Stop guessing what to post.
Let the outliers tell you.
Most content fails because nobody really knows what works in a niche. I built a small bot that scans every post on the top accounts in my niche, calculates which ones crossed 20 times the follower count, and reports those outliers back. The outliers are the posts that took each account from unknown to known. They become the templates for my own breakout posts, on autopilot.
The posts that actually built the account, not the ones that look big.
When you look at a large account, almost every post looks like it did well because the account is large. That is misleading. The posts that actually grew the account were the ones that broke far above the account’s normal numbers, before it was famous. Those are the outliers. This bot finds them. Here is who it is for, what goes wrong without it, how it works, and what you get back.
Who it’s for
Anyone running a social account who needs to know what to post next, and is tired of guessing. Course creators, brand owners, niche page operators, anyone studying competitors to build a content calendar. If you have ever stared at a blank screen wondering what to post, this is for you.
What goes wrong
Without an outlier filter, people copy the wrong posts. They study the most recent or most polished posts on a big account and miss the ones that actually built it. So they spend months making content that looks like a big account’s daily output, and wonder why nothing takes off. You cannot model success by copying the average, you have to copy the outliers.
How the machine works
The bot reads your competitor list, scrapes the last 100 posts on each account along with views, likes, comments, and the account’s follower count. It calculates the view-to-follower ratio for each post and filters everything below 20 times. The posts above that line are the outliers. The bot saves them in a library, sorted by ratio. You are left with proof of what actually works in your niche.
What you get back
A library of roughly 30 to 50 outlier posts a month, broken down into their hook, structure, and payoff. You feed those structures to your AI writer, or you use them as templates yourself. Every post you make from then on starts from a shape that has already proven it can break out.
I was tired of copying the wrong posts.
When I first studied competitor accounts in my niche, I did the same thing everyone does. I looked at their latest posts, their most polished posts, their pinned posts, and tried to copy that style. None of it worked. My accounts stayed small while the big ones kept growing.
After a while I realised the trap. A big account with a million followers gets fifty thousand views on a routine post. That post did not build the account, the audience was already there. The posts that actually built the account were buried months or years back, when the account was small and one post broke far above what its size could explain.
So I built a bot. It scrapes every post on a list of competitor accounts, pulls views, likes, comments, and the follower count, then calculates the view-to-follower ratio. Anything below 20 times the follower count gets thrown out. Anything above is saved into an outlier library. The bot does this across 15 to 20 accounts in my niche at once.
The library gives me 30 to 50 outliers a month. Each one is a post that genuinely outperformed its account, which means the structure itself is doing the work. I take those structures, break each one down into hook, body, payoff, and tone, and feed them to my AI writer as templates. From then on my own posts start from a proven shape, not a blank page.
Then a second layer kicks in. The templates also train my long-term AI assistant on the patterns that work in my exact niche, so it gets sharper over time. Each new month of outliers makes the next month’s writing better. The system compounds because the data compounds.
Proof point: I have shown how I use intelligence pipelines like this to make content decisions in my business in my growth hacking series and walked through how outliers drive my content selection in my $3k+ from one social post breakdown, so the numbers and method on this page are checkable.
Three moves that replace guessing with proof
What made this work was switching the question. Most people ask what is this account posting now. The better question is what posts actually built this account. The outlier lens is the answer to that second question, and the entire framework hangs on it.
Compare every post against its own account, not against absolute numbers
A 5,000-view post on an account with 200 followers is a breakout. The same 5,000-view post on an account with 2 million followers is a flop. Absolute numbers lie. View-to-follower ratio tells the truth. By comparing every post to the account’s own follower count, the bot finds the posts that genuinely went out beyond the account’s normal reach, regardless of whether the account is big or small. That single decision is what separates a useful library from noise.
Set the threshold high enough to keep only real signals
Twenty times the follower count is not arbitrary. It is the line above which a post can only be explained by the algorithm pushing it to non-followers, which is the entire point of an outlier. Lower the threshold and the library fills with posts that did slightly better than average, which is not actionable. Raise it and the rare big breakouts come through cleanly. The threshold is a knob, and the right setting decides whether the library is full of breakouts or full of background noise.
