You wrote a tutorial outline with nine steps, hit record, fumbled step six, and now you’re staring at a four-minute file where 3:40 is perfect and 20 seconds is garbage. On a laptop you’d trim it. On a phone, with no editor in the loop, that whole take is dead. The fix isn’t better nerves, it’s a different answer to how to structure a tutorial in the first place.
Short answer: To structure a tutorial you record on an iPhone without editing, break the script into segments of 60 to 90 seconds that each end at a natural stopping point, and group those segments by which recording mode they need. DemoScope records in-app (with touch indicators and teleprompter) or in External PiP mode (system-wide face cam, no touch indicators), and switching modes mid-take isn’t possible.
Structure by capture unit, not by topic
A capture unit is the smallest chunk of your tutorial you can re-record without re-recording anything else. Traditional tutorial outlines split by topic (“Setup,” “First run,” “Troubleshooting”), which is fine for a written guide and terrible for phone recording, because a topic can run five minutes and a five-minute take is five minutes of retake risk.
Split by capture unit instead. A capture unit ends where the next segment can start cold: a fresh screen, a stated transition (“Next, the settings screen”), no half-finished sentence carrying over. Sixty to ninety seconds is the sweet spot. Long enough to say something complete, short enough that a flub costs you a minute instead of your afternoon.
The practical test: if you can’t describe where a segment begins and ends in one sentence without referencing the segment before it, it’s not a capture unit yet. Keep splitting.
Which recording mode does each segment need?
This is the question most tutorial outlines never ask, and it’s the one that decides your structure when you’re recording on a phone. DemoScope has two recording modes and they carry different feature sets, so a segment that needs touch indicators cannot live in the same take as a segment that needs you to leave the app.
| In-app recording mode | External PiP mode | |
|---|---|---|
| What it captures | Your screen inside DemoScope | Any app on your iPhone, system-wide |
| Face cam | Yes, draggable and resizable PiP bubble | Yes, floating PiP window |
| Touch indicators | Yes, customizable color | No (iOS limitation) |
| Teleprompter | Yes, auto-scrolling, not recorded | No |
| Available tools | Full set | Camera, PiP Border, Text on screen |
| Controls during recording | DemoScope HUD | System recording UI only |
So the structural rule falls out on its own: put every segment that depends on taps being visible, or on you reading a script, into in-app recording mode, and batch the “here’s the thing happening in another app” segments into External PiP mode. Then order your outline so you’re not ping-ponging between the two. Two mode blocks, not seven.
The trade-off is honest: External PiP mode is the rare capability here (almost no iOS apps do system-wide screen capture with a floating face cam), but you give up touch indicators and the teleprompter to get it. If your whole tutorial is finger-driven UI walkthrough, stay in-app for all of it and record the “other app” part as a separate short video. If your tutorial is fundamentally about a third-party app, accept that you’ll be narrating taps out loud instead of showing them.
A three-block structure that survives no-edit recording
Here’s the skeleton I actually use. It’s three blocks, and each block is a separate file you record and keep.
- Cold open, 15 to 25 seconds. State the outcome and the starting condition. “By the end of this you’ll have a shareable link. You need an account and about two minutes.” Record this in in-app mode with the teleprompter on, because opens are the segment people rewrite most and reading beats memorizing.
- The walkthrough, 3 to 6 capture units. One unit per screen or per decision point. Touch indicators on, because a viewer watching a phone tutorial on a phone genuinely cannot tell where you tapped otherwise.
- The close, 20 to 30 seconds. What just happened, what breaks, where to go next. Record it last, and record it twice back to back so you have a spare without setting up again.
If you want the reasoning behind why this order teaches better rather than just records better, that argument lives in how to structure a tutorial that actually teaches: a mobile creators guide. And if retention is your problem rather than clarity, how to structure a tutorial that keeps viewers engaged: mobile recording edition goes deeper on pacing inside each block.
The pre-record checklist for segment-based tutorials
Run this once before the first take, not before every segment. Five minutes of setup saves you the retake spiral.
- Script pasted into the teleprompter, split with blank lines where each capture unit ends, so you can see your own stopping points scrolling toward you
- Teleprompter scroll speed tested against one paragraph read aloud (too fast is the default mistake)
- Face cam bubble parked in a corner that doesn’t cover a button you’re about to tap, and sized so it isn’t eating a quarter of a vertical frame
- Camera shape chosen and left alone (circle, square, vertical rectangle, or horizontal rectangle) so all your segments match
- Touch indicator color set to something that contrasts with your app’s UI, not your app’s accent color
- Do Not Disturb on, notification banners off, low-storage warnings cleared
- Mic route confirmed: if AirPods connected halfway through your last session, your audio level will jump mid-tutorial and you cannot fix that without re-recording
That last one is the failure I hit most. Bluetooth handoff mid-take is silent, invisible, and only obvious on playback.
When this structure is the wrong call
If your tutorial genuinely needs to be one continuous 12-minute flow with no seams, segment-based structure fights you, and you should plan for a desktop editing workflow instead. DemoScope has no trimming, no cutting, and no filters. It records and exports MP4 to your camera roll. That’s the deal, and it’s why the structure work has to happen before you press record rather than after.
Same goes for anything needing captions or subtitles. There’s no transcription in DemoScope, so if accessibility captions are a requirement, budget for a separate captioning step somewhere else in your pipeline.
Ordering segments for the outcome you want
Structure isn’t only about recordability. A tutorial built to teach a concept orders segments by dependency; a tutorial built to drive installs front-loads the payoff. Tutorial structure: the framework mobile creators actually need in 2026 covers the dependency-first version, and if the goal is conversion, tutorial structure: the 5-part framework that actually gets your app downloaded lays out the ordering that puts the “wow” inside the first 20 seconds.
My position: pick one goal per tutorial. Videos that try to teach thoroughly and convert aggressively do neither, and on mobile, where you can’t cut, the compromise gets baked into the file permanently.
Frequently asked questions
How long should each section of a tutorial video be?
Sixty to ninety seconds per section is a good target when you’re recording on a phone with no editing step. That length is long enough to complete a thought and short enough that a mistake costs you one short retake instead of an entire take.
Can I record a tutorial that covers multiple apps on my iPhone?
Yes. DemoScope’s External PiP mode activates a floating face cam window, sends DemoScope to the background, and records your screen across any app on your iPhone using an iOS Broadcast Extension. The limitation to plan around: External PiP mode has no touch indicators and no teleprompter, and your tools are Camera, PiP Border, and Text on screen.
Do I need to edit my tutorial videos afterward?
Not if you structure the tutorial into separate short takes before recording. DemoScope does not include video editing (no trimming, cutting, or filters); it exports MP4 files to your camera roll, so the segmenting decision replaces the editing pass rather than postponing it.
What does DemoScope cost?
DemoScope Pro is $19.99 as a one-time purchase, not a subscription. The free tier records with a watermark, and Pro removes the watermark and unlocks all features. It runs on iPhone and iPad on iOS 14 or later.
Is the teleprompter visible in the recorded video?
No. The DemoScope teleprompter overlays your script on screen for you only and does not appear in the exported recording. It auto-scrolls at an adjustable speed, and it’s available in in-app recording mode, not in External PiP mode.
If your tutorials keep dying at minute four because of a single stumble, rebuild the outline into capture units first and pick your recording mode per block. DemoScope is worth a look if you want touch indicators and a teleprompter on the segments that need them, and a system-wide face cam for the ones that don’t.