1. How We Met on Substack

We started talking on Substack less than a month ago, Keith Keller and I crossed paths while exchanging thoughts on audience acquisition, the subtle difference between followers and subscribers, and how creators can optimize their public profiles.

What stood out to me immediately was Keith’s complete transparency.

Every few days, he shared raw numbers, celebrated small milestones, and openly discussed what failed.

That ongoing dialogue quickly evolved from casual comments into a real collaboration.

2. Meet Keith Keller: The Man Behind Social Audio Decoded

Keith Keller is the creator of Social Audio Decoded.

His publication helps creators host audio events, build communities, and foster real relationships without ever needing to turn on a camera.

Applying a Blue Ocean Strategy, Keith focuses on creating a distinct space for audio conversation rather than copying what everyone else is doing with video or text.

He regularly hosts audio lives, tests growth tactics, and looks for new ways to connect with listeners across Substack and LinkedIn.

3. Tracking Growth in Real Time: From 575 to 606 Subscribers

By October 11, Keith’s dashboard showed 606 subscribers and 60 new subscribers over the previous 30 days.

The screenshots he shared during our conversations captured part of that progress in real time, moving steadily from 575 to 606.

It wasn't an overnight explosion driven by a single viral post.

Keith credits much of his progress to taking the initiative and starting conversations. We haven’t measured how much each individual change contributed.

4. Follow First, Follow Back, Follow Up: Proactive Networking

When I asked Keith what drove the majority of his momentum, his answer was clear: taking the initiative rather than waiting for the algorithm.

"Improving the bio has definitely helped, but the real power is in actively following other accounts first and taking the initiative." Keith Keller

Keith breaks his method down into three steps:

1. Follow First: Proactively search for and follow creators in your niche.

2. Follow Back: Build reciprocal connections with people who show interest.

3. Follow Up: Nurture relationships through genuine replies, direct messages, and chat invitations.

The lesson here isn't to trade subscriptions or join engagement pods.

It's about starting relevant conversations and giving people a genuine reason to stay.

For instance, Keith personally reaches out to followers who haven't subscribed yet, inviting them to join his Substack Chat and asking what topics interest them most.

5. Profile Optimization and Practical Experiments

Keith treats his publication as an ongoing laboratory.

During our early exchanges, he ran his publication through Subshack’s Profile Audit tool, scoring a 95/100.

"There is no doubt that your site has encouraged me to make a few changes and the results have been amazing." Keith Keller

Beyond sharpening his bio, Keith tests various operational shortcuts:

• Updating Evergreen Posts:
  

  When pressed for time, Keith rewrites and refreshes existing articles to improve SEO and readability rather than rushing out unpolished new posts.

• LinkedIn Cross Pollination:
  

  Keith noticed that people who were also active on LinkedIn seemed more likely to subscribe after connecting on Substack.

  He believes LinkedIn is an important discovery source, even though we haven't formally verified that attribution.

6. Facing Unexpected Failures: The "Zero Attendees" Live

Growth experimentation also means confronting data that doesn't go as planned.

Recently, Keith promoted an audio event on LinkedIn and gathered 160 registrations.

However, when the live session started, none of the LinkedIn registrants actually showed up.

Rather than getting discouraged, Keith analyzed the gap between registration and live attendance.

He immediately adapted his workflow by testing tools like Luma for automated calendar reminders and structured follow ups.

Treating failed tests as data points is central to his process.

7. From Product Feedback to Mutual Collaboration

Our relationship quickly grew beyond simple software testing into a two way street.

On September 30, Keith invited me onto his Social Audio Tech Check live stream, where Ana Garibay joined us as well.

During the broadcast, he challenged how Subshack’s audit evaluated creators who rely heavily on audio and live events rather than traditional text posts.

His feedback led me to adjust the audit to better account for podcasts and Lives.

In return, Keith helped me with my audio setup and shared my details with people in his podcasting network.

We ended up helping each other showing that building in public is ultimately about human connections.

8. The Road to 1,000 Subscribers

Keith has set a clear long term target:

1,000 subscribers by the Super Bowl in February 2027.

Having a public benchmark keeps him focused on testing tactics, hosting live sessions, and building genuine relationships one conversation at a time.

9. Key Takeaways for Substack Creators

Keith Keller’s journey offers clear takeaways for any creator looking to grow:

1. Your profile is a storefront:
   

   A clear bio gives strangers an immediate reason to subscribe.
   

2. Relationships drive acquisition:
   

   Initiating conversations (Follow First, Follow Back, Follow Up) outperforms passive waiting.

3. Embrace failed tests:
   

   When registrants don’t show up, as Keith says happened with his LinkedIn sign-ups, it isn’t a defeat; it’s data guiding your next follow-up sequence.

4. Collaboration is a two way street:
   

   Share feedback, support fellow creators, and build real relationships behind the scenes.

I started by helping Keith review his profile.

He ended up helping me improve my product and connecting me with other creators.

That exchange is the part of this story I value most.

Keith writes Social Audio Decoded about hosting audio conversations and building community. To see what a stranger understands from your own profile, try the free Subshack Profile Audit.

On August 12, 2026, I stopped all subscribe-for-subscribe. No reciprocity threads, no “sub back”, no engagement pods.

That chart is the whole story, so let me say it plainly: quitting sub-2-sub did not restore my open rate. The damage was already done. When my list was 53 people, 34% of them opened. Reciprocity grew the list and diluted it.

Half the people on my list didn’t want to read my work. They wanted a subscriber back. Most of them are still on the list today. They just don’t open. That is what a ghost audience looks like in the data.

Here is what happened over 54 days: the drop, the system I used instead, and where the numbers stand today.

Phase 1: The initial shock (days 1-14)

When you quit sub-2-sub cold turkey, the dashboard hits you first.

1. Reach per Note collapsed. I published 54% more Notes to compensate (13 → 20). Total impressions still fell 28% (587 → 420). Per Note, reach fell by more than half: 45 → 21 impressions. What disappeared were reciprocal likes, not readers.

2. The vanity metrics froze. The automatic likes disappeared. The subscriber counter stopped ticking on its own.

This is the moment most creators panic and go back to engagement pods and reciprocity threads.

Phase 2: The 4-part organic system (days 15-54)

To build an audience of real readers without transactional hacks, I committed to four habits.

1. One hour a day of targeted replies

Instead of broadcasting generic Notes into the void, I spent one hour a day writing thoughtful replies under Notes from creators my readers already follow.

Result: 19 subscribers came through Notes in a single week, with no viral Note.

2. A free diagnostic instead of “please subscribe”

I built a free Substack profile audit. Instead of asking people to subscribe, it scores their profile clarity out of 100 in about 10 seconds.

Result: a clear profile converts about 1 in 10 profile visitors into subscribers.

3. Radical data autopsies

On August 19, I published a raw dissection of my own mistakes: I Checked the Data Behind My Own Growth Advice. It Was Worse Than I Thought.

Result: 164 views and 20 comments, my best-read issue since quitting sub-2-sub. Dissecting real failures builds more trust than broadcasting fake success.

4. Real creator pain points

On September 10, I wrote about a friction point every creator was feeling: Substack’s AI Detector Is Creating a Problem Nobody Expected.

Result: a 31.6% engagement rate, my highest of the period, plus 30 likes and 2 new subscribers.

Phase 3: Where the numbers stand at day 54

64 readers were active in the last 30 days.

That is 30.6% of the reachable list, and the real size of this publication. Not the subscriber count.

Behind them: 312 subscribers, all free, and 437 followers.

• Recipients per issue 226 → 302 (+34%)

• Open rate 21.2% → 13.2%. It did not recover.

• People opening each issue between 36 and 48, every single issue

• Best click-through rate 23.7% (Sep 1) and 18.6% (Sep 14), against 0-6% for the issues sent Aug 2-10

• Peak engagement 31.6% on a creator pain-point post

• Subscriber intent from “sub back” to search, recommendations and profile audits

The open rate is the uncomfortable number. It keeps falling because the reciprocal subscribers from my first weeks are still on the list, and they don’t open. The number that matters stayed flat: the same ~40 people read every issue while the list around them grew. My next step is cleaning the list so the open rate shows real readers again.

Key takeaways for Substack writers

1. Transactional growth builds a directory, not a publication.

2. Quality beats volume. 100 subscribers at 34% or 200 subscribers at 18%? Both give you about 35 readers. The second list costs twice the effort to reach the same people.

3. Roots grow in the dark. If your counter feels slow because you chose quality over hacks, keep going. Count the people who open, not the people who subscribed.

Find the 3 things costing your Substack subscribers

The free profile audit scores your Substack profile out of 100 and shows what a first-time visitor sees in 10 seconds.

Audit my profile

+450 likes. 12 profile visits. 0 new email subscribers. That is the end-of-month reality check for a creator posting Notes six times a day and it is far more common than any viral screenshot suggests.

If you spent any time on social media between 2020 and 2022, you learned a specific growth formula: post 3 to 5 times a day, hijack trending topics, hunt for the algorithm’s favor, and volume will eventually force growth.

When Substack introduced Notes, thousands of creators imported this exact playbook. They started treating Notes like a short-form video feed posting 6 times a day, dropping quick hot takes, and celebrating when a Note hit 100 likes or 50 restacks.

Then came the end-of-month reality check:

• Vanity likes: +450

• Profile visits: 12

• New email subscribers: 0

Chasing social dopamine on Substack is the fastest path to creator burnout.

The TikTok illusion vs. the Substack reality

Short-form algorithmic feeds (TikTok, Reels) reward passive dopamine consumption. Users scroll endlessly, hit “like” as a reflex, and forget the creator’s handle 3 seconds later.

Substack operates on an entirely different economic and psychological model: permission-based email relationships.

As Sarah Fay, PhD, highlights in Post Less Often, Please, flooding the Notes feed with low-effort content actually harms your publication. It trains your readers to skim past your name, lowers your perceived authority, and clutters the feed without ever giving a reader a reason to visit your publication.

What 4,500+ Notes reveal about posting cadence

When we analyzed Yana G.Y.’s dataset of 4,553 Substack Notes, the numbers dismantled the “post 5x a day” myth:

1. Volume does not scale conversion. Writers posting 15+ Notes a week saw a steep decline in profile click-through rate per Note.

2. Intent beats virality. Single, high-signal Notes sharing a real milestone, a clear breakdown, or a data drop generated 3x more profile taps than generic inspirational quotes.

3. The 1-to-3 rule. Publishing 1 to 3 thoughtful Notes per day paired with active, helpful replies under other writers’ work consistently outperformed brute-force mass posting.

The 3 shifts every Substack writer needs to make

If you want to stop feeling like you’re speaking into the void, make these three strategic adjustments:

1. From “impressions” to “profile taps”

Stop asking “How many people saw this Note?” Start asking “Does this Note give a reader a compelling reason to tap my profile?” A Note with 200 views that yields 10 profile visits is 10x more valuable than a viral Note with 20,000 views that yields zero clicks.

2. From “broadcast” to “conversation”

Replies under other creators’ Notes predict subscriber growth 2x better than raw likes. Instead of dropping cold links, leave insightful, generous comments where your ideal audience already hangs out.

3. Anchor your profile before you scale distribution

Distribution without an optimized profile is just leaking traffic. Once your bio clearly states who you serve, what problem you solve, and why to subscribe now, every single Note you write becomes significantly more profitable.

