Short-Form Video Editor

The Unpopular View with Michael Brown is an independent podcast about foreign aid, global health, and development. We're hiring one person to grow our audience — and we know a fair amount about why it isn't growing.

The objective is subscribers and followers. Not views. We have views. Last month we drew roughly 10,800 on YouTube and 16,300 on Facebook. Those views produced 8 new YouTube subscribers and a net gain of 2 followers on Facebook. (other platforms include X, TikTok, Instagram, and multiple audio platforms like Apple, Spotify etc).

The reason is retention. Our clips are watched for about 8 seconds on average, whatever their length. Nobody stays long enough to decide they want more. People who already subscribe watch several minutes across a session; everyone else averages about 31 seconds. The show is (arguably!) good. The first eight seconds for clips are not. And there is limited linkage between clips and the longer reels produced in studio.

Right now our clips are selected by AI, which picks segments that make sense in context — meaning they open with setup instead of with the best line. (We also get one clip from the Studio we select that is 1' in length.) 

Your job is to be the human judgment that Opus Clips tool doesn't have, and then to turn the attention that earns into an audience we keep. We have been doing this, but NOT WELL ENOUGH!

Our hypothesis — and what we don't know

We'd rather be straight with you than pretend this is solved. Here's the working theory and how confident we are in each link:

• Better cuts produce higher retention. Fairly confident. Clip length correlates negatively with retention in our data: our two best clips are 21 and 26 seconds, holding 44% and 38%. Our worst are 55 and 105 seconds, holding 9% and 10%.
• Higher retention produces more reach. Assumed, not proven. It's how these platforms are understood to work, but it isn't yet visible in our numbers — our highest-reach posts have had our worst retention.
• Higher retention produces more subscribers. This is the weak link and we know it. Subscribers watch far longer than non-subscribers, but that's probably because they're already fans, not evidence that watching longer creates fans.

Your job is not to execute a plan we're certain about. It's to run this experiment properly and tell us which links hold. If you prove link three is false, that's a genuinely valuable finding and we'll change strategy on it. That outcome is a success in this role, not a failure.

Two things our data already says

We may be clipping the wrong material. Our repeat viewers come almost entirely from the substantive aid and development episodes — USAID, “Can Aid Scale?”, “Is Aid a Scam?” The lighter geopolitical content brings far more traffic and almost nobody who returns. Which conversation gets clipped may matter more than where the clip starts.

Our strongest signal isn't in the clip data at all. Professionals in the foreign aid and conservation (and others in wellness)  world have started asking to come on the show unprompted — including the founder of a major humanitarian organization and an author published by HarperCollins. That suggests real authority with a specialized working audience, which may not be the same audience that scrolls short-form video. Part of this job is figuring out where our people actually are, and we're open to being told we're posting in the wrong rooms. And what to do about it…

Where we post, and why

LinkedIn is the priority. We (Michael Brown) have 11,115 followers there, most of them working in this sector, and recent posts have reached about 368 of them. Engagement is almost entirely reactions — 2 comments and 0 reposts in a week. Every inbound guest approach we've had came from LinkedIn. It is simultaneously our most valuable platform and our worst-performing one, and fixing that is the centre of this job.

YouTube is the archive. Full episodes, built to be found by search years from now rather than to win a Shorts feed.

Facebook gives us volume that doesn't convert. 16,300 views last month to an audience that's 83% US and skews 55+ — not the aid sector. We keep posting there because it's free, not because it's strategy.

Instagram, TikTok, X are syndication. Part of your first-quarter work is telling us which of these deserve any effort at all, and recommending we drop the rest.

What you'll own

• Re-cutting for the hook. Working in our Opus Clip account, deciding where a clip should actually start — usually mid-answer, on the strongest line — and cutting hard after it. Clips under 30 seconds hold nearly twice as well as clips over 45, so length discipline is part of the craft.
• Choosing what gets clipped. Which conversation, which ninety seconds, and what claim goes on screen in the first second.
• Turning viewers into followers. Calls to action, end screens, pinned comments, profile and bio optimization, and the handoff from clip to full episode. This only works once people are still watching when the ask arrives.
• Publishing natively on LinkedIn and YouTube first, with the same asset syndicated elsewhere. Native captions and formatting, not one file copy-pasted five times.
• Measurement and iteration. Weekly numbers, honestly read. Which cuts held, which converted, what changes next week.
• Engagement. Replying in comments, and building the habit of the show showing up in its own threads.