Break outliers into structures, not into copied lines
An outlier is not a script to copy word for word. The reason it broke out was the structure underneath. The bot saves each outlier alongside its breakdown: the hook style, the body structure, the payoff. That breakdown is the asset, not the original caption. From there you, or your AI writer, can produce new posts inside your own brand voice that follow the proven shape. This is the difference between cheap copying and intelligent modelling, and the only one that actually grows an account.
Once those three moves are in place, the library compounds month over month. The outliers from last month become the templates for this month’s writing, and every new outlier sharpens the AI that produces tomorrow’s content.
Before the outlier filter
- Studied the most recent or most polished posts on big accounts
- Copied the surface style and wondered why nothing broke out
- Spent months posting content that looked like a big account’s daily output
- Had no objective way to know what was working in the niche
- Treated content as a guess every single week
After the outlier filter
- Bot scans every post on 15 to 20 accounts in the niche
- Only posts above 20x the follower count are kept as outliers
- 30 to 50 outliers a month, each broken into hook, body, payoff
- AI writes new posts from structures that already proved themselves
- Content decisions backed by data, not by guessing what looks good
Prompt 1: pick the 15 to 20 accounts the bot should watch
The bot is only as useful as the account list it scans. Watch the wrong accounts and the outliers are irrelevant. Watch the right ones and every outlier is a direct signal for your niche. Use this prompt to build a precise list.
Niche account list builder
Act as a social media research strategist. I am building a competitor analysis pipeline that scrapes posts from a list of accounts and surfaces outliers (posts with more than 20x the account's follower count in views). My niche is: [describe in one sentence] My buyer profile is: [describe in one sentence] Platform: [Instagram, TikTok, YouTube Shorts, X, etc.] Give me 20 accounts that meet all four criteria: they post in my niche, they have an active posting schedule over the last 90 days, they have at least 5,000 followers (so there is enough data), and they show variation in their post performance (some posts much bigger than others). For each one, give the handle, follower count, posting frequency, and a one-line note on why their audience overlaps with mine.
These 20 accounts are the input to the bot. A good list produces a library of useful outliers. A weak list produces noise.
Prompt 2: decide on the outlier threshold for your niche
Twenty times the follower count is a strong default. Some niches need a different setting because the natural reach distribution is different. Use this prompt to confirm the right threshold for the niche you picked.
Threshold calibrator
Act as a content data analyst. I am running an outlier filter on social media posts. The filter compares each post's views to the account's follower count and keeps only posts above a certain ratio. The default ratio is 20x (a post's views are more than 20 times the account's follower count). My niche is: [paste from prompt 1] Platform: [paste from prompt 1] Typical engagement style in this niche: [one sentence, for example educational long captions, short visual hooks, and so on] Tell me whether 20x is the right outlier threshold for this niche, or whether I should set it lower (10x, 15x) or higher (30x, 50x). Explain the reasoning in one paragraph, then give a final recommended threshold and a 'caution' note on what would happen if I set it too low or too high.
Lock the threshold once. After that, the bot does the same calculation every run and the library quality stays consistent.
The 3-minute overview of how this works
Before the build steps, watch this short overview. It’s the exact video from the Automations Made Easy page, and it walks through the mechanics behind machines like this one. 1,000+ students have used these mechanics to save two hours a day, with zero coding.
Want to learn the mechanics behind intelligence automations like the outlier finder?
I teach the same mechanics that make this competitor analysis bot work inside Automations Made Easy. 1,000+ students have used these mechanics to make better content decisions in their own niche, without paying expensive research agencies. No coding required.
Week one: roughly 12 outliers, 12 templates.
The first week feels small for a reason. The bot scans 15 to 20 accounts, calculates the ratios, and surfaces the outliers across the entire history of posts available. The early library is whatever the last 100 posts per account contain. In a typical niche, that lands somewhere around 12 outliers in the first week.
Twelve outliers is twelve templates. Twelve genuinely proven shapes you can reuse in your own brand voice across the next month of posts. That alone changes the posting calendar from a weekly anxiety into a structured selection: pick a template, fill in your own angle, ship.
The point of week one is to confirm two things. First, the bot returns enough outliers to be useful in your niche. Second, the structures the outliers reveal actually fit your brand voice. If both check out, every following week compounds the library further.
