Starting Your YouTube Channel
Course 1 · Chapter 6 · Video Ideation & Scripting
A video idea is not the same as a video. Between "I could make something about budget GPUs" and a finished, watchable video lies a process — turning a vague topic into a specific angle, building that angle into a structure, and deciding how you'll actually say what you need to say. This chapter covers how to do that efficiently, so the creative work stays creative instead of becoming a recurring source of dread.
Generating Ideas — Four Reliable Frameworks
Your idea bank from Chapter 5 holds your raw material. These frameworks help you sharpen vague topics into specific, compelling video concepts with a clear angle that differentiates them from what already exists.
Framework 1
The Audience Question Mine
Go to the comments on your competitors' top videos and find the questions viewers ask. Every unanswered comment is a video idea with a proven audience. Look specifically for questions that appear multiple times — that's a high-demand topic that the creator didn't address.
"Why does everyone recommend the 7600X over the 7700? Nobody ever explains the actual gaming difference." — that's a comparison video.
Framework 2
The Angle Pivot
Take a topic that already has well-performing videos and add a constraint, audience, or time dimension that makes it new.
Topic + constraint + audience = angle. The constraint is what differentiates you from the existing videos on the same subject.
"Best budget GPU" exists. "Best budget GPU for 1440p in 2025 under £200" is a specific angle with a defined audience and a clearer answer.
Framework 3
The Misconception Video
Find a belief that's widely held in your niche but is oversimplified, outdated, or just wrong. These videos perform well because they promise to correct something the viewer already believes — which creates curiosity and mild tension.
Disagreement is inherently watchable.
"More RAM doesn't make games faster" — debunks a common belief, forces the viewer to stay to find out when it actually does matter.
Framework 4
The Experience Report
Document something you actually did — a purchase, a build, a month-long experiment — and report the honest outcome. These videos are differentiated because they're personal and specific. They can't be copied by someone who didn't have the same experience.
First-person evidence beats general advice.
"I used a £50 GPU for 30 days — here's what actually happened." Specific, personal, built-in narrative arc.
Building the Hook
The hook is the first 30 seconds of your video. It is the single most important part of any video you make — more so than the quality of the information, the production value, or the length. A viewer who isn't hooked leaves. A viewer who leaves counts against you in the algorithm. Get this right before you worry about anything else.
A good hook does two things simultaneously: it confirms to the viewer that this video is for me, and it creates a reason to keep watching. It should never begin with an introduction, a channel recap, or thanks for watching. The hook is about the viewer, not about you.
Problem Hook
Open with the viewer's pain point
Name a specific frustration your target viewer is experiencing right now. If they feel seen, they stay. Works best for how-to and tutorial content where the viewer has a problem to solve.
"If you've been staring at GPU prices wondering why nothing in your budget actually performs well at 1440p — this video is for you."
Curiosity Hook
Tease the answer without giving it
Introduce the conclusion but withhold the explanation. The viewer must watch to resolve the open loop. Be careful: if the teased payoff doesn't deliver, this creates frustration and a poor satisfaction signal.
"I tested every budget GPU under £200. One of them is significantly better than everything else — and it's not the one reviewers keep recommending."
Promise Hook
State exactly what the viewer will know by the end
Clear, direct, outcome-focused. Works well for informational content where the viewer is outcome-driven rather than curious about the journey. Less exciting, but very high trust signal.
"By the end of this video you'll know exactly which GPU to buy for 1440p gaming on a tight budget, and why — without needing to read a single spec sheet."
Story Hook
Begin mid-action, then rewind
Drop the viewer into a specific moment — the result, the mistake, the discovery — then pull back to explain how you got there. Creates immediate narrative tension. Common in vlog and experience-report formats.
"Three weeks ago I turned on a PC that cost me £287 to build. It ran Cyberpunk at over 60fps. Here's how."
Contrast Hook
Open with a surprising contradiction
State something that contradicts conventional wisdom or the viewer's expectation. Forces a double-take. Best for misconception videos and opinion-driven content where you have a contrarian position to defend.
"The most common advice on this topic is completely wrong — and following it cost me £150 on parts I didn't need."
Video Structure — The Architecture of Retention
After the hook, you need to keep people watching. The underlying principle is simple: a viewer stays as long as they believe the best part of the video is still ahead of them. Structure your video so there's always something coming that they'd regret missing.
0:00 – 0:30
Hook
Open with the problem, promise, or story. No intro music, no "hey guys welcome back." The viewer's attention is at its highest here and drops every second. Use it.
Critical: if retention data shows a large drop in the first 30 seconds, the hook is broken — rewrite it before fixing anything else.
0:30 – 1:30
Context & credibility
Set up why this topic matters and why your perspective is worth hearing. Brief — one or two minutes maximum. This is also where you can introduce yourself, but keep it one sentence.
