YouTube Workflow
A production process for long-form video that survives contact with a normal working week.
Video is the format where an unsystematic approach shows most, because every stage depends on the one before it.
A weak script can't be rescued in the edit. A missing B-roll shot means reshooting or cutting. An unplanned thumbnail means the video is finished and unpublishable.
A process doesn't make the video better. It stops the avoidable problems from consuming the time you'd have spent making it better.
Read the editorial process · Request Your Free Systems Audit
The stages
Idea → Research → Script → Pre-production → Record → Edit → Package → Publish → Measure
Nine stages. Most people do idea, record, edit, publish, and wonder why it takes a full day per video.
1 · Idea
Same discipline as any content: a running list, captured without evaluation, each with the question it answers and who it's for.
Video-specific filter: does this benefit from being watched rather than read? A list of seven tips is better as an article. A process someone needs to see, a comparison, a walkthrough, an argument that needs you present — those earn video.
2 · Research
Time-boxed. Gather sources, examples, data, and what already exists on the topic.
Note the visual needs as you go: what will need to be shown on screen, not just said.
3 · Script
The stage that determines everything downstream, and the one most often skipped in favour of "I'll just talk."
Decide your format:
| Format | Good for | Cost |
|---|---|---|
| Full script | Dense information, precision | Slowest, risks sounding read |
| Beat outline | Most talking-head content | Needs fluency |
| Bullet notes | Interviews, reactions | Most editing time |
For most business content, a beat outline is right: each section's point written out, with the opening and closing fully scripted. The opening because the first fifteen seconds decide retention; the closing because improvised endings ramble.
Write the hook separately and deliberately. What's in the first fifteen seconds decides whether the rest gets watched. It isn't an introduction — it's a reason to stay.
4 · Pre-production
The checklist that prevents reshoots:
- B-roll shots listed, from the script
- Graphics and screen recordings identified
- Anything needing a screen capture recorded in advance
- Location and lighting set
- Audio tested — a short test recording, listened back
- Batteries, storage, settings confirmed
Audio is the one that matters. Viewers tolerate imperfect video and leave over bad audio. Test it every time, including the time you're sure it's fine.
5 · Record
Batch where you can. Setup is a fixed cost; recording three videos in one session costs far less than three times one.
Record the room tone — thirty seconds of silence. The editor will need it.
Keep a take log. Note good takes as you go. Finding them later costs more than noting them now.
Record B-roll immediately after, while the setup is up and the content is fresh.
6 · Edit
A defined sequence, so quality doesn't depend on mood:
- Rough cut — structure only, no polish
- Watch it through, cut everything that doesn't earn its place
- B-roll and graphics
- Audio: levels, noise, music
- Colour, if you do it
- Captions
- Final watch, start to finish, without stopping
Step 2 is where quality comes from. Most videos improve by being 20% shorter.
Step 7 is non-negotiable. Editors stop seeing their own work; one uninterrupted watch catches what scrubbing misses.
7 · Packaging
Title, thumbnail, description. As much of the outcome as the video itself.
Thumbnail: designed, not grabbed from the footage. Legible at small size. Three or fewer words if it has text.
Title: states the benefit or the question. Under 60 characters so it isn't truncated.
Description: the first two lines matter most — they show before the fold. Then timestamps, links, and the resources mentioned.
Timestamps. They improve the experience and they make the video searchable.
8 · Publish
- Title and thumbnail final
- Description with timestamps and links
- Tags and category set
- End screen and cards configured
- Captions uploaded and corrected
- Added to the right playlist
- Preview checked on a phone
- Scheduled or published
9 · Measure
| Measure | Tells you |
|---|---|
| Click-through rate | Whether title and thumbnail work |
| Average view duration | Whether the content holds |
| Retention curve shape | Where exactly people leave |
| Comments and questions | Your next video ideas |
The retention curve is the most useful thing available. A drop at thirty seconds is a hook problem. A steady decline is pacing. A cliff at four minutes is whatever happens at four minutes — go and watch it.
Batching
The single biggest efficiency available.
| Session | Work |
|---|---|
| Monday | Scripts for three videos |
| Tuesday | Record all three |
| Wednesday–Thursday | Edit |
| Friday | Package and schedule |
One week of work for three weeks of output. Compare with doing one video end to end each week, where you pay the setup cost three times.
What to automate
| Worth it | Leave manual |
|---|---|
| Publication checklist generation | Scripting |
| Distribution after publishing | Editing decisions |
| Pulling performance data into one place | Thumbnail design |
| Collecting comment questions into the idea list | Interpreting retention |
That last one in the first column is undervalued — comments are the most reliable source of video ideas, and they're easy to lose.
Realistic time
For a ten to fifteen minute talking-head video with B-roll:
Research 2h · Script 2h · Pre-production 1h · Record 1h · Edit 4–6h · Package 1h · Publish 0.5h
Roughly twelve to fourteen hours. Batching three brings it to about eight per video.
Anyone quoting two hours is either very practised, producing something simpler, or not counting.
Audio quality, the first fifteen seconds, and the thumbnail. Get those three right and the rest is refinement.
Want a content operation that ships consistently?
A free systems audit covers your production process and where the time goes.