Guide
Nobody is watching your first videos. That is the feature, not the problem
Every on-camera person you admire has a terrible first video, and they got better in public, in front of almost nobody, because early videos reach almost nobody. This page is the long game: what actually compounds across your first 100 videos, what to measure instead of views, the plateau that arrives around video 30, and the weekly ritual that turns raw reps into visible skill.
- Videos before you judge whether you are good at this
- 100
- Fix carried out of each weekly review. Never two
- 1
- To 100 videos at about 11 a month, Lite's pace
- ~9 mo
- Cost of a deleted take. Your miss rate is invisible
- 0
The skill curve and the audience curve are different curves
Skill on camera grows from video one. An audience arrives much later, if it arrives. Most early talking-head videos are seen by a few dozen people, many of whom know you.
New creators read that gap as failure. It is actually the best deal you will ever be offered: a full practice environment with a publish button, and consequences near zero. You get to be bad in public without the public.
The stand-up comedy version of this is the open mic: years of small rooms before anyone films a special. On-camera skill works the same way, except your open mic is invisible and the delete button exists.
By the time views arrive, if they do, the awkward reps are behind you, buried under a hundred better videos. Nobody scrolls back that far. Almost nobody scrolls back at all.
What the reps actually buy, roughly in order
The milestones below are what creators commonly report. Your numbers will differ. The order mostly will not.
- Around video 10: your normal voice shows up on camera and stays. The broadcaster impression retires.
- Around video 20: takes drop from six to two, because your mouth has learned the shapes of your own explanations.
- Around video 30: pace. You stop rambling mid-take because you can feel a tangent starting, which is a sense you cannot get any other way.
- Around video 50: you edit while talking. The bad sentence gets repeated correctly without breaking the take.
- Around video 70: topics arrive faster than you can film them, because making videos taught you to notice what people ask.
- Somewhere in the hundred: a stranger quotes your own line back at you. It will be from a video you almost did not post.
This is the whole editor
Highlight a phrase and a clip lands on those exact words. No timeline, no keyframes, no layers.
Measure delivery, not views, for the first stretch
Views measure the platform's mood, the topic, the hour you posted, and lastly you. As a feedback signal on your camera skill they are nearly noise, and treating them as a verdict is how people quit at video 15.
Track things you control instead. Did a video ship this week. How many takes did the keeper need. How many seconds before the first stumble. Whether you watched the final cut without wincing.
Those numbers move steadily while views jump around, and they are the honest scoreboard of the only thing the first 100 videos are for, which is getting good.
Look at retention and topic patterns from around video 30, when there is enough of a body of work for patterns to mean something. Before that, analytics are astrology with graphs.
The weekly review: one video, one fix
Once a week, watch one of that week's videos back. Full screen, sound on, no skipping, the way a viewer would.
Write down one fix. Only one. Swallowing the ends of sentences, looking away between thoughts, starting every video with the same three words. One sentence, on a list.
Carry that single fix into next week's filming and ignore everything else that is wrong. Fixing one thing at a time is slower per week and much faster per year, because each fix actually sticks. Fifty-two weeks of one-fixes is a transformation no course sells.
Keep the list. Reading month one's fixes in month six is the proof of progress that views refuse to give you.
The plateau at 30, and what the delete button is for
Around video 30, most people hit the wall: the videos feel samey, the views have not moved, and the novelty that powered the first month is gone. This is where the majority quit, which is worth knowing, because it reframes the wall as the filter.
The way through is boring and works. Keep the schedule. Change one variable at a time, a topic lane, an opening style, a length, and give each change five videos before judging it. Changing everything at once teaches you nothing about anything.
And keep using the delete button for takes, not for published videos. A bad take costs nothing and vanishes. A published video is a rep that counts, even at forty views. The miss rate that matters, the awful takes, stays invisible forever. The public record only ever shows your keepers.
Leave the old videos up. The visible arc from stiff to natural is not embarrassing. It is the most convincing thing about you, and future clients read it exactly that way.
Reps only compound if each rep is cheap
The thing that ends most 100-video runs is not nerve or talent. It is the hour of editing attached to every rep. At an evening per video, video 100 is four hundred hours away, and no schedule survives that arithmetic.
Cutroom cuts the rep cost to the take itself. Film up to 3 minutes, upload, and it transcribes every word, drops the ums, tightens the dead air, writes the hook, times the captions and lays b-roll on the words that need it. The finished 9:16 MP4 is back in about a minute, and you direct on the transcript, so the review-and-tweak takes minutes too.
A 30 second video is about 110 credits, a 100 credit batch plus export at 20 credits per finished minute. The first 3 are free in your first 7 days, no card, with a Cutroom mark across the middle. Lite is $19.99 a month for 1,250 credits, about 11 videos monthly, which reaches video 100 in roughly nine months. Cancel in one click whenever the habit no longer needs it.
The floor, honestly: a phone and the platforms' caption tools cost nothing, and at one post a week they can carry you. The software matters when the editing hour is the reason videos stop shipping.
Questions people ask
- How long does it take to get good on camera?
- Count videos, not months. Comfort tends to arrive around video 10, competence around 30, and something like ease around 100. At two to three videos a week that is roughly nine months to a year, and no amount of watching other people's videos substitutes for the reps.
- Why is nobody watching my videos?
- Because early videos reach almost nobody, for everyone, on every platform. It is not a verdict on you. Use the empty room: it is the only period where you can be visibly mediocre at zero cost, and it ends precisely when you no longer need it.
- Should I delete my old bad videos?
- Generally no. They are the visible proof of the arc, and viewers who binge your profile read improvement as credibility. Delete only what is factually wrong or professionally compromising. Nobody else is holding your early delivery against you.
- When should I start looking at analytics?
- Around video 30, once there is a body of work. Look at retention and which topics get finished, not raw views. Before that point, publish-rate and takes-per-keeper are better instruments, because they measure you rather than the algorithm's weather.
- What do I do when it feels samey and pointless?
- That is the video-30 plateau, and it filters out most people. Keep the schedule, change exactly one variable, give it five videos, then judge. The feeling is a stage, not a signal, and the creators you follow all went through it in front of nobody.
- What does editing 100 videos cost?
- By hand, roughly an evening each, which is the real reason most runs end early. Through Cutroom, about 110 credits per finished 30 second video: Lite at $19.99 a month covers about 11 monthly, reaching 100 in about nine months. Source takes cap at 3 minutes; exports are 9:16 MP4, and cancelling is one click.
Ship the rep, log the fix, ignore the views, survive video 30. The person at video 100 is unrecognisable, and nobody watched it happen. That was the point.