Optional: use a "here's what we'll cover" roadmap for longer videos — gives viewers confidence that the answer is coming.
1:30 – 11:00
Main content — with internal re-hooks
Deliver the substance. Break it into clear sections. Every 2–3 minutes, add a micro-hook: a teaser for what's coming next in the video ("but wait until you see the performance results"). This prevents viewers from feeling the video is finished before it is.
Each section should have a clear point — if you can't state what the viewer learned from a section in one sentence, it probably needs to be cut or restructured.
11:00 – 12:30
Conclusion & recommendation
Summarise what was covered and give a clear, direct answer or recommendation. Don't hedge with "it depends" unless you immediately follow it with "here's what it depends on and here's the answer for each scenario."
12:30 – 13:00
Call to action
One CTA only — subscribe, or a specific related video to watch next. Not both. Not "like, subscribe, comment, share, and check out my Patreon." One clear next step.
Linking to a related video outperforms a generic subscribe ask for most channels — viewers who watch two videos in a row are far more likely to subscribe than viewers prompted directly.
Scripted vs Conversational Delivery
There is no universally correct approach here. The right choice depends on your subject matter, your natural delivery style, and your production setup. Most creators land somewhere between the two extremes.
Fully Scripted
- Every word planned — no rambling, no filler
- Easier to edit — you know what was supposed to be said
- Works well for complex technical topics needing precision
- Reduces on-camera anxiety for new creators
- Risk of sounding robotic if you read rather than perform
- Takes longer to write upfront
- Spontaneous reactions and moments get lost
Bullet-Point Conversational
- Feels natural and authentic on camera
- Faster to prepare — outline only
- Better for personality-driven and opinion content
- Genuine reactions and enthusiasm come through
- Higher risk of rambling and repetition
- Harder to edit — harder to know what to cut
- Can miss key points without a full script
The most practical approach for most new creators: write a hybrid. Script the hook and conclusion word-for-word (these are the highest-leverage parts of the video), and use bullet points for the main content sections. This gives you precision where it matters and natural delivery in the middle.
What a script actually looks like
A script doesn't need to be a Word document formatted like a screenplay. It just needs to tell you what to say and what to do. Here's a practical format:
== HOOK ==
Three weeks ago I turned on a PC that cost me £287 to build.
It ran Cyberpunk at over 60fps at 1080p medium settings.
Today I'm going to show you exactly how — every part, every decision,
and every compromise I made to hit that number.
== CONTEXT (keep to 60 seconds) ==
[Brief: who this is for, what we'll cover]
This isn't a hypothetical build. These are parts I actually bought,
prices from last Tuesday, and real benchmark numbers from my testing.
== SECTION 1: The CPU Decision ==
[B-ROLL: CPU on table, benchmark screen recording]
ACTION: Show CPU packaging on camera
The biggest money-saving decision in this build was the CPU.
I went with the Ryzen 5 5600, not the 5600X...
[BULLET POINTS — expand naturally:]
· Why 5600 not 5600X (price delta vs gaming perf delta)
· What I'd upgrade to with another £30
· The one scenario where you do want the X variant
== CTA ==
If you found this useful, the next video I'd watch from this channel
is the GPU comparison I did last month — it answers the obvious
follow-up question about which card gets the most out of this CPU.
It's linked on screen now.
Where Viewers Drop Off — and How to Stop Them
Audience retention data (available in YouTube Studio once your video has views) shows you a curve of how many people are still watching at each moment. Understanding the common drop-off points lets you write scripts that avoid them.
Typical retention curve shape — where the losses happen and why
0:00 (hook)
First section ends
Mid-video
Near end
100% watched
0:00–0:30
Largest single drop — viewers who the thumbnail brought in but the opening didn't retain. Fix: stronger hook, immediate relevance signal, no intro.
Section ends
Viewers who got their question answered and leave. Fix: micro-hooks at section transitions — "before you go, there's one thing I haven't covered yet."
Mid-video
Pacing drop or a section that feels like filler. Fix: cut ruthlessly — if a segment doesn't advance the story or add new information, remove it.
After CTA
Normal — once the subscribe ask appears, some viewers leave. Acceptable. Putting genuinely useful content after the CTA can retain this group.
Practical exercise
Pick one idea from your idea bank and apply the full process: choose a hook type, write the hook word-for-word, outline the main sections with bullet points, and write the CTA sentence. That's your first script. Do not make it longer than one A4 page for a 10-minute video — if it doesn't fit, the video has too many ideas and needs to be split.
Next — Chapter 7: Understanding Analytics
Which numbers actually matter in the early days of your channel, which ones to ignore, how to read your retention curve, and how to use analytics to make smarter decisions about what to make next — without drowning in data that doesn't yet have enough signal to act on.