Long-form video
16:9, 5:28, narrated, burned-in subtitlesChapters 9
Chapters jump to the moment in the player. Download the original.
Fractalic Studio, Compound
Don't publish. Compound.
You have already done the hard part. Someone on your team researched a subject, argued about it, and wrote it down properly. We take that piece and build everything else out of it, so it keeps working long after the week it was published.
$490 per package: ten files from one article you already published, checked by a person.
We start by reading what you have already published and telling you which pieces are worth it. See how it is priced.
Running for
Most good writing gets one week of attention
You publish something you are proud of. It does reasonably well for a few days, and then it goes quiet. A month later it is a URL nobody opens. The work was good. It just never got a second life.
Read more
At the same time, the way people find things has been changing under everyone's feet. More of them now ask a question and read the answer instead of working through a page of links. An article that exists only as an article on one page has a much smaller chance of being part of that answer.
The usual response is to publish more. That mostly produces a bigger archive that nobody revisits. The material you already have is the part worth pushing on, because the expensive thinking is finished and paid for.
What one package is
One source in. Ten files out, each one built for where it will actually live, not exported from a template.
Input
One published asset
- 01Long-form videonarrated, burned-in subs
- 02Vertical cut9:16 for Reels and Shorts
- 03Subtitle fileSRT, timecoded
- 04Transcriptplain text
- 05Infographic, landscapePNG and WebP
- 06Infographic, portraitlead-magnet ready
- 07Thumbnail variants3, for A/B
- 08Platform-native copy6 platforms
- 09JSON-LD schemamachine-readable
- 10Episode coverJPG, 1920×1080
White glove, quoted separately
The package price covers the files. If you want them to land without adding work to your team, we upload them to your CMS, build hub pages, and can build and run an authority hub on your own subdomain. That work is quoted separately.
We compounded this page
This page is the source. Every file below was made from it, the same way we would make it from one of your articles, and a person checked each one before it went up.
Vertical cut
9:16, 1:07, for Reels and ShortsOpen files 03 to 10 subtitles, transcript, infographics, thumbnails, copy, schema, cover
Subtitle file
SRT, 127 cues1 00:00:03,160 --> 00:00:04,600 You spend hours researching and 2 00:00:04,600 --> 00:00:06,420 writing something you are proud of. 3 00:00:06,420 --> 00:00:07,500 You hit publish, it gets 4 00:00:07,500 --> 00:00:09,580 traffic, and then, it fades. 5 00:00:12,560 --> 00:00:13,880 A month later, that same
Transcript
plain text, 810 wordsOne week of attention
You spend hours researching and writing something you are proud of. You hit publish, it gets traffic, and then, it fades. A month later, that same link sits completely untouched. All the expensive thinking, all the research your team paid for, is locked inside a URL that nobody opens twice. The work itself is good. The problem is what happens after. Your team did the hardest part of the job, the original thinking, and then let it expire after one short week of attention.
Read the rest of the transcript
How people look for answers now
At the same time, the way people look for information has changed. Fewer people work through a page of blue links. More of them ask a question and read the answer. An article that lives only on one web page has a much smaller chance of being part of that answer. It has no video, no visuals, and nothing that travels to the other platforms where people actually spend their time. The usual reflex is to publish more text. More posts, more pages, more of the same. That mostly produces a bigger archive that nobody revisits, and the distribution gap stays open. The material you already have is the asset worth pushing on. The expensive thinking is finished. The better path is to take those proven ideas and rebuild them in new formats.
One package, ten files
Here is what one package is. One published asset goes in. Ten files come out. First, the heavy media: a long video with subtitles, a vertical cut, an SRT file, a transcript. Then, architecture: landscape and portrait infographics, three thumbnail variants, platform-native copy, JSON-LD schema, and an episode cover. This is not an export from a generic template. Each of the ten files is built for the place where it will actually live: the feed, the search result, the phone screen.