Across a month the system surfaces 30 to 50 outliers, depending on niche. Across a year that is 400 to 600 proven content shapes in your library, every single one of them backed by real data from your competitors. That library trains your AI writer, fills your content calendar, and stops the guesswork that costs most accounts their first six months.
The bot runs in the background, the library grows on its own, and the content decisions you make every week get sharper, not harder.
Prompt 3: break each outlier into its reusable structure
An outlier is most useful when you understand why it worked. Use this prompt on each outlier the bot surfaces to extract the hook, body, and payoff structure that you can reuse in your own posts.
Outlier structure extractor
Act as a content engineer. I have a post that outperformed its account by more than 20 times its follower count, which means the structure is doing real work. Here is the post: [paste the caption, the hook, and a description of the visual or video opening] Niche: [from prompt 1] Break the post into three components. Hook: the first 1-2 lines that earned the click or the stop-scroll. Body: the structure of the middle (steps, list, story arc, comparison, etc.). Payoff: how the post resolves and what the implicit call-to-action is. After the breakdown, give me a one-paragraph template I could feed to an AI writer to produce a new post in this exact shape, in my own brand voice.
The template, not the post, is the asset. Every outlier produces one template you can reuse forever inside your own voice.
Prompt 4: train your AI writer on the growing outlier library
Once a few months of outliers stack up, the library itself becomes training data. Use this prompt to summarise the patterns the AI should internalise so every new post you generate starts inside the proven outlier shape for your niche.
AI writer trainer
Act as a content pattern analyst. I am training my AI writer to produce posts that follow the shape of breakout outliers in my niche. Here is a sample of the outlier library: [paste 5 to 10 outlier breakdowns from prompt 3] Niche: [from prompt 1] My brand voice: [describe in one sentence] Summarise the recurring patterns across these outliers. Common hook types. Common body structures. Common payoff shapes. Common emotional beats. Then write a single system instruction I can paste into my AI writer that tells it to follow these patterns whenever it produces a new post for me. Keep the system instruction under 300 words.
From now on, every post the AI writes for you starts inside the outlier shape your niche has already proven works. The guessing is over.
The exact build, step by step
Pick 15 to 20 accounts in your exact niche
Sit down and pick the 15 to 20 accounts whose audience matches yours and that post in the same niche. They do not need to be huge. They need active posting in the last 90 days and a mix of post sizes (so the data is varied). Save the handles in a sheet with their current follower count. This list is the bot’s input.
Connect a scraper that pulls views, likes, comments, and follower count
Wire up a scraper (Apify actor, no-code scraping tool, or a custom run) to pull the last 100 posts from each account in your list along with views, likes, comments, post type, and a snapshot of the account’s follower count. Run it once a week. Dump the result into a single sheet with one row per post.
Calculate the view-to-follower ratio per post
In the sheet, add a column: views divided by follower count. That column is the outlier signal. A value of 20 or higher means the post broke out far beyond the account’s normal reach. The math is trivial but the move is everything: you are now comparing posts against their own account, not against each other in absolute terms.
Filter the sheet to keep only the outliers
Apply a filter: keep only rows where the ratio is above 20. Everything else is noise for this purpose. Save the filtered view as your outlier library. In a typical niche you will see 5 to 15 outliers from the first run, then 10 to 20 per week after that as new posts roll in.
Break each outlier into hook, body, payoff
For every outlier in the library, write a one-paragraph breakdown that names the hook style, the body structure, and the payoff. Use the prompt above to make this automatic. Add the breakdown as a column next to each outlier. This is the asset you actually reuse, not the original caption.
Feed the breakdowns to your AI writer as templates
Take the breakdowns and paste them into your AI writer as system instructions. From then on, when you ask the AI for a new post, it produces one that follows a proven outlier shape from your niche, in your own brand voice. Every new outlier the bot finds expands the AI’s repertoire further. The library compounds, the AI gets sharper, and content stops being a weekly fight.
Account list
Fifteen to twenty competitor accounts in your niche, mixed sizes, active posting in the last 90 days.
Per-post stats
Scraper pulls views, likes, comments, and follower count for the last 100 posts on each account.
Outlier filter
View-to-follower ratio per post, keep only above 20x, throw out the rest as noise.
Templates and AI
Each outlier broken into hook, body, payoff. The breakdowns train your AI writer for every future post.