The high-converting Notes cadence

Four habits and the order matters: the profile comes before the volume.

Fix the leak before you post again

Ten Notes a day into an unanchored profile is the 2021 playbook with extra steps. The free Substack profile audit scores your profile out of 100 in 10 seconds no signup and tells you what to fix before your next Note.

Audit my profile

How many Notes a day are you posting right now and how many profile visits did they bring last month? Drop both numbers in the comments. 👇

Imagine walking down a busy street looking for dinner. You stop in front of two restaurants:

• Restaurant A has a clear menu posted on the door: exact dishes, prices, ingredients, and dining options. In 3 seconds, you know whether it fits your taste and budget.

• Restaurant B has a closed wooden door with a tiny brass plaque that reads: “Exploring the quiet culinary spaces between taste and memory.”

Which restaurant do you walk into?

On Substack, 9 out of 10 creators are running Restaurant B. They spend dozens of hours writing articles and Notes, only to send interested profile visitors to a bio that refuses to state what the publication is actually about.

The cost of a “closed door” bio

According to Sarah Fay, PhD, the single biggest headline and profile mistake writers make is confusing poetic self-expression with functional clarity.

When a reader taps your name on Notes or clicks your link from social media, you have a 3-to-5 second window before they decide to hit “Back” or tap “Subscribe”.

Across hundreds of profile audits, three recurring “closed door” bio patterns destroy conversion rates:

1. The Personal Diary: “Notes on books, coffee, life, and my random thoughts.” A complete stranger has no incentive to give you their email address for your personal diary.

2. The Vague Visionary: “Rethinking the future of humanity and work.” Too broad it promises everything and delivers nothing specific.

3. The Silent Storefront: a bio that states a topic but includes zero call-to-action, and no indication of what happens next.

The 3-part high-converting bio formula

To convert cold profile taps into dedicated subscribers, your bio must act as a clear storefront menu. In Yana G.Y.’s Conversion Blueprint and Subshack’s 100-point profile framework, high-converting bios follow a strict 3-part structure:

1. The specific target & transformation (who & what)

State exactly who you write for and what specific problem you solve.

• Weak: “Writing about newsletter growth.”

• Strong: “Helping Substack writers convert profile visits into email subscribers with real data.”

2. The proof or cadence anchor (why trust you)

Give the reader a reason to trust your authority, or a predictable expectation.

• Examples: “Based on 4,500+ analyzed Notes” · “10+ years in B2B SaaS” · “Every Saturday morning.”

3. The micro call-to-action (what to do next)

Never leave the visitor guessing. Direct them to your free tool, pinned post, or welcome offer.

• Example: “Audit your bio in 10 seconds below 👇”

Passing Sarah Fay’s “Bus Test”

Sarah Fay challenges every Substack creator with the Bus Test:

Imagine sitting next to a tired stranger on a bus. If they ask what you write about, can you explain it in one sentence so clearly that they immediately understand its value even if they aren’t your target audience?

If your bio requires a reader to scroll through 5 past articles just to figure out what you do, you fail the Bus Test. Readers don’t subscribe to vague content; they subscribe to clarity and predictable transformation.

The 100-point profile menu checklist

Five checks, weighted to 100 the same grid the free audit scores. Positioning alone is worth more than a third of it.

Put a menu on your door in 10 seconds

Stop letting interested profile visitors walk away because your front door is locked. The free Substack profile audit runs the 100-point diagnostic on your profile in 10 seconds no signup and tells you which of the three parts your bio is missing.

Audit my bio

What does your current bio look like? Drop it in the comments below, and let’s rewrite it together into a high-converting 3-part menu. 👇

1 in 10 profile visitors subscribes. That is the number that decides whether a Substack grows and most writers never look at the page where it happens.

When Yana G.Y. analyzed 4,553 Substack Notes using AI and raw database metrics, the funnel was brutally simple:

• 567,867 impressions generated 9,289 profile visits a click-through rate of just 1.64%.

• 9,289 profile visits generated 929 new subscribers an exact 10.0% conversion rate.

This exposes the single most important rule of Substack growth: the job of a Note is only to earn the profile tap. Your profile description bears 100% of the conversion burden to turn that visitor into a subscriber.

When 1 in 10 profile visitors subscribes, a publication grows predictably. But when a profile is vague, poetic, or unanchored, that 10% collapses to less than 1%. You can publish 50 Notes a week, but if your profile leaks visitors, 99% of your distribution effort is completely wasted.

The quicksand illusion: why it looks like a Notes problem

Most Substack writers spend hours perfecting their Notes, testing hooks, and chasing viral impressions. They see their subscriber count tick upward, celebrate the milestone, and assume their growth engine is running smoothly.

Then they notice an uncomfortable contradiction: open rates begin to drop falling from 34% down to 18% or 14% while their reader list becomes increasingly silent.

Why? Because they treat Substack Notes as the entire growth strategy, when Notes is actually just the billboard. The real conversion happens on the public profile and right now, 90% of creators are sending traffic to digital real estate built on quicksand.

The 3 structural cracks in most Substack profiles

Across hundreds of profile reviews and audits, three recurring structural flaws explain why visitors leave without subscribing.

1. Failing Sarah Fay’s “Bus Test”

Sarah Fay, PhD, poses a fundamental test for every Substack creator: Imagine you are seated on a bus next to an exhausted construction worker who asks, “What’s your Substack?” How do you answer so that your value is instantly clear, even if it’s not their specific topic?

Most writers fail this test in their bio. They write poetic statements like “Exploring the quiet spaces between thoughts” or “Notes on life, literature, and coffee.” A stranger reading that in 3 seconds has no idea what problem you solve or why they should hand over their email address.

2. The attribution gap (chasing vanity metrics)

When we analyzed the raw data behind our own publication at Subshack, we discovered that 19 subscribers arrived through Notes, yet zero were attributed to a specific individual Note. Overall, 66% of that subscriber cohort came through the Notes channel as a whole.

Writers who rely solely on default dashboards often assume their Notes aren’t working because no single Note shows “+1 subscriber”. In reality, readers discover you through multiple touchpoints, tap your profile name, and subscribe based on the clarity of your profile setup.

3. Treating your profile like a personal bio instead of a store window

Your profile description is not an obituary or a personal diary. It is a high-converting storefront. In 10 seconds or less, it must answer three non-negotiable questions:

1. Who is this for? (Target audience)

2. What specific transformation or insight do you deliver? (Clear value proposition)

3. Why subscribe now? (Proof + clear CTA)

The 100-point foundation: five categories, weighted

The free audit scores a profile out of 100 across five categories. Positioning and conversion alone are worth 60 points fix those first.

Find the 3 things costing your Substack subscribers

Before you write your next Note or publish another article, stop pouring water into a leaky bucket. The free Substack profile audit scores your profile out of 100 in 10 seconds no signup and flags the 3 priority issues to fix first.

Audit my profile

Fix those 3, and your profile stops being quicksand and becomes a permanent growth asset.

What score did your profile get? Drop your score and your Substack handle in the comments below, and let’s review your bio together. 👇

• 124 km on Foot, 21 Hours of Running: What Our Journey from Cergy to Rouen Taught Us About Autism and Endurance

• From a 5:00 AM Start to a 2:00 AM Finish: Running 124 Kilometers for Autism Awareness

• Cergy to Rouen (124 km): 3 Men, 1 Bike, 21 Hours for Autistic Children

The air at Cergy Prefecture at 4:30 in the morning has a taste. Cold, metallic, a little damp. It gets into your lungs before you have decided to breathe deeply.

It was pitch black. The square was empty. No spectators, no starting gun, no race number pinned to anyone's chest. Just three men, one bicycle, a few bottles of water, and a distance that nobody in their right mind runs in a single push: 124 kilometers, all the way to Rouen.

We were quiet. Not the confident quiet of athletes who know exactly what is coming. The other kind. The kind where everyone is doing the math in their head and nobody wants to say the number out loud.

There were three of us:

• Rachi, the runner. This was his challenge, his idea, his legs on the line.

• Me, the challenger. My job was simple and slightly cruel: run next to him and keep pushing him past the point where he would have stopped alone.

• Our support rider, on the bike. Water, food, dry socks, photos, videos, and the small miracle of someone always being ten meters ahead, checking that the road was safe.

And one reason.

We did not run 124 kilometers for a medal, a personal best, or bragging rights. There is no medal. We ran to put a spotlight on autistic children and their families, who fight quiet, daily, invisible battles that nobody applauds and nobody posts about.

At 5:00 AM, we started running into the dark.

Act I: Dawn and the first miles (0 to 55 km)

The first hour is a strange kind of lie. Your legs feel fresh. Your breathing settles. Your headlamp carves a small moving circle out of the night, and you start to believe this will be easy.

It will not be easy. But the lie is useful, so you let it carry you.

At 6:00 AM, a friend was waiting for us on the side of the road with supplies. Nothing dramatic: water, something to eat, a few words. But when you have been running in the dark for an hour and a familiar face appears out of nowhere, it lands like a full night of sleep. We refilled. We laughed. We left.

Around 6:30, the sky started to let go. The black turned deep blue, then grey, then a thin orange line opened over the fields. We were running on empty country roads between crops still wet with dew, and the light arrived slowly, the way it does when nobody is watching.

Our rider dropped back, framed a shot, rode ahead, framed another. Those first photos and videos are the ones where we still look human.

A rhythm settled in, and it held for hours:

• Two runners side by side, matching cadence without talking about it.

• One bicycle a few meters ahead, reading the road, spotting the cars, calling out the potholes.

• Nobody speaking much. Just the sound of three sets of wheels and feet, and the occasional bird deciding the day had started.

That is what the first 55 kilometers looked like. Long, patient, almost peaceful. The kind of running where the effort hides behind the scenery.

Act II: Kilometer 55, and the power of a community

We reached the Vetheuil and Vernon area after more than five hours of continuous running.

We were tired but intact. And then we turned a corner and found people waiting for us.

Members of the local autism association were there. The town mayor was there. They had come out to meet three sweaty strangers who had been running since before sunrise, and they treated us like the effort mattered.

I want to be honest about what that does to you physically. It is not a metaphor. Your shoulders drop. Your face changes. The pain does not disappear, but it stops being the main thing in your head.

We stopped for twenty minutes. We ate, we drank, we sat down for the first time in five hours, which is its own small trap.

And we recorded a video message for social media, right there, on the spot. Not polished. Not scripted. Just the three of us, out of breath, saying why we were running: for autistic children, for their parents, for the families who do this every single day without a support bike and without anyone waiting at kilometer 55.

Then something better happened. Two other runners joined us. They ran 23 kilometers at our side.

There is a thing that happens when strangers start sharing your effort. The suffering does not go away, it just stops belonging to you alone. Conversation replaces counting. Kilometers pass without being noticed. For 23 kilometers, it stopped feeling like an endurance test and started feeling like a group of people going somewhere together.

That is exactly what the cause is about, by the way. Nobody should have to carry it alone.

Act III: The night, the wall, and the funfair

Then the day turned, and so did the run.

Somewhere between kilometer 80 and 90, we passed the sign for Val-de-Reuil. I remember the sign better than I remember the kilometers around it, which tells you everything about that stretch.

By then the legs were heavy in a way that food and water do not fix. Every small hill felt personal. The conversation died. Running became a negotiation, renewed every few minutes: one more kilometer, then we decide.

