A social activist

AI content engine: up to 800,000 impressions

A social activist whose message landed in every room and nowhere online. I wrote the platform strategy, then built an AI engine that turns one rough video into the week's work: finished videos, the blog post, and the whole marketing flow for the idea it argues. Monthly impressions went from a few thousand to a peak of 800,000, and one income stream became four.

Client
A social activist Name protected
My role
Strategist and builder
What I built
The platform strategy, the AI content engine, the video pipeline, and the handoff to their content lead
Results
A peak of 800,000 impressions in a month, up from a few thousand, across eight platforms. One income stream became four: books, services, speaking and courses.

The room was full and the internet was empty

In person, the work landed. Online, the same message reached a few thousand people a month across every platform combined.

Content went out weekly-ish, made by hand by an assistant and the activist themselves. Every video, blog post, newsletter and digital event came one at a time, and each one cost money and hours. A talk that took days to prepare was seen by the room and then by nobody.

That reads as a marketing problem and it is not. The message was already good. Nothing carried it, and nothing turned attention into income.

What I did

I measured before I wrote anything. I went through what had already been published and sorted it by what travelled and what died. The pattern set the strategy, not my taste.

The strategy document, 25 pages. Platform by platform. Not "post more". A weekly cadence, a content-pillar map, a measurement frame, and the exact experiments to run in the first 90 days. I wrote it so their content lead could run it without supervision. It became the operating system their content lead still runs.

The bet. The work went out on eight platforms: LinkedIn, YouTube, Facebook, Instagram, TikTok, X, Threads and Bluesky. I did not treat them equally. I picked two to overinvest in and one to underinvest in, on purpose. The two got the weekly cadence, because that is where the people who buy this work were already reading. Production time went where it converted.

The video pipeline. The activist records something quickly, with no editing and no crew. The engine takes that raw video and returns a finished one, ready to publish. I started from an open source video tool and extended it well past what it does out of the box, so each video is assembled in code instead of by hand in an editor. The same source video also produces the blog post and the rest of the marketing flow for the idea it argues. That means the posts, the newsletter, and the copy that sells the product or event behind it.

Their words, not a model's. The engine drafts from what the activist has already published and said, so the language stays theirs.

The production line. A weekly queue, a review step, and a publishing window per platform. The cadence held because the week's work was ready before it was due.

The handoff. I trained their content lead on the system and stayed on retainer. Not "I am the agency now". The work is "you have the system, and I am the second brain when the platforms change". Algorithm shifts, new platform launches, policy changes: that is the retainer.

The part most content work skips

Reach is not the product. Revenue is. I built the engine so each piece of content pointed at something the activist sells. One recording can carry an idea all the way from a clip to a page that asks for the sale.

That is what changed the business. Before, the income came from one stream. After, four: books, services, speaking and courses.

What changed

  • Monthly impressions grew from a few thousand a month to a peak of 800,000 in a month.
  • Eight platforms carried the work, and two carried the weekly cadence. The two I backed outperformed the one I did not, as planned.
  • One income stream became four: books, services, speaking, courses.
  • More content at a higher quality, with less money and fewer hours going into producing it.
  • A 25-page strategy sits with their content lead, who runs it without supervision.
  • I stayed on retainer for the platform and algorithm shifts that come next.

Questions buyers ask

What is an AI content engine? A system that turns one piece of source material into the week's work: the finished video, the posts, the blog, the newsletter, and the copy that sells whatever the idea points at. A person approves each piece.

Will it sound like me? It drafts from what you have already published, so the phrasing stays yours.

Do I need a studio? No. Here the input is an unedited video recorded quickly, and the pipeline does the rest.

If your message is stuck in the room

You have the audience, the message and the demand. What you do not have is the team to build the machine that carries it. I build it: strategy, design, code, security and launch, all mine. Then I train the person who runs it and stay on retainer for the platform shifts that follow.

Product building