Build this outlier finder inside the same playbook 1,000+ students use
Automations Made Easy teaches the mechanics behind intelligence pipelines like this one. Step by step, no code, plain English. Save two hours a day and stop guessing what your next post should look like.
The six months after I switched it on
Here is the shape of the first six months of the outlier library after I switched on the bot for one niche. The line is intentionally steady from month three onward, because that is how a compounding library behaves once the early backlog is processed.
Outliers added to the library per month
Three things matter on this chart. The library stabilises around 40 to 50 outliers a month. The outliers are niche-specific, not generic content tips. And the AI writer trained on the library produces content that consistently lands inside the shape of what already works.
What other students built
I teach the simple skills behind machines like this in Automations Made Easy. Students who built their own version sent back what changed in their first month.
“I thought I was studying the right accounts but I was looking at all the wrong posts. The outlier filter showed me which posts actually grew them, and my own engagement jumped within three weeks of using those structures.”
“I run a coaching account and I used to write four posts a week from scratch. Now I write from the outlier templates and I publish twice as often because the blank page is gone.”
“The breakdown into hook, body, payoff is the part nobody else does. Most competitor tools show you what is trending. This one shows you what is structurally working in my niche.”
“We trained our AI on three months of outliers from our niche. Our content team now spends most of their week filming, not writing. The writing is solved.”
What’s inside Automations Made Easy
AME isn’t a library of pre-built automations. Every business is slightly different. What’s reusable across all of them is the underlying mechanics: how to set up little machines that listen, write, and follow up while you sleep, and how to wire the pieces together without writing code.
The program walks you through six modules: The Right Tools (the cost-effective, no-code stack I actually use), Task Selection Mastery (which automations are worth building first), Design Secrets (mapping an automation before you build it), Zero to Hero (complete beginner to confident automator), Real-World Application (we build a full automation together, end to end), and Monetization Mastery (turn the skill into a side-business).
It also includes done-for-you templates you import in two clicks, over-the-shoulder training videos, and the same playbook 1,000+ students have used to save two hours a day. No coding required. If you can copy and paste, you can build this.
The outlier finder: common questions
Pulled from what readers and Automations Made Easy students ask most.
Why 20x specifically?
Twenty times the follower count is the line above which a post can only be explained by the algorithm pushing it to non-followers at scale. Below that line, a post can do well from its own audience and a small viral bump, which is harder to learn from. Above it, you have something genuinely breakout. In some niches the right threshold is 15x or 30x, which is why prompt two has you confirm it for your niche before locking it in.
Do I need a paid scraper to do this?
You need some way to pull post stats and follower counts at scale. There are paid scrapers (Apify actors, scraping APIs) and there are no-code options. The cost is usually a few cents per account per week, which is far cheaper than a content researcher. The exact tool depends on the platform you are analysing. The mechanic, view-to-follower ratio above a threshold, is the same regardless of how you get the data.
What if my niche has very few large accounts?
You do not need large accounts. You need accounts that post in your niche with variation in their performance. A small account with one breakout post is more useful than a huge account with a flat line, because that breakout is exactly the signal you want to learn from. Mix sizes in the list.
Will I get banned for scraping these accounts?
You are reading public data: post counts, view counts, and follower counts that any visitor to the profile can see. The risks come from making too many requests too quickly from a single IP. Most paid scrapers handle that for you with proxies and rate limits. If you build your own, use sensible delays and rotate IPs and you stay inside the safe zone.
Will the AI writer copy the outliers and get me flagged?
No, because the AI uses the structure, not the original text. It produces new content inside a proven shape, in your brand voice, on your topic. The structure is generic enough to be reused (hook + steps + payoff is not copyrightable). What you are reusing is the architecture, not the actual lines.
More Money Makers like this one
Built from the same handful of mechanics. Each one captures, converts, or earns money that would have walked out the door.
Run this outlier finder yourself, or learn the mechanics inside Automations Made Easy.
If you want to learn the mechanics behind intelligence automations like this and build your own in your niche, Automations Made Easy is the playbook. Step by step, no code, plain English. If you want help mapping the right accounts and threshold for your specific niche first, I take a small number of consulting clients each month.
€497 one-time · Lifetime access · 1,000+ students