You never decide. You just run one more kilometer and open the same negotiation again.

Around kilometer 100, late at night, we came across a funfair. Lights, music, fried food, the whole thing, still running long after we thought the world had gone to bed.

We stopped to eat and to find something resembling strength.

And that is where the strangest, warmest moment of the whole day happened.

We met the travelers and the carnival workers who run the fair. Older men, a few drinks in, absolutely not the crowd you would cast for an ultra-marathon cheering section. We told them what we were doing. Cergy. Rouen. On foot. Since five in the morning. For autistic children.

They went wild.

They cheered us like we had just won something. They asked questions, they shook our hands, they wanted the details, they wanted to know how much was left. They insisted on a group photo, all of us squeezed together under the fairground lights, three exhausted runners and a crew of men who had been working since morning themselves.

No one in that photo had anything in common except the moment.

That is the thing about running for a cause: it builds bridges you never planned to build. You leave with your legs destroyed and your faith in people quietly repaired.

Act IV: Rouen, 2:00 AM

The last 14 kilometers were the longest of my life.

Entering Rouen at night is a silent thing. Empty streets, closed shutters, traffic lights changing color for nobody. Our footsteps echoed off the buildings. The bike rolled beside us without a word.

At that point the body is no longer the problem. The body has already said everything it had to say and been overruled. What is left is the mind, running its own loop: you can stop, nobody would blame you, you have already done enough.

You keep going anyway. Because stopping at kilometer 118 would not have been a failure of the legs. It would have been a failure of the reason.

We finished in Rouen at 2:00 in the morning. Twenty-one hours after leaving Cergy.

124 kilometers. Battered feet. Legs that no longer took instructions. Three people standing in an empty street at two in the morning, unable to explain what they were feeling to anyone who was not there.

And here is the part I keep coming back to.

Twenty-one hours is nothing.

It hurt, it was long, it was hard, and it is still nothing compared to what autistic children and their parents live every single day. Their run has no finish line at 2:00 AM. There is no support bike, no friend waiting at kilometer 55, no mayor at the roadside, no funfair applause. There is a lifetime of small battles, repeated daily, most of them invisible.

We got to choose our suffering, and we got to stop. They do not.

If these 124 kilometers made even one person look differently at an autistic child, ask a better question, judge a parent a little less quickly, or simply understand that the family next door is running an ultra of their own, then every single meter was worth it.

What you can do in ten seconds

We ran for visibility. Visibility only works if it travels.

Please share this post. Restack it. Send it to one person. That single click costs you nothing and takes the cause further than our legs ever could. We covered 124 kilometers; a share covers more ground than that in a second.

And if you know a family living with autism, do the simplest thing: tell them you see them.

Thank you

• To the autism association at kilometer 55, for coming out to meet us and for the work you do the other 364 days of the year.

• To the mayor and the town, for standing at the roadside when you had no obligation to.

• To the two runners who joined us for 23 kilometers and carried more than they realize.

• To the friend at kilometer 20 with the water and the good timing.

• To the travelers and carnival workers at kilometer 100, for the loudest and most unexpected encouragement of the entire day.

• To our support rider, who spent 21 hours on a bicycle so that two runners would never have to think about anything but the next step. There is no version of this story where you are not in it.

• And to Rachi, who decided this was possible before anyone else did.

124 kilometers. 21 hours. Three of us on the road, and one reason worth all of it.

By August 12th, it had dropped to 21%.

By August 20th, it hit 16%.

And on my last two issues in September, it bottomed out at 14%.

During that exact same period, the number of people receiving each email grew from 53 to 266. On paper, the publication looked like it was expanding. In reality, every new issue was reaching a smaller fraction of the inbox than the one before it.

If you publish on Substack, you have probably felt this specific tension: celebrating a milestone subscriber number while watching your open rate quietly bleed out.

Here is the raw data from my emailed posts, why open rates lie (in both directions), what my own subscriber export showed, and the exact protocol I ran on September 15 to clean my list without deleting real readers.

1. The Raw Data: A 20-Point Open Rate Slide

Here is the exact progression across a selection of my emailed posts:

• 3 Aug: 34% open rate (Why Subscribe-for-Subscribe Is Not Really Win-Win) : 18 opens out of 53 recipients · 77 views

• 5 Aug: 23% open rate (How I Grew From 25 to 67 Substack Subscribers...) : 16 opens out of 69 recipients · 91 views

• 7 Aug: 22% open rate (I Hit 100 Substack Subscribers in My First Week...) : 34 opens out of 154 recipients · 223 views

• 12 Aug: 21% open rate (I Stopped Subscribe-for-Subscribe...) : 48 opens out of 226 recipients · 89 views

• 20 Aug: 16% open rate (My Subscriber Count Grew. My Open Rate Fell From 34% to 18%) : 39 opens out of 244 recipients · 85 views

• 1 Sept: 14% open rate (I Checked My Last 40 Substack Subscriptions...) : 36 opens out of 255 recipients · 87 views

• 10 Sept: 14% open rate (Substack’s AI Detector Is Creating a Problem...) : 37 opens out of 266 recipients · 80 views

The pattern looks undeniable. It is not the whole story.

Look at the opens, not the percentage: 18 people opened on August 3. 48 opened on August 12. 37 opened on September 10. The number of people actually opening did not collapse. It grew, then settled around 35 to 40, while the list kept growing underneath it.

Most of that growth came through Notes while I was doing subscribe-for-subscribe: 162 of the 215 people who joined between August 1 and August 12 subscribed from a Note. People who clicked Subscribe, not people who read. The percentage fell because the denominator grew.

2. Why “0% Open Rate” Does Not Mean “Dead Reader”

When faced with a 14% open rate, the instinct is to export your CSV, filter by Activity = 0, and hit delete.

That is a mistake. Here is why:

1. Email pixels lie in both directions: Apple Mail Privacy Protection (MPP) pre-loads the tracking pixel, so it can count an open that never happened. Email clients that block images or show text only do the opposite: a reader can read every word without triggering a single open event.

2. The Substack App Factor: Readers who read your posts exclusively inside the Substack iOS/Android app or on the web do not always register as email openers.

3. Channel vs Exact Attribution: Just as 19 subscribers came through Notes without being attributed to a single Note, engagement often happens on profile surfaces rather than in the inbox.

Deleting people purely because an open-tracking pixel returned zero will silently wipe out some of your most attentive app-based readers.

When I exported my own list on September 15, the data proved it:

• 275 subscribers in total. 215 of them joined between August 1 and August 12, while I was still doing subscribe-for-subscribe.

• 154 of those 215 had never opened a single email.

• But 62 of those 154 still showed activity: 54 had viewed my posts on the web or in the app. A “0 opens” filter would have deleted them.

• 71 had never received an email at all (email turned off, app-only readers). “0 opens” says nothing about them.

• 23 had clicked a link inside my emails while Substack counted zero opens. Their pixel was blocked. They are readers.

A delete-by-0-opens cleanup would have removed at least 67 people who were demonstrably reading.

3. The List Cleaning Protocol I Actually Ran

Here is exactly what I did on September 15, and what happens next.

Step 1: Isolate the Silent Cohort

I kept only free subscribers who met every one of these conditions:

• Joined between August 1 and August 12, the subscribe-for-subscribe window. My publication is only six weeks old, so a “joined more than 60 days ago” rule would have matched nobody.

• Received at least one email (among subscribers with no other activity, this removed 36 app-only readers).

• Zero recorded email opens.

• Zero post views on the web or in the app.

• No activity and zero link clicks (this removed 5 more people who had clicked a link).

That left 51 subscribers out of 275. I did not filter on Notes activity: the export does not include likes, restacks or comments on Notes, so I check those names by hand before removing anyone.

Step 2: Send a “Stay or Leave” Re-engagement Post

Before removing anyone, I sent a short email to those 51 people only (a saved Substack segment) on September 15:

> Subject: Should I remove you from the Subshack Growth list?

> On August 12, I stopped doing subscribe-for-subscribe to focus only on high-value content.

> If you want to keep receiving my case studies and raw analyses, click the button below:

> [ Keep my free subscription → ]

> If you do nothing, your address will be removed in 7 days.

A P.S. also pointed to two free tools: the Subscribe-Back Checker and the profile audit.

Nothing happens automatically: Substack does not remove anyone for you. On September 22, I will export the list again and keep anyone who clicked, opened or read a post in the meantime. Then I check the remaining names against my Notes activity and remove the rest by hand. I will publish the results in a follow-up post after September 22: who stayed, who left, and whether this actually helped my open rate. Subscribe so you don’t miss it.

Two traps I hit while setting this up: a saved Substack segment built with an “email is any of” filter silently lost that filter and grew to my entire free list (273 people). Rebuild the cohort with activity filters, and re-open the segment before you send. And a post sent to a segment is still public in your archive.

4. Fix Your Landing Page First

Cleaning your list fixes your retention. But if your public Substack profile is leaky, new visitors will continue to bounce.

Whether or not you prune your list, run your profile through the free audit tool:

👉 Score My Substack Profile /100 →

It checks your positioning (35 pts), conversion hooks (25 pts), brand trust (15 pts), activity (15 pts), and SEO (10 pts) in 10 seconds without requiring a login.

What is your current Substack open rate? Have you ever purged inactive subscribers? Hit reply or drop a comment below. I read and answer every response.

Subscribe now

On August 12, I stopped doing subscribe-for-subscribe to focus only on high-value content.

If you want to keep receiving my case studies and raw analyses, click the button below:

Keep my free subscription →

If you do nothing, your address will be removed in 7 days.

— Subshack Growth by Kacoulib

P.S. Did you test sub-for-sub too? Check who actually subscribed back to you and who didn't (no account needed): 👉 Run the Subscribe-Back Checker →

P.P.S. Want to fix the leak on your public profile instead? Get your free score: 👉 Score my Substack profile /100 →

If you publish on Substack, you’ve probably noticed the noise: endless advice telling you to post 5 times a day, trade subscriptions, or chase viral vanity metrics.

Subshack Growth exists to filter out the noise.

I share only what actually moved my numbers the successful tests, and especially the ones that failed.

What You Get Here

1. Raw Data Playbooks: Real experiments with exact numbers on open rates, Notes attribution, and reader retention.

2. Zero-Fluff Autopsies: Honest breakdowns of platform mechanics (like why 19 subscribers came through Notes without a single one being attributed to a specific Note).

3. No Automation Fake Hacks: No auto-posting, no subscribe-for-subscribe, and no bot engagement.

4 Things to Do Right Now

1. Audit Your Substack Profile in 10 Seconds (Free)

Before changing your content strategy, fix your landing surface. The free Profile Audit checks your profile against 5 deterministic axes: Positioning (35 pts), Conversion (25 pts), Trust (15 pts), Activity (15 pts), and SEO (10 pts).

👉 Score My Substack Profile /100 → (No login or account needed)

2. Read the 3 Core Data Experiments

If you want to understand what actually drives growth on Substack, start with these three autopsies:

• Retention & Deliverability: My Open Rate Collapsed From 34% to 14%. Here Is the Exact Protocol I’m Using to Clean My Substack List.

• Notes Attribution: 19 Subscribers Came Through Notes. Not One Was Attributed to a Note.