The seenlabs run
Here is how we know it works. We ran it for SeenLabs, a digital signage company. They had a library of genuinely useful articles that were almost never read twice. We picked the pieces worth multiplying and turned them into a video series. We produced 39 complete packages, about two a week. Every one of them is still live, and anyone can open it. We kept that up for five straight months, without skipping a single week between the first episode and the last. One or two videos would have told us very little. So we kept going, week after week, and watched who the episodes actually reached.
Strangers, not followers
The view counts on those 39 episodes are only part of the story. What we did not expect was who watched. Almost none of it came from people who already followed the accounts. This chart compares followers to strangers. On YouTube, 99.6% of watch time came from non-subscribers. On Facebook, 99.8% were not followers. Instagram: 95.4%. LinkedIn sits lower, at 52%, because that account has a real audience behind it. We left it in, rather than show you only the comfortable numbers. And note what these numbers measure: how far the work travelled, not how many people came to the website afterwards. Reach and traffic are different numbers, and we do not blur them together.
One article, before and after
One article had been sitting quietly. In the 151 days before the rebuild: 795 search impressions. In the same span after: 17,449.
What we do not claim
But look at the boundaries. During those same months, we also ran internal linking and technical fixes on that site. So the videos alone did not cause this. And nobody can honestly promise you a search ranking, or a mention in an AI answer. Those systems change too often for that. What we can do is agree up front on what we will watch.
How it works
Then we look at the numbers with you, against a baseline we set before anything ships. Here is how it works. First, we pick. We go through your library together and select the pieces worth multiplying. Not all of them. Only the ones that will carry. Next, we build the base: a knowledge base for your brand. Your voice, your visual rules, and the claims and facts we are allowed to assert. Then we produce, one package at a time, each built for its topic. We hand the finished files to your team, and measure against the baseline we set on day one. The pace is up to four packages a week, and that is a real limit. Each package carries its own research, prompts and fact-check. A person checks every file against your source before it ships. That check is what separates this from fully automated AI output. Accuracy comes before speed.
Start with five
Start with your own archive. Pick five articles you were proud to publish and watched disappear after a week. Let us rebuild them. We agree beforehand on what we are watching. Then you look at your own numbers, not ours, and decide whether this becomes a habit. Don't publish. Compound.
Infographic, landscape
1920x1072, WebP
Infographic, portrait
1072x1920, lead-magnet ready
Thumbnail variants
3, for A/B
Platform-native copy
6 platforms, plus YouTubeLinkedIn, company page 169 words
99.6% of the YouTube watch time on our last client series came from people who were not subscribed to the channel.
That series was SeenLabs. A digital signage company with a library of useful articles that were almost never read twice. We picked the pieces worth the effort and rebuilt each one as a package: narrated video, vertical cut, infographics, thumbnails, copy for six platforms, schema. About two a week, for five months, without missing a week. 39 episodes are live today.
One of those articles went from 795 search impressions in the 151 days before the rebuild to 17,449 in the 151 days after. Other work ran on that site in the same months, so we do not credit the videos alone, and we say so on the page.
We made the video below from our own product page, with the same process.
If you have a dozen articles you were proud of and watched go quiet, start with five.
#ContentMarketing #B2BMarketing #ContentStrategy #VideoMarketing
First comment: https://fractalic.studio/products/compound/
---
LinkedIn, founder 66 words
Short version (58 words)
We turned our own product page into a Compound package this week. Same process a client gets: the video, the vertical cut, infographics, copy for six platforms. The part I care about most is in the middle of the film: one article went from 795 to 17,449 search impressions, and we still say other work ran too. Full breakdown in the comments.
Instagram 79 words
39 episodes. Zero weeks skipped.
Most good writing gets one week of attention. Then it goes quiet.
We rebuild one published article into ten files:
→ a narrated video and a vertical cut
→ two infographics and three thumbnails
→ copy for six platforms
For SeenLabs, 95.4% of the accounts the episodes reached on Instagram were not following them.
Made from our own product page, the same way. Full breakdown on our site.
