I Used Claude to Analyze 30 Viral Substack Posts. The Pattern Nobody Talks About
I analyzed posts with 862, 997 and 1-2k likes from creators my size. Claude for Chrome helped me find what they all had in common — and it completely changed the way I write articles
Recently I came across a post that got 947 likes and almost 300 comments. The author had 1.3k subscribers. Same as me. I've been on Substack for a year and my best post has 152 likes. Most of mine land somewhere between 20 and 40.
Same audience size. 6x the result.
I could have told myself she got lucky, or that her topic was more popular, or that she’d been on the platform longer. Instead, I opened Claude (with extension Claude for Chrome) and spent a Saturday doing what my PhD training taught me to do: treat it as a research problem.
I found the same pattern in every single viral post I analyzed. And it has nothing to do with posting consistently or engaging in the comments — the things everyone tells you to do.
Here’s what I found.
What I Actually Studied
I analyzed 30 posts from 30 different creators in the Substack. All of them had subscriber counts similar to mine — between 1,300 and 1,500.
All of them had one post that dramatically outperformed everything else they’d published — from 862 to more than 1-2k likes, with hundreds of comments and restacks. And I’m not talking about Notes, which go viral much more easily. I mean long-form posts.
I read every word of every post and then fed them into Claude for Chrome, using a specially designed prompt for this purpose. I analyzed the structural patterns that made these posts go viral.
Here's what came back. And this is what will make your posts more readable and re-shared!
They're fairly easy to implement. Just try them.
The First Pattern: Every Viral Post Had a Reusable Resource Layer
This is the thing I hadn’t seen anyone name directly.
Most posts are consumed and forgotten. You read them, think “interesting,” close the tab, and move on. They might get likes from people who already follow you. But they rarely travel beyond your existing audience because there’s nothing to take away, nothing to save, nothing to send to a friend.
Every single viral post I studied had something different: a layer of content that readers could use independently of the post itself. Something with value after they close the tab.
In the 862-like post, it was The 4 Things the Successful 89 Creators Did Differently.
In the 997-like post, it was 10 Claude Prompts that Can Replace an Expensive Consultant.
In the 1,2k-like post, it was 3 KPIs I’m Tracking + Your Creative Curriculum.
Every post gave readers something to do, not just something to read.
That single element is the difference between a post that gets 40 likes from your subscribers and a post that gets 400 likes from people who’ve never heard of you. Saved posts get reshared. Reshared posts bring new subscribers.
It compounds. But only if you give people something worth saving.
→ Try this:
Look at your last published post.
Ask yourself: is there anything in it that a reader could screenshot, save, or forward to a friend?
If not, add one thing — a short list, a framework, a set of questions, a single reusable prompt. It doesn’t need to be long. It needs to be detachable.
The Second Pattern: The Opening Had to Create Recognition, Not Introduction
We all know how important it is to start a post correctly so the reader wants to read the article to the end. All the viral posts I analyzed with Claude had one thing in common.
The viral posts opened by describing a situation the reader was already in, in more precise language than the reader could use themselves.
The 884-like post opened like this: “I spent 9 months commenting on Notes daily, restacking generously, building relationships. My engagement was great. My revenue was $0. So I analyzed 100 Substack writers making over $10,000 per month. 89 of them barely engaged compared to what I was doing.”*
That’s not just a hook. That’s a mirror. Every struggling Substack writer recognized themselves in that opening paragraph before the author had offered a single piece of advice. And when readers feel genuinely seen, they share the post because sharing it feels like showing someone else a mirror that finally fits.
The question to ask yourself before publishing: does my opening describe my reader’s situation, or does it describe my topic? Those are different things, and the first one converts at a much higher rate.
→ Try this:
Ask yourself before publishing: does my opening describe my reader’s situation, or does it describe my topic?
If it describes your topic, rewrite the first two sentences of your post. Instead of introducing your topic, describe the exact moment your reader is in right now — the frustration they feel, the thing they tried that didn't work, the question they can't stop asking.
Read it back and ask: would someone forward this opening to a friend and say "this is literally us"? If yes, you're done. If not, go one level more specific.
The Third Pattern: The Title Was a Mechanism, Not a Label
Most titles describe what a post is about. Viral titles do something structurally different, they create a reason to click before the reader has read a single word.
Looking across the 30 posts I studied with Claude, every high-performing title used at least one of three mechanics:
A specific number that makes a vague promise feel concrete and deliverable. “10 Claude Prompts that Replaced a $500 Consultant” works because the number 10 implies a list you can finish, use, and keep. “Prompts that replaced a consultant” without the number feels like an opinion. With it, it feels like a tool.
