving subscriptions and paid conversions.
Here’s where to focus:
Engage First, Broadcast Second: You’re new. Don’t worry about being heard yet. Spend time engaging with other people’s Notes—liking, replying, restacking work you genuinely admire. The algorithm will start showing your content to their audiences.
Show Up Consistently: Post regularly—ideally once a day if you can manage it. Consistency teaches the algorithm your voice and helps new readers recognize you over time.
Be Crystal Clear About Your Value: Don’t make people guess what they get by subscribing. Show them. Share your ideas, tease your best long-form work, post videos—whatever communicates your unique value.
Build Genuine Connections: Restack and recommend other creators you actually believe in. Authentic support gets amplified. The algorithm rewards superfans who lift others up.
Test Strategic Partnerships: Find complementary creators and commit to mutually recommending each other’s work. When you both use explicit language like “I recommend subscribing,” the algorithm notices and both audiences grow.
The Bottom Line
Substack’s algorithm isn’t trying to trap you. It’s trying to help you build something real.
Publish consistently. Engage authentically. Be clear. Show your value. Help others. Recommend publications. Repeat.
Do that, and your audience will find you. More importantly, when readers subscribe to your newsletter. You own that relationship. No algorithm change can take that away from you.
➤ Next week: I came to Substack with traditional email marketing tactics. Here are 8 lessons I learned that changed my approach after 30 days of daily publishing. Read next week’s issue.
➤ If you liked this article, help me restacking it 🔄