#contentmarketing #b2bmarketing #contentstrategy #videomarketing #repurposing
---
Facebook 172 words
You publish something you are proud of. It does well for a few days, then it goes quiet, and a month later it is a link nobody opens.
The writing was not the problem. The problem is that an article on one page has a small chance of reaching anyone who is not already looking for it.
So we take one published article and rebuild it into ten files: a narrated video, a vertical cut for Reels and Shorts, infographics, thumbnails, and copy written for each platform. A person checks every file against the source.
For SeenLabs we did that 39 times, about two a week for five months. 99.8% of the 6,100 people the episodes reached on Facebook were not following the page. On YouTube, 99.6% of watch time came from non-subscribers.
Those numbers measure how far the work travelled, not website visits. We keep the two apart.
The video below was made from our own product page, the same way.
Start with five of your own articles: https://fractalic.studio/products/compound/
#ContentMarketing
---
X 41 words
99.6% of YouTube watch time on 39 episodes came from people who weren't subscribed.
One article in, ten files out. Here is the film we made from our own product page.
https://fractalic.studio/products/compound/
(192 characters with the link counted as 23)
---
Google Business Profile 65 words
One article went from 795 to 17,449 search impressions in the 151 days after we rebuilt it. Compound turns one published article into ten files: a narrated video, a vertical cut, infographics, thumbnails and copy for six platforms, each checked by a person. Other work ran on that site too, and we say so. Full breakdown below.
Button: Learn more, https://fractalic.studio/products/compound/
Image: infographic-landscape or cover.
YouTube Shorts 47 words
Title (35 chars)
One Article, Ten Files, 39 Episodes
Description
Most good writing gets one week of attention. We rebuild one published article into ten files. For SeenLabs that meant 39 episodes in five months and 99.6% of YouTube watch time from non-subscribers. Full breakdown: https://fractalic.studio/products/compound/
---
YouTube, long cut 325 words
Title option A (problem, 44 chars)
Most Good Writing Gets One Week of Attention
Title option B (mechanism, 39 chars)
One Article In, Ten Files Out: Compound
Title option C (number, 41 chars)
39 Episodes, 99.6% Strangers: What We Saw
Recommended: A. It names the problem the viewer already has; B and C work better as the Short title and the LinkedIn hook.
Description
Your best article got one good week. Then it became a URL nobody opens twice.
Compound takes one article you already published and rebuilds it as ten files, each made for the place it will live: a narrated video, a vertical cut, subtitles, a transcript, two infographics, three thumbnails, copy for six platforms, schema and a cover. A person checks every file against the source before it ships.
→ The SeenLabs run: 39 episodes, about two a week, five months, zero weeks skipped.
→ Who watched: 99.6% of YouTube watch time came from people who weren't subscribed.
→ One article went from 795 to 17,449 search impressions, 151 days before and after the rebuild. Other work ran on that site in the same months, so we do not credit the videos alone.
This video was made from our own product page, with the same process.
Start with five of your articles: https://fractalic.studio/products/compound/
Chapters
0:00 One week of attention
0:37 How people look for answers now
1:25 One package, ten files
1:58 The SeenLabs run
2:40 Strangers, not followers
3:28 One article, before and after
3:40 What we do not claim
4:09 How it works
5:06 Start with five
Tags
content repurposing, B2B content marketing, video from blog posts, content marketing strategy, AI search visibility, repurpose articles into video, done for you content, content distribution, blog to video, YouTube for B2B, Fractalic Studio, Compound
Subtitles: upload long.srt as the English caption track (Subtitles, Add language, English, Upload file, With timing). The burned-in subtitles are pixels; the SRT gives YouTube the words.
---
JSON-LD schema
VideoObject, in the source of this page{
"@type": "VideoObject",
"name": "Most Good Writing Gets One Week of Attention",
"uploadDate": "2026-09-28",
"duration": "PT5M28S",
"contentUrl": "https://www.youtube.com/watch?v=2KKthgorwCI",
"embedUrl": "https://www.youtube.com/embed/2KKthgorwCI",
"hasPart": [ 9 chapter clips ]
}The full block, with the transcript, sits in this page's code. Search engines and AI crawlers read it without running any script.