• The Reciprocal Trap: Why Subscribe-for-Subscribe Is Not Really Win-Win

3. Leave a Comment and Introduce Yourself

I read and answer every comment. What is your Substack about, and what is your current biggest growth bottleneck?

4. Join the Weekly Chat Rituals

Tuesday Feedback Thread. Bring two options: a headline, a profile bio or a Note hook. Option A against Option B. You get feedback from me and from the room, and you leave with a decision instead of a doubt. Join this week’s thread →

Friday Wins. First one on Friday 25 September. Post the one number that moved this week and what you think moved it. Small wins count. Patterns show up faster when several of us line them up side by side.

Talk soon,

Kacoulib | Founder at Subshack

I wasn’t planning to write an article about AI detection.

Then I started posting about Substack’s new AI detector.

One of those Notes reached 134 impressions, 11 likes and 4 replies, with 69% discovery.

Another reached only 48 impressions but generated a new subscriber.

For my small publication, that was enough to tell me something.

Writers have opinions about this.

And I think the most interesting problem with AI detection isn’t actually AI.

It’s what happens to human writers once they know they’re being scored.

What Substack actually launched

In July, Substack introduced Scan for AI text, powered by Pangram.

Readers can scan eligible Posts, Notes, comments and replies and receive an estimate of how much of the text appears human-written or AI-assisted.

Creators can also scan their own drafts, report a detection error and disable detection on individual pieces of content.

Substack’s reason for doing this makes sense.

Trust matters.

If a reader thinks they’re paying to read someone’s ideas and instead gets hundreds of automatically generated articles, that changes the relationship.

Chris Best even acknowledged something important when Substack announced the feature:

The detector cannot tell whether great human care went into a piece. It also cannot tell whether AI was simply used as a source or research tool. And Substack explicitly says the detector is not perfect.

So far, fair enough.

But the feature introduces another question writers didn’t have to ask before.

“Is this good?” becomes “Does this look human?”

Imagine spending four hours writing something.

You research it.

Rewrite the introduction.

Delete paragraphs.

Change your mind halfway through.

Finally, you’re happy with it.

Then you run the detector.

82% AI.

What happens next?

You know you wrote it.

But your readers don’t.

Do you publish it anyway?

Do you add a statement defending yourself?

Do you disable detection?

Or do you start changing sentences until a machine agrees that you sound human?

That last option is the one I find fascinating.

Because suddenly the detector isn’t just observing writing.

It’s influencing writing.

This is already becoming a real dilemma

Today, I found a writer describing almost exactly this situation.

They said they wrote their first Substack feature entirely themselves, using Grammarly only for proofreading.

Substack’s detector reportedly classified the text as 100% AI-generated.

They then tested the same text with several other AI detectors, which reportedly classified it as human-written.

That is one person’s report.

It does not prove that Pangram has a widespread false-positive problem.

But the reaction is more interesting than the score itself.

The writer essentially asked:

Should I change writing I know I wrote because an AI detector thinks AI wrote it?

That is a completely new writing problem.

Even a highly accurate detector can change behavior

This is why I think the false-positive debate is only part of the story.

Imagine the detector becomes extremely accurate.

The behavioral problem still exists.

Once a score exists, people optimize around the score.

We already do this everywhere online.

We change headlines for click-through rates.

We change articles for Google.

We change thumbnails for YouTube.

We change posts for algorithms.

Now writers have another metric available to optimize:

How human does my writing look?

And that creates a bizarre feedback loop.

Write something.

Scan it.

Get an AI score.

Change your writing.

Scan it again.

Repeat until the machine thinks you sound sufficiently human.

At that point, who exactly are you writing for?

“100% human-written” could become a status symbol

I posted another thought about this recently.

I think we might eventually see writers putting things like:

100% human-written

in their bios, About pages or marketing.

Not because the writing is necessarily better.

But because authorship itself becomes part of the product.

For some publications, that could genuinely matter.

Readers may want essays, personal stories or journalism created directly by a person.

Other readers may not care whether AI helped with editing, research or structure.

That’s fine too.

The problem starts when a percentage becomes a shortcut for quality.

Because these are very different questions:

Was AI involved?

and

Is this worth reading?

A perfectly human article can be terrible.

An author can also use AI somewhere in their process while contributing original research, experience, judgment and ideas.

The percentage alone cannot answer that.

I think writers should optimize for trust, not the detector

I’m not going to tell writers to ignore AI transparency.

Quite the opposite.

If you use AI heavily, tell people.

If you don’t, say that too if it matters to your audience.

Substack even created a “How I make this” statement specifically so creators can explain their process.

But I wouldn’t rewrite a sentence I believe is good simply because a detector dislikes it.

And I definitely wouldn’t start deliberately making my writing worse, stranger or less polished to increase a “human” percentage.

Because then we arrive at the strangest possible outcome:

Humans changing how they write so that an AI believes they are human.

That seems like the exact opposite of what these tools are supposed to achieve.

The question I want to optimize for is still much simpler:

Was this worth reading?

Everything else comes second.

I’m curious how other writers are handling this.

Have you scanned your own writing yet?

And if Substack told you something you wrote yourself was mostly AI-generated, would you change it?

I’m documenting my Substack growth experiments and the data behind them while building Subshack.

Next experiment: I currently have 277 subscribers, but only 61 opened an email in the last 30 days. I want to understand what a “subscriber count” actually tells us.

Subscribe now

My open rate was falling—from 34% to 18%.

A subscriber is not always a reader. Some people find your work and genuinely want more. Others arrive through a subscribe-for-subscribe thread, ask you to return the favor, and eventually drift away.

Both increase your subscriber count. But they do not build the same audience.

The question Substack makes difficult to answer

You’ve probably seen this message:

“Subscribed ! Your turn.”

You subscribe back. A few days later, you have no simple way to check whether that account is still subscribed to you. Once is easy. Forty times is not.

So I built the Substack Subscribe-Back Checker. It reads a public profile’s subscription list and sorts every publication into three groups:

• Mutual — both sides are subscribed today.

• One-way — you subscribe to them; no subscription back is publicly visible.

• Publicly unverifiable — they keep their subscriptions private, so nobody can check either way.

I ran it on my own list first

I subscribe to 488 publications. The free scan reads the 40 most recent. The ones where a subscribe-back would still be recent enough to matter.

7 mutual. 19 one-way. 14 publicly unverifiable.

Seven. I had expected something closer to half.

That third number matters as much as the first. A private list stays private, and Subshack never opens it. Unverifiable means unverifiable not “they didn’t.”

A one-way subscription is not a betrayal

I subscribe to plenty of writers who will never subscribe to me. That is normal. I read them.

The better question is the other one: how many newsletters sit in your inbox only because you once felt obliged to return a subscription?

The checker never accuses anyone. It describes a relationship as it looks today. Someone may have used another account, or unsubscribed later. That is why a one-way result is a lead to review, never a verdict.

Check your own list

Before you scan, guess your number. Under 20%? Around 40%? More than 60%? Then run it and compare.

Check who subscribed back — free

No account, nothing to install. The free scan covers your 40 most recent subscriptions and gives you every total. To cover all of them — mine were 488 — the Subshack Chrome extension runs the same check from your own browser, under Audit › Sub-for-sub.

A bigger number feels like progress. Knowing who is actually behind it is more useful.

For seven days, I stopped doing the thing that had grown my Substack the fastest.

No more:

• “Subscribe and I’ll subscribe back.”

• “Follow me and I’ll follow you.”

• Transactional support Notes.

• Subscribing to publications only because they subscribed to mine.

I expected my subscriber growth to slow down.

What I did not expect was this:

I published 54% more Notes, but my total reach fell by 28%.

And the average measurable reach of each Note was cut roughly in half.

That result was uncomfortable.

It was also exactly what I needed to see.

Why I stopped a strategy that was clearly working

When I started this Substack, I wanted to understand how a completely new publication could grow from zero.

Subscribe-for-subscribe worked.

At least, it worked according to the most visible metric: subscriber count.

Two reciprocal-support Notes alone brought me:

• 867 combined impressions

• 55 combined replies

• 63 subscribers directly attributed by Substack

One Note generated 50 attributed subscribers.

The other generated 13.

For a new account, those numbers were difficult to ignore.

My subscriber count grew quickly. I eventually reached 253 subscribers after publishing 130 Notes and 12 articles.

But another problem appeared.

I could see how many people had subscribed.

I could not tell how many had subscribed because they genuinely wanted to read my work.

Would they open my next article?

Would they return?

Would they interact when there was no subscription expected in exchange?

Was I building an audience, or simply increasing a number?

The only way to find out was to remove the tactic responsible for the fastest growth.

The rules of the experiment

For the following period, I stopped all explicit reciprocal-growth activity.

What I stopped

I did not publish any subscribe-for-subscribe Notes.

I did not promise to follow people back.

I did not subscribe to publications simply because their writers had subscribed to mine.

I did not ask people to comment “done” so that I could return the action.

What I continued

I kept publishing original Notes.

I replied to other creators when I had something relevant to add.

I published articles.

I participated in conversations.

I promoted my work selectively.

I continued improving my profile, content and tools.

The goal was not to stop growing.

The goal was to see what remained when the transactional activity disappeared.

What happened

The following comparison comes from my latest seven-day analytics window, synced through August 23, 2026.

• Notes published — 13 → 20 (+54%)

• Total impressions — 587 → 420 (–28%)

• Likes — 62 → 54 (–13%)

• Replies — 5 → 5 (no change)

• Restacks — 3 → 1 (–67%)

• Discovery share — 40% → 36% (–4 points)

• Attributed subscribers — 0 → 0 (no change)

One of the 20 recent Notes did not have impression data available, so it should not be treated as receiving zero impressions.

Across the 19 Notes with measurable data, the average was approximately 22 impressions per Note.

During the previous period, the average was approximately 45 impressions per Note.

That means the measurable reach per Note fell by roughly half.

I created more content.

I generated less total distribution.

And each individual Note received significantly less attention.

Publishing more did not solve the problem

The obvious assumption was that more Notes would create more opportunities for discovery.

That did not happen.

I published seven additional Notes, but received 167 fewer impressions.

This does not necessarily mean that publishing more directly caused the decline.

This was not a controlled scientific experiment. Timing, topics, audience activity and Substack distribution could all have influenced the results.

But it does show something important:

Publishing more content does not guarantee more distribution.

My problem was not a lack of output.

I had already published 130 Notes.

The problem was that most of those Notes did not have a strong distribution mechanism behind them.

The reciprocal Notes had one.

They gave readers a direct reason to respond:

Take this action, and I will return it.

Original Notes have to compete differently.

They need to be interesting enough to stop someone in the feed.

Useful enough to make them remember you.

Strong enough to earn a restack.

Relevant enough to attract the right reader.

That is much harder than offering an immediate transaction.

Zero attributed subscribers does not necessarily mean zero impact

Substack attributed no new subscribers directly to Notes in either seven-day period.

That does not prove that none of the Notes influenced a subscription.

Someone can:

1. See a Note.

2. Visit the profile later.

3. Read an article.

4. Return through another source.

5. Subscribe without the original Note receiving credit.

Attribution is incomplete.

That is why I do not want to claim: “My original content generated zero subscribers.”

The more accurate conclusion is:

Substack directly attributed zero subscribers to these Notes during the measured period.

That distinction matters.

It is easy to turn incomplete attribution into a confident but false growth lesson.