A broken rule or violated assumption. “I Analyzed 100 Substack Writers Making $10K+/Month. 89 Broke This ‘Rule’” works because the reader arrives with a belief — that successful people follow the rules — and the title immediately contradicts it. The dissonance is the click.
An unexpected word in an expected phrase. “How to Become Disgustingly Creative in 2026” would be invisible as “How to Become More Creative.” The adverb does the entire job. It makes the reader pause, smile, or frown and any of those reactions is better than scrolling past.
None of these titles label a topic. Each one creates a small, unresolved tension that can only be relieved by reading.
→ Try this:
Take the title of your next post and run it through these three questions. Does it contain a specific number?
Does it break an assumption the reader holds?
Does it contain at least one word that would make someone pause mid-scroll?
If the answer to all three is no, rewrite it until at least one answer is yes. You’re not changing what the post is about, you’re changing whether anyone opens it.
The Fourth Pattern: The Argument Attacked a Belief the Reader Already Held
This is different from having a surprising title. A surprising title gets the click. This pattern is what earns the share and shares are what carry a post beyond your existing audience.
Every post in my study that crossed 800-1000 likes made the same structural move inside the content: it identified something the reader already believed to be true, presented evidence that it was wrong, and then offered a better framework to replace it.
The 997-like post didn’t just say AI can replace consultants. It opened with a concrete story — a friend who paid $5,000 for a two-day strategy session with a boutique firm — and then showed, step by step, how the same analysis took 45 minutes with Claude.
The belief being attacked wasn’t abstract. It was something the reader had probably done, or considered doing, or told themselves was worth the money.
The 2k-like post on anxiety didn’t just introduce an interesting theory. It took a belief that millions of people carry as identity — I have an anxiety disorder — and reframed it as a learned behavior that can be unlearned. That’s not a surprising title mechanic. That’s a direct challenge to something deeply personal, backed by an argument.
The pattern is always the same: name the belief explicitly, show the evidence against it, and give the reader something more accurate to hold instead.
When that works, readers don’t just like the post. They send it to the specific person in their life who needs to read it. That’s the mechanism behind every share spike I found in the data.
→ Try this:
This one is optional — not every post needs to challenge conventional wisdom, and forcing it where it doesn’t fit will feel hollow.
But if you have a post coming up where you genuinely see something differently from most people in your space, try this: write one sentence before you start drafting that completes “Most people believe ____. I think they’re missing ____.”
Don’t use it as a rule. Use it as a lens. Sometimes it will crack the whole post open.
When it does, make sure that tension appears on the page, name the belief explicitly, then show your readers why you see it differently. That’s the move that turns a good post into one people send to a friend.
One Last Thing
None of the four patterns I found require a bigger audience, a more popular topic, or years on the platform. The creators who got 1-2k likes have the same subscriber count as me.
What they all had in common wasn’t luck or timing. It was a structure — something they built into the post before they hit publish. A reusable resource. An opening that felt like a mirror. A title that created tension. An argument that named something the reader believed and then quietly took it apart.
You don’t need all four in every post. But the more of them you use, the further the post travels — beyond your subscribers, beyond your niche, into the feeds of people who’ve never heard of you and might just hit follow when they’re done reading.
That’s the compounding effect. And it starts with the next thing you write.
If this post was useful and you feel like saying thanks, a small coffee donation goes a long way. It truly fuels my drive to keep exploring and writing. Every bit of support means the world! ☕️
Dear Substackers, what's your biggest bottleneck right now - your title, your opening, or finding the reusable thing worth saving? Drop it in the comments. I read everything and I'll tell you exactly where to start.
Many of the things you mentioned, I have been applying by default. But I come from communication and I have written a lot of content for others. I think the most important thing is if the information you provide gives real value to the reader. Does it strike a nerve? Does it create tension? Does it validate something we often miss? One of my llatest post did that and performed well. Not in likes, but in views. I also gree from 6 to 21 subscribers thanks to one post. I did leverage a dual system in which I pulled people from my network to Substack, but that is always a gamble. As people rarely want to leave the main platform. That post was saved on LinkedIn 50 times, so I know the content resonated. I even got a DM from someone saying they liked my post. The reason was simple: people felt seen.
Even a quiet presence here carries weight. You’ve already built something meaningful — now it’s about continuing to engage and evolve.Please subscribe and leave a reaction on my posts — it genuinely helps.