Episode cover
1920x1080
How we know it works
SeenLabs is our founder's own company: a digital signage business Vahagn Ter-Sarkisyan founded in 2015, with a library of articles that were genuinely useful and almost never read twice. We went through it with them, picked the pieces worth the effort, and started turning them into episodes.
39
episodes produced and published, every one live and clickable today
~2
per week, sustained for five months
0
weeks skipped between the first episode and the last
Strangers, not followers
Share of the people who saw an episode who were not following the account at the time.
YouTube
99.6%
of watch time came from people who weren't subscribed
Facebook
99.8%
of 6,100 people reached weren't followers
Instagram
95.4%
of accounts reached weren't followers
LinkedIn
52%
of impressions landed outside the existing network
LinkedIn sits lower because that account has a real audience behind it, so getting past it is a harder thing to do. We left it in rather than showing you only the three comfortable numbers. All of this measures how far the work travelled, not how many people came to the website afterwards.
What happened to one of the articles
This one had been sitting quietly for months. We measured the 151 days before we rebuilt it and the 151 days after.
151 days before
795
151 days after
17,449
search impressions
Read the article this happened to →
We show both ends rather than the multiple, because the small number at the left is the part that makes the right one mean anything. A second article started picking up search clicks it had never had before, fifty of them over the same months.
We ran other work on that site during the same five months. Internal linking, technical fixes, a few other things. That means we cannot honestly hand you a single number and say the videos caused it, and we are not going to pretend otherwise.
The part most people leave out
So here is what we can actually put in front of you, and what we can't.
What we can show you
- Thirty-nine finished packages, all of them still online and open to anyone
- An article that went from 795 search impressions to 17,449, and a second one that started earning clicks it never had
- How far each episode travelled past the existing audience, platform by platform
- A whole package, file by file, before you spend anything
- The account screenshots behind every number here, including the platforms where nothing much happened
What we won't claim
- That the videos alone caused the search results. Too much else was running at the same time
- That every channel worked. Some did very little, and we will tell you which ones
- A guaranteed ranking or a guaranteed mention in an AI answer. Nobody can promise that honestly
- That any of this is automatic. Each package is built for its subject and checked by a person before it ships
When someone shows you a clean percentage for this kind of work, ask how many pieces it covers and what else was running that quarter. We would rather run it on five of your own articles and look at your numbers with you.
How it works
Pick
We go through your library together and select the pieces worth multiplying. Not all of them, the ones that will carry.
Build the base
A knowledge base for your brand: voice, the claims you can and cannot make, visual rules, the facts we're allowed to assert.
Produce
One package at a time, engineered per topic with its own prompts. Human fact-checked and brand-approved before anything ships.
Hand over and measure
We hand over the finished files, then look at the numbers with you against the baseline we set at the start.
How it is priced
One price per package, whatever the topic. You start with five, and the first five are where we learn your voice.
Your first batch
$2,450 for five packages
Five packages at $490 each, paid upfront. This is where we build your brand base: your voice, the claims you can and cannot make, your visual rules, the facts we are allowed to assert.
You get two rounds of revisions on each of the five packages. That is what they are for: getting the base right, so the packages after it need none.
Every package after that
$490 per package
Delivered final, at the standard pace of up to four packages a week. No revision rounds: by now the base is calibrated, and each package is checked against it before it ships.
Faster pace and monthly billing
When you need it faster
$686 per package
Up to ten packages a week, for when you need to build a library of video and visuals quickly. The price is 40% above standard.
Monthly, for high-volume blogs
$490 per package, billed monthly
For teams whose blog publishes 24 or more articles a month. You keep a reserved place in our weekly queue, and each month we agree with you how many packages to make and from which articles. Above four a week, the faster price applies.