I have already made that mistake before.

The real benefit of stopping subscribe-for-subscribe

Stopping reciprocal growth did not immediately improve my numbers.

It improved the quality of the information behind them.

When transactional Notes dominated my dashboard, the obvious recommendation was: publish more reciprocal Notes, because they generate the most subscribers.

That recommendation was technically correct.

It was also strategically dangerous.

If I repeated it indefinitely, I could continue growing the subscriber count while making it increasingly difficult to understand whether anyone actually wanted the content.

Removing the tactic exposed the baseline.

It showed me what my publication could generate without an explicit exchange attached.

The baseline was weaker than I wanted.

But weak, honest data is more useful than impressive data I cannot interpret.

The experiment exposed my actual growth problem

I initially thought the question was: can original content replace subscribe-for-subscribe?

That was not the most useful question.

The real question was: can I build a repeatable distribution system for original content?

Right now, the answer is not yet.

Creating more Notes is not the system.

Publishing another article is not the system.

Posting a link repeatedly is not the system.

A real distribution system needs to answer:

• Which topics attract the right people?

• Which Notes generate profile visits?

• Which readers return more than once?

• Which creators amplify the content?

• Which interactions lead to subscriptions?

• Which subscribers become active readers?

• Which growth is organic, community-driven or reciprocal?

Substack gives me the totals.

I need to understand the mechanism behind them.

Why I started building Subshack

This experiment is also the reason I started building Subshack.

I did not need another dashboard telling me that one Note had more impressions than another.

I needed to know why.

Was the Note successful because the idea was strong?

Because another creator amplified it?

Because it triggered reciprocal activity?

Because the same people repeatedly interacted with my content?

Or because it genuinely reached new readers?

Subshack now helps me separate those signals and identify:

• Organic and reciprocal activity

• Repeat engagers

• Potential relationships

• People amplifying my work

• Subscribers becoming inactive

• Content generating reach without conversion

• Content generating conversion without obvious engagement

The goal is not to automate relationships.

The goal is to stop treating every impression, like and subscriber as if they meant the same thing.

They do not.

What I am changing now

I am not returning to mass subscribe-for-subscribe.

I will still subscribe to publications I genuinely want to read.

I will still support creators whose work interests me.

But that is a relationship, not a transaction.

For the next phase, I am changing five things.

1. Fewer repetitive Notes

Publishing more did not increase total reach.

I will focus on fewer Notes with stronger hooks, clearer opinions and more useful evidence.

2. More distribution around each idea

Instead of publishing an idea once and immediately replacing it, I will reuse strong ideas through different formats:

• A data point

• A contrarian observation

• A practical lesson

• A question

• A personal example

3. More targeted interaction

I will stop interacting randomly.

I want to focus on people who are relevant to my topic, repeatedly engage with similar work and are likely to become genuine readers.

4. Separate acquisition from audience quality

Subscriber growth will remain useful.

But I will also measure:

• Returning engagement

• Article opens

• Repeat interactions

• Restacks

• Replies

• Activity after subscribing

5. Judge content by its job

Not every piece of content needs to generate subscribers.

Some content creates discovery.

Some builds trust.

Some starts conversations.

Some converts.

The mistake is judging all of them with a single number.

The next experiment

My next test will be more targeted.

For seven days, I will use Subshack to identify the people most relevant to my publication.

Not simply the people most likely to follow back.

I will focus on:

• Repeat engagers

• Relevant creators

• People already interacting with Substack-growth content

• Readers who have shown interest but have not subscribed

• Existing subscribers who appear likely to become active readers

I will engage manually and compare the results with my previous random interaction.

Then I will measure:

• Profile visits

• Replies

• New relationships

• Attributed subscriptions

• Returning readers

• Time spent per result

That should answer a more valuable question: can targeted relationships outperform transactional growth?

The conclusion

Subscribe-for-subscribe gave me something every new creator wants: momentum.

It helped me leave zero.

It introduced me to other writers.

And two Notes generated 63 directly attributed subscribers.

Pretending it did not work would be dishonest.

But the experiment also revealed its limit.

Once I removed the transaction, publishing more content did not preserve the same distribution.

I published 54% more Notes.

My impressions fell by 28%.

My measurable reach per Note was roughly cut in half.

That does not mean I should return to the old strategy.

It means I now know what needs to be built next.

Not a faster way to increase the subscriber count.

A better system for reaching the right people, turning attention into genuine interest and identifying who actually comes back.

That is a much harder problem.

It is also a much more valuable one.

Want to see which parts of your own Substack growth are actually working?

Try Subshack — the Chrome extension for understanding your Notes, audience and growth.

Subscribe now

The Truth About My Substack Growth — Part 4 of 5

Data snapshot: 23 August 2026, 16:43 UTC — 130 Notes, 253 current subscribers.

The cleanest comparison in my Substack experiment happened by accident.

On 12 August, I stopped using explicit subscribe-for-subscribe Notes. By my latest synced snapshot I had published exactly 65 Notes before that cutoff and 65 after it — two groups almost perfectly matched in size.

The result looked brutal.

My first 65 Notes were credited with 69 subscribers. Among the next 65, Substack returned usable attribution statistics for 64 — and those 64 were credited with zero.

If I had stopped at the Note-level dashboard, the conclusion would have been obvious:

I removed the tactic that worked, and my Notes stopped generating subscribers.

Then I opened the subscriber-level data. 29 current subscribers had joined after the cutoff, and 19 of them were marked as having come through Substack Notes.

Zero attributed to an exact Note. Nineteen acquired through the Notes channel. Both numbers are correct — I had been asking two different questions as if they were the same one.

The 65-versus-65 comparison

Here is the complete split from the synced data:

The shift was not limited to subscriber attribution. Impressions fell by roughly 54%, replies from 62 to 14, likes more moderately from 207 to 170. Restacks moved the other way, from 2 to 6, though the numbers are too small to treat as a stable pattern.

The reciprocal Notes in the first group gave people an immediate reason to act. Remove that incentive and the visible activity becomes much less dramatic.

But “less dramatic” is not the same as “zero growth”. That distinction was hidden one level deeper.

Exact-content attribution is not channel attribution

The Note-level metric answers a narrow question:

Which exact Note received credit for this subscription?

The subscriber record answers a broader one:

Through which Substack surface did this subscriber arrive?

After the cutoff, the 29 subscribers still on the list at the time of the sync broke down like this:

• 19 — Substack Notes

• 4 — reader profile discovery

• 3 — Substack app

• 1 — direct-to-app

• 2 — unknown

So roughly 66% of that cohort was associated with the Notes channel.

That does not identify the Note that persuaded each person, or prove that one of my own Notes was the decisive touchpoint. Someone could have found me through a reply, visited my profile after several interactions, or subscribed after seeing me repeatedly across the Notes ecosystem. The available data cannot isolate those paths.

The Note-level zero was not fake either: for the 64 post-cutoff Notes with returned attribution data, no individual Note received subscriber credit. The mistake would be translating that into:

Notes produced no subscribers.

That goes beyond what the metric can support. It is the same problem my classifier exposed in Part 3: the raw number is correct while the lesson extracted from it is wrong. There the missing context was intent; here it is grain — one table measures an exact piece of content, the other measures the wider acquisition surface.

The fix is not to pick whichever number makes the better story. It is to keep both, and label them correctly.

What the comparison actually supports

1. Removing the reciprocal incentive reduced visible activity. The second group received fewer impressions and far fewer replies — consistent with reciprocal Notes being unusually effective at creating immediate action. It is not a controlled causal experiment: topics, timing, audience size and content mix also changed.

2. Exact Note attribution disappeared. No post-cutoff Note with returned attribution data received direct subscriber credit. That is a real weakness in the second group.

3. The Notes channel still participated in acquisition. Nineteen current subscribers who joined after the cutoff were associated with Substack Notes at the channel level. The data cannot identify the decisive interaction.

4. Acquisition is still not the final outcome. What matters is what those people did after joining — whether they received the next article, opened it, came back, replied or restacked without being offered anything in exchange. Part 2 showed how wide that gap can get: my subscriber count grew while my open rate fell. That is the measurement layer I still need to improve.

The experiment I am redesigning

My first growth experiment optimised for speed — I described that version in Part 1. The next one needs to preserve the full path:

• Distribution — how many people saw the Note, and how many were not already connected to me?

• Channel acquisition — did the subscriber arrive through Notes, the app, profile discovery, search or another surface?

• Exact-content attribution — when Substack can connect a subscription to a specific Note or article, which content receives credit? When it cannot, the answer should remain unknown, not silently become zero.

• Reader behaviour — after joining, did they receive an email, open an article, return and interact again?

• Editorial signal — which useful ideas generated replies, restacks and repeat attention without depending on a reciprocal promise?

I still want to publish useful Notes, write substantive replies and produce deeper articles. I also want to test one external community at a time instead of dropping links everywhere and calling that distribution.

But subscriber speed is no longer the score that decides whether the experiment worked. The system needs to distinguish:

Who arrived, where they arrived from, what received credit — and whether they came back.

What I am changing in Subshack

I am adding this distinction to Subshack. A zero in the content table should not silently become a zero in the growth story.

The product should show exact attribution when it exists, channel-level attribution when that is all the data supports, unavailable data as unavailable, and post-subscription behaviour separately.

No invented precision. No automatic causal story. Just the strongest conclusion the available data can actually support.

Install Subshack — Free on Chrome

If you want to see the same distinction on your own account before I ship it, run the free Substack profile audit — it reads your public profile and scores it without asking for a login.

The uncomfortable conclusion

My first 65 Notes told a simple story: attention, replies and attributed subscribers, quickly. My next 65 looked much worse in the Note-level dashboard.

But the subscriber records showed that growth had not stopped. It had become harder to attribute.

That is less satisfying than a clean before-and-after victory. It is also more useful. Because the real question is no longer whether Notes generated zero subscribers or nineteen. It is:

Which exact content can be credited, which channel participated, and did those subscribers become readers?

Until a dashboard separates those questions, the cleanest number may produce the weakest lesson.

Have you ever seen two Substack screens appear to tell different growth stories? Which number did you trust?

Tell me in the comments — especially if you have hit a similar attribution gap.

Next in Part 5: I compare what happened when I published on X and Bluesky, contributed on Reddit, and waited for Google — because publishing everywhere is not the same as building distribution.

Method. Subshack snapshot of 23 August 2026, 16:43 UTC. The post-cutoff cohort counts people who joined on or after 12 August and were still subscribed at the sync. One of the 65 post-cutoff Notes returned no impression or attribution data. Channel source does not identify the exact content that caused a subscription. Observational, not a controlled experiment.

The first version of my growth classifier made a reasonable mistake.

It labelled my own article about subscribe-for-subscribe as reciprocal content.

The article was not asking anyone to subscribe back.

It was analysing why reciprocal growth had distorted my early results.

But the classifier saw the language inside it:

• subscribe-for-subscribe

• follow-for-follow

• reciprocal support

• subscribing back

It recognised the subject.

It completely missed the intent.

That false positive exposed a larger problem in the way growth tools interpret performance.

A metric can be correct while the lesson drawn from it is wrong.

The problem I was actually trying to solve

In Part 1, I analysed 105 Notes.

Two networking Notes accounted for 56 of the 62 subscribers Substack attributed to individual Notes.

