How We Built It: 22M Views in 6 Months, Plus 3 More Case Studies
Your product can be copied. An audience that already knows you can't. This page shows how we build that audience, with four results from the same engine, including our own account.
One of our clients crossed 22M+ combined views in 6 months and gained 80K+ organic followers. Instagram only, one platform. He filmed once, for 30 minutes, at the start.
This is the architecture that did it. Not a tool list.
And the part most people get wrong: the machine is the easy half. Anyone can buy an avatar for $30. Almost nobody can make it sound like a person worth following. That gap is strategy, and it is the half that actually did the work here.
Four results, one engine
| Client | Format | Result | Time | What it turned into |
|---|---|---|---|---|
| Founder A | AI clone of face and voice | 22M+ views, 80K+ new followers | 6 months | Inbound leads from organic |
| Founder B | AI clone of face and voice | 2K to 400K+ followers, 12M+ views | 8 months | Signups for a coaching program |
| A faceless brand we built | Faceless videos, no one on camera | 100M+ views in a single month | One month | Reach for a brand with no face to show |
| Our own account | The same formats we sell | 0 to 3,000 followers | One month | Proof we run the engine on ourselves first |
Different formats, same engine underneath: a point of view, a voice, one platform, and a cadence that doesn't stop. The rest of this page breaks down Founder A. The other three are below.
The wall he hit first
He already believed video mattered. That was never the gap.
The gap was production. Scripting, filming, editing, posting, every week, forever. Running the company ate every hour, and he hated being on camera. So while quieter competitors shipped weekly, he stayed invisible.
Most founders do not have an idea problem. They have a production problem.
But solving only production is how you get 6 months of content nobody remembers.
Why most AI video fails
AI video does not fail because it looks fake. It fails because it sounds like everyone else.
That is not a tool problem. Generic AI content is a missing strategy layer. And it splits the market cleanly:
- Production-only agencies ship on schedule and produce slop, because they skipped the strategy
- Consultants and brand strategists hand you a positioning deck you never execute
- Neither puts a point of view into something that ships every week without you
We do both halves, and we bake the strategy into the production so it survives contact with the calendar. Strategy that never ships is a document. Production without strategy is noise.
What he did vs what we did
| Him | Us |
|---|---|
| One 30-minute recording session, once | Everything else, every week, for 6 months |
| Approves the direction | The strategy half: point of view, positioning, Voice DNA |
| Runs his company | The production half: scripting, rendering, editing, distribution, all-in on one platform |
That ratio is the whole point. If the founder is still in the loop weekly, it is not a machine. It is a second job.
The 5 layers
This is what the engagement actually looks like. Layers 1 and 2 are the strategy half, the part competitors skip. Layers 3 to 5 are production: necessary, and increasingly a commodity.
1. Brand strategy
Before a single frame, we build your point of view. What you believe, what you will argue against, what you are willing to be wrong about in public.
Slowest part, no shortcut. It is why the videos read as a person with a position instead of a feed of tips.
Output: a positioning spine every future script has to serve.
2. Voice DNA
We extract how you actually talk. Rhythm, the words you reach for, the ones you never use, how you open, how you land a point.
This is the file that makes the output sound like you instead of like AI. Highest-leverage asset in the system, and why our client's audience never flagged his videos as synthetic.
You keep it. Walk away in a year and the point of view, the Voice DNA, and the angles are yours. You own the strategy, not just rent the output.
3. Capture
One 30 minute session, once. We generate every video from it. That is your entire on-camera obligation for the life of the engagement. This is what our 22M client chose.
4. The production machine
Scripted against your spine and your Voice DNA. Rendered, edited, packaged. A weekly show plus daily shorts, on a cadence that does not depend on your calendar, your travel, or your mood.
Cadence makes it compound. Voice makes it worth compounding.
5. Distribution and signal
Everything went to one platform, deep instead of wide. We watch who engages there. Not vanity views. Which specific people in his market keep showing up.
That is what turns attention into inbound. For our client, founders who had watched him for months eventually landed in his DMs asking to work with him. Knowing which signal closed the deal is what separates a content channel from a pipeline.
What 6 months looked like for him
| Window | What happened |
|---|---|
| Month 1 | Strategy, Voice DNA, the one recording session. First episodes ship. |
| Month 2 | Cadence locks in. Nothing looks impressive yet. This is where most founders quit. |
| Months 3 to 4 | Compounding starts. Back catalogue keeps working. Recognition in his market. |
| Months 5 to 6 | 22M+ cumulative views. Inbound arriving unprompted. |
This is one client's run, not a forecast. Your numbers will be yours. What repeats is the shape: a slow start, weeks where nothing looks like it is working, and compounding for whoever stays in.
Founder B: 2K to 400K+ followers in 8 months
Same setup as Founder A: one 30-minute capture session, then an AI clone publishing on a steady cadence.
The first five months were quiet. Month six broke. Eight months in, the account had gone from about 2,000 to 400,000+ followers and 12M+ combined views, and that traffic turned into signups for a coaching program.
The lesson is the same one as above, only longer: the months that look like nothing are the ones that make the rest possible. Read the full case study.
A faceless brand: 100M+ views in a single month
Not every business has a founder who should be the face. For this one we built a faceless brand: nobody on screen, the same strategy layer underneath, and videos shipped on a fixed cadence. It reached 100M+ views in a single month.
If you want to see the formats we use for this, start with The Faceless Story Kit, The AI Mascot Playbook and AI Video Examples.
Our own account: 0 to 3,000 followers in a month
We don't sell anything we don't run on ourselves. We started our own account from zero and grew it to 3,000 followers in its first month, using the same formats on this page: an AI clone, faceless stories, mascots and AI creators, shipped on a fixed cadence.
It's the smallest number here and the one we're proudest of, because it's the test every new format goes through before a client sees it.
These are individual results, not a forecast. They show what the engine can do when it runs unbroken, not what any one account will do.
What we guarantee, and what we do not
We do not promise overnight. Nobody honest can put a date on attention.
What we guarantee is the strategy and the shipping, because that is what sits inside our control:
- A point of view and a Voice DNA built before anything ships, and every script gated against them
- Founder video shipped every week, from your clone or from footage you send, each one building the same authority instead of chasing whatever is trending
- Distribution locked to the platform that fits your audience, not spread thin across all of them
- A report of who engaged, so attention becomes names
That combination is the whole bet.
Strategy makes each video say something worth remembering.
Cadence makes them stack.
Videos that stack in one direction build authority.
Videos that chase trends just accumulate.
The outcomes are real. They are just not a schedule. 22M views and inbound arriving unprompted is what this produced once it ran unbroken for 6 months. Month two looked like nothing. That is the trade: we guarantee what we control, and the authority compounds for the founders who stay in.
Anyone quoting you a guaranteed view count by a date is selling you the commodity half and lying about the rest.
Who this works for
It works when three things are true:
- You are past the point where this is an investment and not a bet
- You are willing to be the face, but not to run the production. Clone or weekly footage, either way we run everything after
- You think in quarters, not posts. Month two will look like nothing and you keep going anyway
Three is the one that decides it. Most founders clear the first two on paper. The ones who compound are the ones still shipping in month six, before anything looked like it was working.
Related knowledge
- Explore the AI Avatar map
- Build your first avatar in HeyGen
- Avatar production workflow
Want this set up for your business?
We turn these workflows into working marketing and sales systems.