Your price, before the call
What one package includes
| Input | One asset you have already published, usually an article. |
|---|---|
| Output | Ten files: narrated long-form video with burned-in subtitles, 9:16 vertical cut, SRT subtitle file, plain-text transcript, landscape and portrait infographics, three thumbnail variants, platform-native copy for six platforms, JSON-LD schema, episode cover. |
| Checked by | A person, against your source and your brand base, before it ships. |
| Pace | Up to four packages a week. Up to ten at the faster price. |
| Revisions | Two rounds on each of your first five packages. Delivered final after that. |
| Delivered as | Files, handed over to your team. |
| You provide | The source, your brand assets, and one person who approves. |
| Quoted separately | CMS upload, hub pages, an authority hub on your own subdomain, publishing to your accounts. |
Webinars, podcasts and other long recordings are priced on the call. Prices in USD, taxes additional. Offer details are confirmed in the order.
Who this fits
A fit if
- You have a real library, a dozen or more pieces worth multiplying
- Someone owns content or marketing on your side
- You sell something considered, where authority shortens the sales cycle
- You want a system that runs monthly, not a one-off burst
Not a fit if
- You only need one or two pieces. The first batch is five, because that is how long it takes to get your voice right
- You want it fully automated with nobody reviewing output
- You're looking for a guaranteed ranking or traffic number
- Your library is thin. Fix that first, we'll say so
Questions we get asked
How much does Compound cost?
$490 per package. One package turns one published asset into ten files. You start with a
batch of five packages, $2,450 paid upfront, which includes building your brand base and two
rounds of revisions on each of the five. After that every package is $490, delivered final.
Up to ten packages a week costs $686 per package.
What is in one package?
A narrated long-form video with burned-in subtitles, a 9:16 vertical cut, an SRT subtitle
file, a plain-text transcript, a landscape and a portrait infographic, three thumbnail
variants, platform-native copy for six platforms, a JSON-LD schema block and an episode
cover. A person checks every file before it ships.
How fast do you deliver?
Up to four packages a week at the standard price. If you need a library built quickly, we
go up to ten a week at $686 per package, 40% above standard.
How many revisions are included?
Two rounds on each of your first five packages. Those rounds are how we calibrate your
voice, your claims and your visual rules. After that, packages are delivered final.
Do you publish the content for us?
Not within the $490. The package is the set of files, handed to your team. Uploading to
your CMS, building hub pages, running an authority hub on your own subdomain and publishing
to your accounts are white glove work, quoted separately.
Is there a monthly plan?
Yes, for teams whose blog publishes 24 or more articles a month. It is $490 per package,
billed monthly, with a reserved place in our weekly queue. We agree each month how many
packages to make and from which articles. Above four a week, the faster price applies.
Why not do this in-house with AI tools?
You can, and some teams should. The tools each cover one step: one makes a video, another
cuts clips, another drafts posts. None of them checks the facts against your source, writes
the schema, or assembles ten files that agree with each other. The cost is in the hours
between the tools and in the mistakes you make while learning them. We made fourteen over
five months and wrote every one down.
How does the price compare with agencies and freelancers?
Agencies that repurpose webinars and podcasts usually sell monthly retainers starting around
$3,000 a month, priced on a call. Bought from a freelancer, a narrated video of this length
typically costs more than our whole package. Doing it yourself with AI tools costs less in
cash and more in your team's hours.
What can you start from?
A published article is the easiest input. Whitepapers and PDFs, case studies, product documentation, sales decks and onboarding material work too, and so does the explanation your best rep gives every week. Webinars, podcasts, conference talks, recorded customer calls and other long recordings are priced on the call.
Do you guarantee rankings or mentions in AI answers?
No. Nobody can honestly promise a ranking or a mention in an AI answer. We agree before we
start what we are going to watch, and then look at your numbers with you.
Start here
Start with five articles
Pick the pieces you were proudest of and watched disappear. Five packages, $2,450. We rebuild them, we agree beforehand on what we are watching, and then you look at your own numbers and decide whether this becomes a habit.
Calendar not loading? Open it in a new tab or write to we@fractalic.studio.