One was explicit:

Follow = Follow
Subscribe = Subscribe
Let’s start.

The other told writers I was reading, subscribing and supporting them, then invited them to connect.

The results were real.

The first Note recorded 479 impressions, 39 replies and 44 attributed subscribers.

The second recorded 355 impressions, 14 replies and 12 attributed subscribers.

But neither result proved that readers wanted my next article.

Then, in Part 2, I found that my subscriber count continued rising while later article open rates clustered between 18% and 23%.

That still did not prove that reciprocal subscribers caused the decline.

It showed something narrower—and more useful:

Subscriber acquisition and reader attention were no longer telling the same story.

I’m building this classifier into Subshack, the Chrome extension I’m creating for Substack writers.

So I wanted Subshack to answer a harder question than:

Which Notes produced the biggest numbers?

I wanted it to ask:

What mechanism produced those numbers—and should that mechanism influence my editorial recommendations?

The first version failed for an important reason

A basic classifier can look for obvious exchange language:

• “subscribe and I’ll subscribe back”

• “follow for follow”

• “drop your Substack”

• “support me and I’ll support you”

Those patterns are useful.

They are also insufficient.

An article criticising subscribe-for-subscribe may contain exactly the same words as a Note promoting it.

One describes the mechanism.

The other activates it.

My first classifier treated both as equivalent.

That meant it could make two opposite mistakes:

1. Exclude genuinely editorial analysis because it mentioned reciprocal growth.

2. Treat reciprocal performance as evidence that the underlying topic created genuine demand.

The problem was not the raw data.

The problem was allowing a weak interpretation of that data to control the recommendations.

I stopped treating classification as a moral judgement

“Real growth” and “fake growth” make dramatic labels.

They are not precise enough for an analytics product.

A reciprocal subscriber is still a person.

A community interaction can become a genuine relationship.

A subscriber acquired through an editorial Note may still never open another article.

So the purpose of classification is not to declare one person valuable and another worthless.

It is to separate how an outcome happened from the outcome itself.

I now use four categories:

Organic content

The Note primarily shares an idea, observation, story, argument or useful resource without requesting an exchange.

Community or networking content

The Note is designed to meet people, participate in a creator community or start a conversation, but it does not explicitly promise reciprocal action.

Reciprocal content

The Note contains an explicit or strongly implied exchange: subscribe back, follow back, support in return or another action tied to reciprocity.

Unknown

The available evidence is not strong enough to classify the mechanism honestly.

Unknown is not a failure.

It is better than forcing certainty into ambiguous data.

How I changed the classifier

I did not solve the problem by asking an AI model to guess harder.

I changed the order of decisions.

1. Explicit exchange rules come first

Clear promises such as “subscribe and I’ll subscribe back” can be classified deterministically.

The same input should produce the same result every time.

2. Analytical context acts as a safeguard

If the content is reviewing, criticising or reporting on reciprocal growth, the presence of those keywords is not enough to classify it as reciprocal.

The system must distinguish participation from analysis.

3. Ambiguous content stays uncertain

Semantic classification is used only when the deterministic evidence is incomplete.

If the signals still conflict, the result remains unknown instead of being forced into the nearest category.

4. Manual correction always wins

No automated system understands every context.

If I correct a classification, that override persists. The same Note should not be misclassified again during the next analysis.

5. Recommendations require enough organic evidence

Subshack does not make a strong editorial recommendation from one lucky Note.

It requires at least five organic Notes before presenting a pattern as something worth repeating.

Below that threshold, the honest answer is:

There is not enough evidence yet.

The rule that changed the product

The simplest version of the rule is:

Raw truth stays intact. Recommendations get filtered.

Subshack never deletes the impressions, replies, subscribers or other activity produced by reciprocal Notes.

Doing that would rewrite history.

Instead, the dashboard shows two different views:

• All activity answers what happened.

• Organic analysis helps decide what editorial content to repeat.

In the later dashboard snapshot, 106 Notes had synced—one more than the original 105-Note audit.

The classifier separated them into:

• 92 organic Notes

• 8 community or networking Notes

• 6 reciprocal Notes

The 14 community and reciprocal Notes still remained in the raw analytics.

They were simply excluded from recommendations about what editorial content worked.

That distinction produced a less flattering but more useful organic view:

• 1.3 subscribers per 1,000 impressions

• 4 subscribers gained

• 3,072 impressions

• 45% discovery rate

Those numbers do not erase my early growth.

They stop two exceptional reciprocal Notes from teaching the software that I should publish more reciprocal Notes.

What the classifier still cannot know

Classification adds context.

It does not create perfect attribution.

Subshack still cannot know with certainty:

• why every person subscribed

• whether a community interaction was sincere

• whether an organic subscriber will become a regular reader

• whether a reciprocal connection will eventually become genuine

• how much algorithmic distribution affected a Note

The categories describe the most likely mechanism visible in the content.

They do not read people’s minds.

That limitation matters because clean labels can create false confidence.

The purpose of the system is not to remove uncertainty.

It is to stop obvious distortions from silently becoming advice.

A small classification test

Consider these three Notes:

“Subscribe and I’ll subscribe back. Drop your link below.”

“I tested subscribe-for-subscribe for seven days. Here is what happened to my open rate.”

“Looking for other SaaS writers to exchange feedback and learn from.”

The first is reciprocal.

The second is editorial analysis.

The third is harder.

It may be community networking.

It may become reciprocal depending on what “exchange” means in practice.

That is exactly why I need an unknown state, reviewable evidence and manual correction—not just a confident label.

How would you classify the third Note?

I would genuinely like to see where other writers draw the line.

The uncomfortable product lesson

I originally built analytics to find what worked so I could repeat it.

That sounds sensible.

But “repeat what worked” becomes dangerous when the software cannot distinguish:

• attention from intent

• transactions from amplification

• community participation from editorial demand

• acquiring a subscriber from creating a reader

The biggest number is not automatically the strongest signal.

Sometimes it is simply the easiest mechanism to activate.

My first classifier made the same mistake I had made personally.

It saw the surface result and moved too quickly toward a conclusion.

Correcting that mistake changed Subshack from a performance dashboard into something more useful:

a system that preserves what happened while remaining sceptical about what it means.

If your analytics showed that your best-performing content was driven by reciprocity, would you still want it recommending more of the same?

I’m building this classification system into Subshack. If you want to test it as I build it, join the Founding Beta.

Next in Part 4: I tested four distribution channels outside Substack. Together, they produced one subscriber.

Subscriber count told me I was growing.

Open rate told me to slow down.

Neither number told the whole truth.

I went from zero to more than 150 subscribers during my first week on Substack.

By the time of this audit, the publication had reached 241 subscribers.

From the outside, the direction was obvious:

More subscribers meant more growth.

But after auditing the Notes responsible for that growth, I looked at what happened after people subscribed.

The result was less flattering.

On August 3, one of my articles recorded a 34% open rate.

Ten days later, my newest article recorded 18%.

That is a decline of 16 percentage points—or roughly 47% relative to the earlier rate.

My list was getting bigger.

The percentage of it opening my new articles was getting smaller.

The pattern across my first nine articles

The decline did not happen in a perfectly straight line.

These were the recorded open rates:

The 34% result was the high point, not the permanent baseline.

But the later pattern is still difficult to ignore.

Every article published after August 5 recorded an open rate between 18% and 21%.

Meanwhile, the subscriber count continued rising.

The list was expanding faster than reader attention.

A falling open rate does not automatically mean fewer readers

This distinction matters.

A percentage can fall while the absolute number of readers still rises.

For example:

• 100 subscribers at a 34% open rate means 34 opens.

• 200 subscribers at an 18% open rate means 36 opens.

The second list has a much weaker rate but slightly more total readers.

So I cannot honestly say that my audience disappeared simply because the percentage declined.

A larger list can still create more reach.

The problem is efficiency.

I needed almost twice as many subscribers to produce roughly the same number of opens.

That means subscriber growth and audience growth are not always the same thing.

One measures how many people joined.

The other measures how many people continue paying attention.

The explanation I cannot prove—but cannot ignore

The most obvious explanation is acquisition quality.

In Part 1, I found that the two strongest networking Notes accounted for 56 of the 62 subscribers Substack attributed to individual Notes.

One explicitly offered reciprocal support.

The other told writers I was reading, subscribing and supporting them.

Both generated subscribers quickly.

But someone who subscribes because they expect support in return is not necessarily subscribing because they want my next article.

When the next email arrives, the original reason for subscribing matters.

Did they join because they wanted:

• Substack growth experiments?

• data and practical tactics?

• updates about building Subshack?

• or simply another subscriber in return?

Those motivations all produce the same +1 in the subscriber count.

They do not produce the same reading behaviour afterward.

The falling open rate is consistent with audience dilution.

But it does not prove that reciprocal growth caused it.

What the data cannot tell me

This audit covers only nine articles from a publication that was a few weeks old.

The articles also had different:

• titles

• subjects

• publication dates

• levels of relevance

• amounts of promotion

The 34% article discussed why subscribe-for-subscribe was not really win-win.

The 18% article documented my first Substack Chat experiment.

The difference in subject alone may explain part of the gap.

I also do not have a clean cohort analysis showing:

• which acquisition source produced each reader

• whether reciprocal subscribers opened less often

• whether organically acquired subscribers returned more frequently

• how engagement changed for the same subscribers over time

So the accurate conclusion is not:

Reciprocal subscribers caused my open rate to fall.

It is:

My subscriber count increased while the open rates on later articles remained materially below my early high points. Reciprocal acquisition is one plausible explanation, but the available data cannot isolate the cause.

That sentence is less dramatic.

It is also more honest.

The subscriber count hides different types of growth

Before this audit, I treated every new subscriber as an identical unit of growth.

Now I see at least three different possibilities.

Transactional growth

Someone subscribes because I subscribed, replied or offered support first.

Community growth

Someone subscribes because we interacted, share the same experience or belong to the same creator network.

Editorial growth

Someone subscribes because the content itself created enough interest for them to want the next piece.

All three subscribers are real people.

All three may eventually become genuine readers.

But their original motivation is different.

A dashboard that combines them into one number cannot explain why the list grows while the open rate weakens.

What I should have measured from the beginning

Subscriber count is still useful.

It just needs context.

I should have tracked it alongside:

• article open rate

• total views

• comments and replies

• restacks

• returning readers

• unsubscribes after each publication

• the mechanism that produced each subscriber

The most important question is no longer:

How many subscribers did this tactic produce?

It is:

What did those subscribers do after joining?

Did they open the next article?

Did they read a second one?

Did they reply without expecting anything in return?

Did they share the work with their own audience?

That behaviour is much closer to audience growth than the initial subscription.

What I am changing now

I am not deleting reciprocal subscribers or pretending the early growth never happened.

The raw numbers remain part of the story.

But I am changing what I allow those numbers to teach me.

A networking Note can be excellent for meeting creators.

It should not automatically become evidence that the subject of the Note created editorial demand.

A reciprocal subscription can begin a real relationship.

It should not automatically carry the same weight as someone who subscribed after reading an article.

And a growing list should not be celebrated without checking whether people continue opening the work.

For the next articles, I am watching both sides:

• acquisition: who subscribes and why

• retention: who returns and continues reading

Growth needs both.

The uncomfortable conclusion

My publication reached 241 subscribers quickly.

