Short-form content agencies are dead. Are they?

Four videos, near enough the same title, going back as far as 2023. "Short Form Content Agencies Are Dead." "I am shutting down my short-form content agency!" "i quit my short-form content agency...what now?" "Why Short Form Content Agency Don't Work Anymore?"

And a search, "short form video company that failed", with people typing it in.

The written version is thin on the ground. So this is it, from someone still inside it rather than filming the exit. So this is it, from someone still inside it rather than filming the exit.

Two of those four are probably right about the business they were running. That is the complication, and I am not going to pretend otherwise for the sake of a tidier post. The shape of agency they are describing does not work, and I would have quit it too.

The argument is about what "it" is.

Play 1: read what the quitting videos are describing

Watch them and the same business shows up each time.

A client sends footage. The agency delivers edited videos, priced per video or as a monthly bundle of videos.

The client posts them. Nobody agreed on what happens next, because the deliverable was the file.

That business had one defence, which was that good short-form editing was scarce. Hard captions, punch-ins, sound design, a hook cut tight in the first second. Three years ago that combination was rare enough to charge for.

It is not rare now. So the defence is gone, and with it the business.

Play 2: name what broke, in order

Three things broke, and they broke in sequence rather than all at once.

The style got commoditised. What was a skill became a template, then a preset pack, then the default look of the whole platform. A portfolio full of it stopped distinguishing anybody, because every portfolio has it.

AI took the first pass. Auto-captioning, auto-reframing to vertical and automatic selection of likely moments now sit inside the editing tools and the platforms themselves. That is most of the volume work in a clipping job, and it went from billable hours to a button.

Then the price collapsed. Identical deliverable, global supply, and a client who genuinely cannot tell two showreels apart. Price settles at the cost base of whoever is cheapest, and if you are competing on a file, that is who you are competing with.

None of that is a market disappearing. It is a service becoming abundant, which is the normal fate of anything that can be described precisely enough to outsource.

Play 3: stop selling the keystrokes

Editing is not the product.

I mean that literally and it took me too long to act on it. The editing is the visible part, which makes it the part clients ask the price of, and the part an agency instinctively organises itself around. But the value sits in the decision before the edit: which 40 seconds out of a two hour recording are worth cutting, and where the hook goes.

Get that decision right with a mediocre edit and it travels. Get it wrong with a beautiful edit and nothing happens, and the client has a folder of beautiful videos and no reason to renew.

That decision is what VALORAE Arc sells. The edit comes with it, but the edit is not the thing being bought, and pricing it as though it were is how an agency ends up competing against automation on automation's terms.

Play 4: own a number the client can check

The per-video agency has a structural problem that nobody names. It has no number.

When the invoice is per video, the only thing the client can measure is whether videos arrived. Videos arriving is not an outcome, so the renewal conversation has nothing in it except price, and price only moves one direction.

An agency that owns a number has a different conversation. Videos shipped against a cadence, accounts posted to and managed, view and retention movement tracked across months rather than screenshotted from the good week. Arc has shipped 500 plus videos and passed 10 million views across them, and the reason I can put that in a post is that somebody counted it from the start.

The number does not have to be flattering. It has to be checkable.

Play 5: build the supply side before you need it

The second structural problem is hiring, and it is the one that ends most of them. Delivery quality is what breaks first when an agency grows, because the standard lived in one person's head and never got written down.

Finding a short-form editor is easy. Finding one who makes the same selection call you would, repeatedly, without supervision, is the whole business.

VALORAE Cast exists for that. It trains editors and clippers and routes paid work to them, which means the supply side is a community that already exists rather than a job post sent out in a panic when a client signs. Building it before it was needed was the single highest-leverage thing I did, and I did it for the wrong reason, which was that I enjoyed it.

Play 6: run more than one thing off the same back office

A single-service agency carries the full cost of its own back office. Contracts, invoicing, onboarding, the CRM, the people who answer clients.

The VALORAE group runs Arc for clipping, VALORAE Media for social page management, and Cast for the editor and clipper side, with NovaTechRay as an alias agency in AI search visibility. The full map is here.

That is not diversification for its own sake. It is the same back office, the same editors, and the same distribution knowledge serving five doors instead of one, which changes what a slow month means.

It also splits attention, which is a real cost and the reason I would not recommend this to someone with one client and no systems. Margin in the early years is not income, it is the thing that pays for the next hire, and where that margin comes from decides whether you can afford a second door at all.

Play 7: what would make me post the quitting video

The honest test of any argument like this is what would falsify it.

Three things would.

If clients stop being able to tell a well-selected clip from an automatically selected one, judgment stops being sellable and this becomes a software business. If platform economics shift so far that distribution is worthless without paid spend, the outcome I sell disappears.

And if the editor supply gets so good and so cheap that a client can hire directly with no quality risk, the agency layer has no reason to exist.

I do not know whether any of those happens. What I know is that none of the four videos in that search is describing them. They are describing a per-video shop with no number, no supply side and no pricing power, and that business was already dying when the market was easy.

Same shape as the SMMA argument, and the same answer. The service did not die. The version you could start without being able to deliver anything did.

The short version

1. Stop selling files per video, because that is the part that got automated.
2. Sell the selection and the hook, which is the part that did not.
3. Own one number the client can verify, and count it from day one.
4. Build the editor supply before a client forces you to.
5. Spread one back office across more than one service if you can carry the attention cost.
6. Treat early margin as hiring budget, not income.
7. Write down what would prove you wrong, and go look at it every quarter.

The group I run all of this through is VALORAE, and the thresholds that keep moving underneath it are worth watching more closely than the quitting videos.