Realistic volume: 3–5 clips a week. We'd rather have three clips that hold attention than fifteen that don't.

What success looks like

The number we judge this role on is net new followers and subscribers. Everything below is a step toward it.

Days 1–30 — instrumentation and baseline. Working with our Web & Digital Systems Lead, get per-clip attribution in place so we can finally tell which clips produce subscribers. Establish per-platform baselines. Start cutting to 30 seconds or under.

Days 31–90 — fix the front door. Average retention above 40%, up from roughly 20%. One documented hook experiment per week: same content, different opening, measured. This is the gate — conversion work before this point is optimizing a door nobody reaches.

Days 91–180 — convert. Follows per 1,000 views above 5, up from under 1 today. Build and test the CTA layer. Get LinkedIn reach above 10% of followers, from roughly 3% now.

Through end of term — compound. Scale what works. Deliver a written playbook — which hooks work for this show, which platforms earn the effort, what converts — that someone else could run from cold.

Working with the team

You'll work most closely with our Web & Digital Systems Lead, who handles the site, integrations, and analytics plumbing. The division is clean: you decide what to measure and what to build — which links get tracked, what a landing page needs to do, what the CTA is and where it sends people. They build it. Your attribution setup is their first-month deliverable, and their metadata work depends on your keyword judgment.

You'll also overlap with an Information & Research Lead, who is making our episode archive searchable. Once that exists, finding the right ninety seconds becomes a search rather than an afternoon of scrubbing.

This is a small team. Everyone's work touches everyone else's.

What we're looking for

Required

• You can edit short-form video and have work to show. CapCut, Premiere, DaVinci, Descript — the tool doesn't matter, the eye does.
• Hook instinct. You know why one opening stops a scroll and another doesn't, and you can explain it.
• You think past the view. A clip that gets watched and produces nothing is a failure.
• Willingness to read analytics and change your mind. Half this job is craft; half is looking at a retention graph and admitting the cut you liked didn't work.
• Comfort with contested subjects. The show covers foreign aid and development, and its guests disagree with each other and sometimes with the host. You needn't agree with anyone, but you should find the arguments interesting.

Helpful: short-form work of your own that performed; caption and subtitle workflows; Opus Clip or Descript; growing any account from a standing start; podcast or student media experience.

We are not looking for someone to schedule posts. The editing decisions are the growth strategy here.

Terms, stated plainly

This is an independent project with no revenue and no funding. There is no pay. It's structured for academic credit and we'll work with your school on the paperwork. Roughly 5–10 hours a week, remote, through May 31, 2027.

What you get instead: full analytics access, real authority over what gets made, a named credit, a reference, a documented before-and-after on a live channel, and proximity to guests worth knowing in this field.

First project

The Donnelly episode will be published alongside Lifeline: The Story of PEPFAR on or about October 15th, coordinated with HarperCollins. You'd build the short-form campaign around it — clip selection, release timing against the publication date, platform-specific posts. A real deadline with a publisher attached.

How to apply

Upload three things:

• Your resume.
• Work samples — links to 2–3 short-form videos you cut yourself. Personal, class, or client work all count. One line each on what you were going for.
• The exercise, as your additional document.

Watch any 2-3 clips from the Emily Brearly Segments found at  https://www.youtube.com/@TheUnpopularView_MichaelBrown. It underperformed. Then watch the source at https://www.youtube.com/watch?v=5OO5nhXXO_k&t=6s ("Can Aid Be Fixed"); https://www.youtube.com/watch?v=XgT1hmcYghw&t=1s ("Can Aid Scale?") or https://www.youtube.com/watch?v=6qPPIyEM9bc&t=2s ("is Foreign Aid a Scam?").

• Where would you start the clip instead, and why? What's the first spoken line?
• What would you title it?
• Someone watches your version all the way through and likes it. What do you want them to do next, and exactly how do you get them to do it?

Text is fine — there's no need to actually edit. We want the reasoning.

We read the exercise and work samples before the resume—applications reviewed on a rolling basis.

And as an FYI, there are several other interns along with the Podcaster on this team.