That achievement is real.

But the open-rate pattern suggests that the list may have grown faster than the audience’s interest in my work.

I did not necessarily build a smaller audience.

I may have built a larger audience with weaker average intent.

That is not the same failure.

But it requires a different strategy.

The goal is no longer to make the subscriber count rise as quickly as possible.

The goal is to make the next subscriber more likely to become a reader.

That starts before the subscription, on the profile they land on. I built a free Substack profile audit that scores what a first-time visitor sees, out of 100.

If your subscriber count doubled while your open rate nearly halved, would you still call it growth?

Would you prefer a smaller, highly engaged list—or a larger list that produces more total reach but weaker average engagement?

Tell me which one you would choose in the comments.

Next in Part 3: How I taught Subshack to distinguish organic, community-driven and reciprocal growth—after the first version classified my own article as reciprocal content.

Subscribe now

Two weeks ago, I thought I had found a fast way to grow a new Substack.

I started from zero, used Notes heavily, joined conversations, and hit my first 100 subscribers in one week, before crossing 150 shortly afterward

The number looked good.

The method looked repeatable.

So I began sharing what I was learning.

Then I built a tool to analyse my own growth.

The first thing it showed me was that the story I had been telling was incomplete.

I analysed 105 of my Notes.

This audit covers 105 Notes published between July 28 and August 17, 2026.

Only two of them were responsible for 90% of the subscribers I could attribute to individual Notes.

And those two Notes were the exact kind of content I had publicly decided to stop writing.

The number I couldn’t ignore

The first Note was direct:

Follow = Follow
Subscribe = Subscribe
Let’s start.

It generated:

• 479 impressions

• 39 replies

• 44 attributed subscribers

The second was less explicit:

To everyone publishing on Substack — I’m reading. I’m subscribing. I’m supporting you. Let’s connect.

It generated:

• 355 impressions

• 14 replies

• 12 attributed subscribers

Together, these two Notes represented only 1.9% of the 105 Notes in my dataset.

But among the six Notes for which subscriber attribution was recorded, they accounted for 56 of the 62 attributed subscribers.

That is 90%.

The other four Notes carrying attribution data accounted for the remaining six subscribers.

For the other 99 Notes, attribution was unavailable—not zero.

The obvious recommendation from the data would be simple:

Write more networking Notes.

But that recommendation would optimise the number while ignoring the reason behind it.

The tactic didn’t fail. It worked too well.

Subscribe-for-subscribe is often criticised because it creates low-quality subscribers.

That description is too comfortable.

My data suggests a more dangerous problem:

Reciprocal growth can outperform editorial content so dramatically that it becomes difficult to see what your audience genuinely values.

Here is what my data showed:

12.8 times more impressions and 189 times more replies per Note, every normal analytics dashboard will tell you to repeat it.

The metric is not fake.

The activity happened.

The subscribers are real people.

But the recommendation can still be wrong.

Those people may have subscribed because they wanted support in return—not because they wanted to read my next article about Substack growth.

That distinction does not appear in the subscriber count.

The signal that was missing

The two strongest acquisition Notes generated zero restacks.

That matters.

A like is cheap.

A reply can be part of an exchange.

A subscription can be reciprocal.

A restack is different.

When someone restacks a Note, they put it in front of their own audience.

They attach a small part of their credibility to it.

The two Notes that generated most of my attributed subscribers were successful at activating people.

But nobody chose to redistribute them.

They created transactions, not amplification.

That does not make the Notes worthless.

One of them helped me meet other new writers, and some of those relationships may become genuine.

But it changes what I can honestly claim they proved.

They proved that reciprocal and community-driven Notes can grow a subscriber count quickly.

They did not prove that the underlying content created demand.

The limitation I need to state clearly

My attribution data is incomplete.

Subscriber attribution was recorded for only six of the 105 Notes in this audit.

That means I cannot honestly claim that these two Notes generated 90% of every subscriber I acquired.

The accurate claim is narrower:

Among the six Notes for which subscriber attribution was recorded, the two strongest accounted for 56 of the 62 attributed subscribers—90%.

Small datasets make dramatic headlines easy.

They also make methodological honesty essential.

At the time of this audit, I had:

• 241 subscribers

• 105 analysed Notes

• 9 published articles

• a publication that is only a few weeks old

This is not a universal study of Substack.

It is an audit of my own behaviour.

What this audit changed

I no longer want my analytics to answer only this question:

Which Notes produced the biggest numbers?

I need them to answer a harder one:

Which Notes produced behaviour that is consistent with genuine interest?

That means separating raw performance from the mechanism that produced it.

The raw numbers must remain untouched.

Hiding reciprocal subscribers would be dishonest.

But recommendations should not automatically treat these three things as the same signal:

• a reciprocal interaction

• a community conversation

• an editorial response

They may produce similar numbers for completely different reasons.

That is what I am now building into Subshack.

The uncomfortable conclusion

My early growth was not imaginary.

The subscribers were real. So were the replies.

But I still do not know how much of that growth represented genuine demand for my work.

And if I want to give other writers growth advice, that is the first question I need to answer.

I built a tool to analyse my Substack growth.

One part of it is already free and public: a Substack profile audit that scores what a first-time visitor sees on your profile, out of 100.

The first thing it told me was that my growth story did not survive contact with the data.

I’m publishing the numbers anyway.

If you removed your two best-performing Notes, would your growth story still hold?

Mine didn’t.

Have you ever used reciprocal growth? Did those subscribers continue reading afterward?

Tell me what happened in the comments.

Next in Part 2: My subscriber count kept growing while my article open rate fell from 34% to 18%. I’ll show what happened to engagement while the list got bigger.

Subscribe now

The thread generated 34 replies.

Here’s exactly how I started it—and the unexpected way Substack promoted it in the Notes feed.

Why I created a Chat

During my first week on Substack, subscribe-for-subscribe was my main acquisition strategy.

It helped me grow from zero to more than 150 subscribers quickly.

This week, I stopped completely.

No reciprocal subscriptions.

No “subscribe and I’ll subscribe back.”

I wanted to discover whether I could continue growing by publishing useful content, writing meaningful replies, and building relationships.

My early observation is that organic growth becomes much harder when your publication doesn’t have a community around it.

Publishing brings people to your work.

Community gives them a reason to return.

The exact message I published

I created a Chat called Subshack Growth by Kacoulib and used this welcome message:

👋 Welcome everyone!

I created this chat as a place to talk about growing on Substack—what’s working, what isn’t, and the experiments we’re testing.

Feel free to share your publication, ask questions, or talk about a growth problem you’re currently facing.

To get things started:

What’s the #1 thing you’re struggling with on Substack right now?

I’ll try to help whenever I can—and I’ll also share what I learn from my own experiments here. 🚀

The thread received four likes and 34 replies.

The likes weren’t the most interesting result.

The replies were.

Why this question generated conversations

I think the message worked for four reasons.

1. The purpose was immediately clear

The Chat wasn’t presented as a general networking group.

It had one specific purpose: helping people discuss and experiment with Substack growth.

People could quickly decide whether the conversation was relevant to them.

2. People knew what they were allowed to share

I explicitly invited members to:

• Share their publication

• Ask questions

• Describe a growth problem

• Discuss their own experiments

They didn’t have to guess what was acceptable.

3. I asked only one question

I didn’t ask people to introduce themselves, describe their publication, share their goals, and explain their entire strategy.

I asked:

What’s the #1 thing you’re struggling with on Substack right now?

It was specific, relevant, and easy to answer.

4. The replies created more replies

Once people started sharing their problems, the thread no longer looked empty.

Each response gave someone else an example of what they could discuss.

The first participants helped make the conversation easier for everyone who arrived afterward.

Substack promoted the Chat automatically

After publishing the thread, I noticed something unexpected.

Substack displayed a native card in the Notes feed with the label:

From the Subshack Growth by Kacoulib chat

The card showed part of my message, the participants, and the number of replies.

Naturally, I tried to reproduce it myself.

I copied the thread link and shared it inside a regular Note.

It didn’t work. 😂

The Note contained the same destination, but Substack didn’t display it as the same native Chat card.

This appears to be an internal distribution mechanism that creators cannot manually trigger by simply sharing a link.

I don’t know exactly when Substack decides to show these cards or who receives them, so I’m not calling it an algorithm hack.

It’s only an observation from my experiment.

But it shows that Substack Chat may provide an additional native way to reach your existing audience inside the Notes feed.

Subscribers are not the same as a community

Subscribe-for-subscribe can increase a subscriber count.

But it doesn’t automatically create an audience that remembers you, reads your work, or participates in what you’re building.

A community is visible through behavior:

• People return

• People ask questions

• People help one another

• People share what they’re testing

• People participate without expecting something in exchange

That is why the difference between four likes and 34 replies matters.

A like is a quick reaction.

A reply starts a relationship.

Chat can also become a content engine

The replies revealed the actual problems people face while growing on Substack.

Those conversations can become:

• New Notes

• Detailed articles

• Growth experiments

• Tutorials

• Tools

• Future Chat discussions

This creates a useful loop:

Ask a question → understand the problem → publish an answer → attract new readers → invite them into the next conversation.

Instead of guessing what my audience wants, I can listen to what members are already asking.

What this result does not prove

Thirty-four replies do not mean the Chat generated 34 subscribers.

It also doesn’t prove that every new Substack publication should create a Chat.

This experiment only shows that one focused question generated significantly more conversation than I expected.

The next step is measuring whether those conversations lead to:

• Returning participants

• More profile visits

• New subscribers

• Better-performing Notes and articles

• Recommendations from members

That is what I’ll be tracking next.

Profile visits only count if the profile answers who you are and what you publish. I later built a free Substack profile audit that scores what a visitor sees, out of 100.

The simple setup you can reproduce

If you want to test Substack Chat for your publication:

1. Give the Chat one clear topic.

2. Explain what members can discuss.

3. Start with one easy, specific question.

4. Respond to early participants.

5. Use their problems to create future content.

6. Continue the conversation with focused follow-up threads.

You don’t need hundreds of active members.

Ten people who regularly return and participate can be more valuable than hundreds of silent subscribers.

My current hypothesis is simple:

Sub-for-sub can help you escape zero. Community may be what helps you grow after that.

I’ll continue sharing the real results as I test it.

If you’re trying to grow on Substack, join the Subshack Growth by Kacoulib Chat and tell us the biggest challenge you’re currently facing.

Subscribe now

It helped me move from zero to more than 150 subscribers quickly.

That result was real.

But it left one important question unanswered:

Can I continue growing without reciprocal subscriptions?

Getting someone to subscribe because you subscribed to them is not the same as attracting someone who genuinely wants to read your work.

So this week, I’m stopping subscribe-for-subscribe completely.

No “subscribe and I’ll subscribe back.”

No reciprocal-growth threads.

No promising a follow in exchange for a follow.

For seven days, I’ll rely on original Notes, meaningful replies, useful articles, a custom domain, and improved SEO settings.

This is day two of the experiment.

Why I stopped subscribe-for-subscribe

Subscribe-for-subscribe solved one immediate problem:

It helped me escape zero.

When a publication is new, almost nobody knows it exists.

Reciprocal-support threads put my profile in front of other creators who were also trying to grow.

That created momentum quickly.

But acquisition speed is not the only thing that matters.

