From ¥2,000 to ¥20,000 a video: three years of video editing, and the turning point of "direct sales after the director quit"

There’s plenty of information out there about turning video editing into a side hustle or freelance career, but usually only the post-success numbers get cut out and shared. The record covered here is the opposite: the client relationships that didn’t work out are listed side by side, coded not by real name but as “Client B,” “Client C,” and so on — and every single one of them is shown falling apart along the way.

The writer goes by “Yume.” After four years as a new-graduate engineer with regularly 70+ hours of overtime a month, she took medical leave. Watching a YouTube video titled “How to Earn ¥100,000/Month Through Video Editing” during a stalled job search became her starting point. Three years later, her endpoint: one client, ¥20,000 per video, ¥200,000–300,000/month. We trace the substance of those three years through her own numbers.

Three years, in stages

On the first gig’s ¥5,000 rate, she writes that at the time she thought, “For a beginner, ¥5,000 is actually pretty good, isn’t it?” The eventual ¥20,000 rate is four times that.

The breakdown of the seven that fell through

The same article lists seven engagements besides Client A. Pulling out just the rates and how each ended paints a vivid picture of what this line of work actually looks like.

Of the seven, only Client B fell through over rate. The rest fell apart over the quality of the working relationship, sloppy instructions, the client’s own operational chaos, or communication that simply stopped. This list shows that the early problems in video editing aren’t only about pay.

The moment the trajectory changed

The turning point came in June 2023, when Client A’s director quit. Rather than wait for the gig itself to disappear, she pitched the client company directly for continuation, with the departing director’s blessing. The result was a direct contract, and the rate doubled from ¥5,000 to ¥10,000.

Line up the before-and-after numbers and the gap is stark. Before the direct contract, she was earning ¥5,000/video and, despite trying seven other clients, never landed on anything that lasted. After the direct contract, the rate rose in two steps (to ¥10,000, then ¥14,000, then ¥20,000) and she now earns ¥200,000–300,000/month from that one client alone. This moment is the turning point, precisely including the fact that it started with something outside her control: the intermediary director’s resignation.

Why the direct contract worked

On the surface, this reads as “the middleman’s cut disappeared, so the rate went up”, but that alone doesn’t explain the subsequent rise from ¥10,000 to ¥20,000. The mechanism breaks down into three parts.

The counterparty in negotiations changed. In crowdsourcing or director-mediated work, the client-side contact sets price on the assumption that the editor is interchangeable. In a direct contract, the client evaluates editing quality and turnaround directly. The reason “the workload increased” worked as grounds for a rate hike is that she was talking directly with the person who actually recognized that increase in workload.

Another editor’s departure became an opportunity for a raise. When a second editor quit, the work consolidated onto her, and the rate went to ¥14,000. If an intermediary had still been in the loop, that gap likely would have been filled by replacing the editor. Being directly connected meant the gap on the supply side became her own gain.

On top of that, her track record concentrated in one place. As the work scattered across clients B through H consolidated onto Client A, she became someone with a high cost of replacement in the client’s eyes. The factors she names last are “actually meeting deadlines,” “actually fixing mistakes,” and “actually responding to messages”. The same things she was already doing in her relationship with the director on the first gig. What made the difference was not the substance of her skill but whether she got directly connected to someone able to evaluate that skill.

The risk this record reveals

At the same time, her current setup is dependent on a single client. As the seven failed engagements show, the probability of a deal falling apart for reasons on the client’s side (a staff departure, a change of policy, communication going silent) is not low. She herself frames accepting that dependency as a deliberate choice: “a trusted relationship with a single client, over unstable engagements across multiple companies.” The flip side is that the moment this one client goes away, the ¥200,000–300,000/month reverts to zero.

The initial investment shouldn’t be ignored either. The roughly ¥200,000 for the PC and the Premiere Pro subscription came out before any income existed. That was funded by savings from her salaried years, for someone starting with no savings, alongside taking leave or quitting a job, there’d be no choice but to keep taking on work like Client B: ¥2,000 a video for six hours of labor.

How much of this can be replicated

What’s replicable is keeping a record of why each gig fell through and using it as input for future judgment calls, and preparing an original portfolio in advance rather than relying on a test project. Landing a ¥5,000 gig on her third application after registering isn’t unrelated to the fact that she could show a self-made piece she appeared in, filmed, and edited herself.

The conditions that are hard to replicate are equally clear. One is that the client worth pitching directly happened to be her very first gig. The other is the coincidence that the director quit and, on top of that, gave her blessing to reach out directly. Pitching a client without the director’s consent risks being seen as someone who broke the existing business relationship, likely ending it. Rather than a turning point you can engineer on purpose, the direct contract should be read as a question of whether you can act when the opportunity arrives.

One more point: her university experience editing video for a light-music club, and her background as an engineer with no hesitation around picking up new tools, both fed into how fast she was able to self-teach. This wasn’t three years starting from an absolute zero.

Related reading

• Until MENTA was sold to Lancers — a case examining, from the platform side, the conditions under which individual skill-based transactions succeed.
• Hitode’s blog — a long-term record of an individual media outlet where the numbers moved through continuity rather than dramatic turning points.