A subscriber gained through reciprocity may have subscribed because of the exchange—not because they were interested in the publication.

That does not make the subscriber worthless.

It simply means the original motivation may be different.

The real test comes afterward:

• Will they open the next article?

• Will they read again?

• Will they reply?

• Will they stay subscribed?

• Will they care about the topic when reciprocity is no longer involved?

This experiment is not designed to prove that subscribe-for-subscribe is bad.

It is designed to find out whether I can build a second growth engine that does not depend on it.

What I’m testing instead

For the next seven days, I’m focusing on five things.

1. Original Notes

During my first week, I published more than 40 Notes.

Some promoted articles.

Some shared observations.

Some asked questions.

Some documented what I was learning about Substack growth.

This week, the Notes will have one additional constraint:

They must be useful even if nobody subscribes afterward.

I’ll publish Notes around:

• practical growth observations

• experiments

• mistakes

• content performance

• questions creators are already discussing

• lessons from building a publication from zero

The goal is not simply to publish more.

The goal is to find out which ideas create genuine interest.

2. Meaningful replies

I’ll continue replying to other creators.

But I will not promise a subscription or follow in return.

Each reply should add something useful to the conversation.

That could be:

• a concrete observation

• a relevant question

• a nuance

• a lesson from my own experiment

• a related article when the connection is genuinely useful

The reply should still make sense if the promotional link is removed.

This matters because replies can create discovery without turning every interaction into a transaction.

3. Original articles

During my first week, I published several articles about what I was testing.

The articles were not my largest direct acquisition source.

But they served another purpose.

They gave profile visitors context.

Context only helps if the profile itself is clear. I built a free Substack profile audit that scores what a visitor actually sees, out of 100.

A Note may make someone curious.

An article can help them understand:

• what I write about

• how deeply I explore a topic

• whether the publication is relevant to them

• whether they want to keep reading

This week, I’ll continue turning short observations into deeper articles.

Notes create entry points.

Articles create depth.

4. A custom domain

I also connected a custom domain to the publication.

This will not magically create traffic.

And I do not expect a meaningful SEO result after only seven days.

The reason is longer-term ownership.

I want the publication to become associated with a domain that I control, not only a platform profile.

That should make it easier to build:

• a recognizable brand

• direct traffic

• search visibility

• links that remain connected to my own domain

• a future home for the tools I’m building

The short-term experiment is about Substack growth.

The custom domain is a longer-term asset.

5. SEO improvements

I reviewed the publication and post settings to improve basic SEO.

That includes:

• clearer titles

• stronger descriptions

• relevant topics

• internal links between related articles

• useful image previews

• a more consistent publication position

Again, seven days is too short to properly evaluate search performance.

Google may take longer to discover, crawl, and rank the content.

So I will treat any search traffic this week as an early signal—not proof that the SEO work succeeded.

What I will measure

Subscriber count is still useful, but it will not be the only metric.

I’ll track:

• new subscribers

• article views

• open rates

• replies

• comments

• restacks

• direct traffic

• traffic from Substack Notes

• search traffic

• unsubscribes

• returning readers

I also want to compare the quality of the signals.

For example:

A Note that generates ten replies may reveal more genuine interest than one that brings five silent subscribers.

An article with fewer views may still be valuable if readers return afterward.

A slower week may still be better if the audience is more engaged.

What would count as success?

Success does not require matching the speed of subscribe-for-subscribe.

My first week was optimized for rapid acquisition.

This week is testing a different system.

The experiment will be useful if I can answer three questions:

1. Can I continue acquiring subscribers without reciprocity?

2. Do the new subscribers engage more with the content?

3. Which Notes, replies, and articles create the strongest signals of genuine interest?

Growth may slow down.

That would not automatically mean the experiment failed.

A smaller number of genuinely interested readers may be more valuable than a larger number of inactive subscribers.

The limitation of this experiment

This is not a perfectly controlled test.

I am changing several things at once:

• stopping subscribe-for-subscribe

• changing my Notes strategy

• publishing new articles

• improving internal linking

• using a custom domain

• updating SEO settings

That means I will not be able to attribute every result to one specific change.

If subscriber growth continues, I cannot say that SEO caused it.

If growth slows down, I cannot say that stopping subscribe-for-subscribe was the only reason.

The goal is simpler:

Can this new system sustain growth without reciprocal subscriptions?

If it works, I can isolate individual parts in future experiments.

My hypothesis

My current hypothesis is:

Growth will be slower, but the people acquired may show stronger signs of genuine interest.

I may get fewer subscribers.

But I may see:

• better conversations

• more article engagement

• more returning readers

• fewer transactional subscriptions

• a clearer understanding of which content attracts the right people

That is what I want to test.

Why this experiment matters

My first week answered one question:

Can subscribe-for-subscribe create fast initial momentum?

For me, the answer was yes.

This week is designed to answer a harder question:

Can original content, useful interactions, and long-term discovery channels keep the publication growing without reciprocity?

I’m currently on day two.

At the end of day seven, I’ll publish:

• subscriber growth

• traffic sources

• article performance

• engagement

• unsubscribes

• what appeared to work

• what failed

• what I still cannot prove

The results may be weaker than the first week.

That is part of the experiment.

Subscribe to receive the full seven-day results—whether the numbers are good or bad.

Subscribe now

In my first week, I reached 150+ subscribers.

I didn’t use paid ads.

I tested three things:

1. Subscribe-for-subscribe

2. Substack Notes

3. Original articles

The goal was simple:

Understand what actually creates growth.

The experiment

Period:
First week

Actions:

• 5 published articles

• 40+ Notes

• daily engagement with creators

• subscribe-for-subscribe interactions

Strategy 1: Subscribe-for-subscribe

This was the biggest acquisition driver.

It helped me escape zero quickly.

Why it works:

New creators already have a growth mindset.

They are actively looking for:

• new writers

• communities

• interesting publications

The limitation:

A subscription does not automatically mean a reader.

The next challenge is retention.

Strategy 2: Substack Notes

Notes became my discovery channel.

They helped me:

• start conversations

• share experiments

• bring people to my profile

• test ideas quickly

Short content created repeated visibility.

Notes bring people to your profile. What they see there decides the rest, so I later built a free Substack profile audit that scores it out of 100.

Strategy 3: Original articles

Articles were not the biggest acquisition source.

But they played another role.

They answered:

“Why should I subscribe?”

A Note can get attention.

An article can build trust.

The surprising lesson

I initially thought:

“Better articles = more subscribers.”

The data showed something different.

Growth came from combining:

Distribution + Trust.

Reciprocal interactions created the first contact

Notes created repeated exposure.

Articles created context.

What I will test next

Now that I understand acquisition, the next experiment is retention.

I will measure:

• open rates

• returning readers

• replies

• unsubscribes

• article conversion

Because 150 subscribers is only the beginning.

The real question:

How many become active readers?

I’m documenting the numbers, mistakes and results as I go.

Subscribe to follow the next experiment — when I stop measuring how many people subscribe and start measuring how many actually stay.

Subscribe now

Audit your Substack free →

A few days ago, I had no Substack audience.

Then I reached 100 newsletter subscribers during my first week.

Before that first week was even over:

100 → 150 subscribers.

The obvious reaction would be to celebrate the number.

But I’m more interested in understanding why the growth continued after 100 — and whether those subscribers become real readers.

I’m documenting these experiments as they happen: the numbers, what I tried, what appeared to work, and what I still cannot prove.

The baseline

I started this publication from zero.

During my first few days, I tested several ways of getting discovered:

• participating in creator and subscribe-for-subscribe Notes

• replying to other writers

• publishing my own Notes

• sharing growth experiments publicly

• writing articles about what I was learning

• linking those articles when they genuinely matched a conversation

That helped me reach my first 100 subscribers.

But the interesting part is that the growth didn’t stop at 100.

Before my first week was finished:

Subscribers: 150

That means another 50 subscribers arrived after the 100-subscriber milestone.

What changed after 100?

I didn’t suddenly discover a new growth hack.

I kept doing largely the same things.

But one thing had changed:

I now had more content for people to discover when they landed on my profile.

Early on, someone visiting my profile saw almost nothing.

Now they could see:

• growth experiments

• original Notes

• opinions about Substack growth

• actual results

• mistakes and lessons

• articles explaining what I was learning

The profile itself had become more useful.

If you want to see what your own profile shows a first-time visitor, I built a free Substack profile audit that scores it out of 100.

Instead of asking:

Who is this person?

a visitor could start asking:

Do I want to keep reading this?

Notes created the surface area

One thing is becoming increasingly clear to me:

Notes have been one of my most important discovery surfaces.

Every Note creates another opportunity for someone to:

• see my name

• read an idea

• reply

• visit my profile

• discover an article

• subscribe

That doesn’t mean posting for the sake of posting.

The Note still needs to give someone a reason to stop.

But original Notes continuously create new entry points into the rest of my content.

Reciprocal support helped — but it is not enough

I also participated in reciprocal-support Notes.

Things like:

Subscribe = Subscribe
Follow = Follow

They helped create initial visibility.

For a publication starting from zero, that can be useful.

But someone subscribing because you subscribed to them is not necessarily subscribing because they want your next article.

That’s why I started publishing original Notes alongside those interactions.

Reciprocity can create the introduction.

Your content has to give people a reason to stay.

Different content may have different jobs

Another pattern is starting to appear in my article statistics.

My more opinionated article about subscribe-for-subscribe attracted more attention and discussion than my straightforward growth reports.

But one of my practical articles was better at directly generating a subscriber.

That suggests different content may serve different purposes:

Opinionated content → attention

Practical content → potential conversion

Notes → discovery

Replies → conversations

Growth reports → credibility

I don’t have enough data yet to call this a rule.

But it changes how I think about publishing.

150 subscribers does not mean 150 readers

This is the part I care about next.

150 subscribers sounds good.

But it doesn’t tell me:

• how many will open the next article

• how many will return

• how many will reply

• how many will unsubscribe

• how many joined mainly because of reciprocity

• how many genuinely care about the topic

Subscriber count measures acquisition.

It does not automatically measure audience quality.

What I still cannot prove

I cannot accurately attribute all 150 subscribers to specific actions.

Several things were happening simultaneously:

• Notes

• replies

• reciprocal-support threads

• articles

• profile visits

• Substack discovery

So this is not:

“Do these exact things and you’ll get 150 subscribers.”

It’s an experiment happening in public.

The growth is real.

The attribution is still incomplete.

What I’m testing next

Getting subscribers is no longer the main question.

Now I want to understand what happens after the subscription.

I’ll be tracking:

• article views

• open rates

• replies

• comments

• profile visits

• new subscribers

• returning readers

• retention

I’ll also continue testing different types of Notes:

• practical tactics

• opinions

• questions

• experiments

• progress updates

• warnings about common growth practices

The question becomes:

Which actions attract people who actually keep reading?

Conclusion

Getting from zero to 100 subscribers was the first milestone.

Going from 100 to 150 before the first week was even over was more interesting.

It suggests that early distribution becomes more valuable as your profile builds up more content worth discovering.

Notes create discovery.

Articles provide depth.

Conversations create familiarity.

Now the real experiment begins:

Can those 150 subscribers become an actual audience?

I’m documenting the numbers, mistakes and results as I go.

Subscribe to follow the next experiment — when I stop measuring how many people subscribe and start measuring how many actually stay.

Subscribe now