How to record a software tutorial on a Mac that people can follow along with.

The recording workflow

1. Define the result and the starting point

   Write one sentence: “After this video you can [do one thing].” Then list what the viewer needs before starting: app version, a sample file, an account level.

2. Split the lesson into checkpoints

   Three to six checkpoints, each ending in something the viewer can see on their own screen. If they cannot tell whether a step worked, it is not a checkpoint yet.

3. Script one sentence per action

   Name the control, say what you do, say what should happen. “Open the Data menu and choose Import. A file picker appears.” Avoid “click here”.

4. Set up the screen for small players

   Enlarge the interface, close everything unrelated, turn on Do Not Disturb, and record only the app window or a fixed 16:9 area.

5. Record checkpoint by checkpoint

   Press Shift-Command-5, choose your microphone in Options, turn on Show Mouse Clicks, and record. Pause for two seconds on every result before moving on.

6. Review the narration against the screen

   Watch once with sound for wrong or missing instructions. Fix a wrong step by re-recording that step only, not the lesson.

7. Follow your own tutorial

   Reset to the starting state and do exactly what the video says, nothing more. Any place you had to guess is a step to re-record.

How to structure a software tutorial with checkpoints

A tutorial that runs as one unbroken flow gives the viewer no way to notice they fell behind. Checkpoints fix that: each one ends with a result the viewer can compare with their own screen.

The example column uses a spreadsheet lesson, “build a monthly sales summary”.

A tutorial script skeleton

You do not need a word-for-word script. You need the order of actions fixed and the names of controls correct. Fill in this skeleton and keep it next to the screen while you record.

• Opening: “In this video you will [result]. You need [prerequisites].”
• Starting state: “Start with [file or screen] open, like this.”
• Each step: “[Open, choose, or type] [exact control name]. You should now see [result].”
• Each checkpoint: “Checkpoint: your screen should show [state]. If it does not, [most common cause].”
• Closing: “You now have [result]. To go further, [one next step].”

How to keep small UI details readable

Your recording monitor is the best case. Most viewers watch in a help center column or a phone, where a full desktop capture turns small labels into smudges.

• Record the app window or a fixed 16:9 area, not the whole display. Record Selected Portion does this on any current macOS; Record Selected Window arrives in macOS 26 Tahoe.
• Enlarge the interface before recording. In most browsers Command-Plus does it; many desktop apps have their own zoom or text size setting.
• Turn on Show Mouse Clicks in Shift-Command-5 > Options so each click is visible.
• Keep the pointer off the label you are talking about, and keep it still while you explain.
• Zoom in on fields the viewer must read. QuickTime Player’s editing features include no zoom; in Qreata, Auto-zoom on clicks (off by default) places zooms from your real clicks, Add zoom places one by hand, and each zoom’s duration, scale (1.2–4×) and focus point can be edited. Clicks can also be highlighted after recording.
• Test at delivery size: play ten seconds in a window about 700 pixels wide. If you cannot read a label, neither can the viewer.

Pacing, narration, and whether to show your face

Pacing has one rule: say it, do it, hold it. Announce the action, perform it, then leave the result on screen for about two seconds. Cut the time you spend waiting for pages to load. Keep the pauses after results, because that is when the viewer checks their own screen.

Narrate in the present tense and name controls exactly as they appear. If you say “Export” and the button reads “Download”, the viewer stops to hunt.

Show your face in the intro and the recap, where trust matters, and hide it during dense UI work, where it covers the interface. Leave it out of short help center clips that will be re-recorded whenever the product changes. Apple’s screen recording article describes no camera overlay for the built-in recorder; Qreata records the camera as its own track, so after the take you can keep it in the intro and recap, hide it on the segments with dense UI, or move and resize it for one fragment only. In Qreata a slow page load can also stay in the video at up to 4× speed instead of being cut.

Best format for a help center or YouTube tutorial

Export MP4 with H.264 unless your platform says otherwise. Apple itself labels its H.264 capture option “SDR (Most Compatible)”. Built-in recordings save as .mov; Qreata exports MP4 in H.264 at up to 4K.

The lengths below are our editorial guidance. Captions matter in both places, and there are two kinds. Burned-in captions are part of the picture and show in every player; Qreata adds them with Captions in video, word-highlighted, in one fixed style. Caption files (SRT or VTT) let the viewer switch captions on and off; Qreata does not export them, so create those in the destination platform after upload, or in a separate captioning tool.

Record a tutorial with the built-in Mac recorder, and where it ends

For a single short lesson, the tools in macOS are enough.

It ends at editing. Apple’s list of QuickTime Player features is trim, split into clips, rearrange, rotate or flip, add a clip, and remove audio or video. There is no zoom, no captions, and no way to re-record one step inside the lesson. On macOS 14, 15, and 26 the recorder also captures the microphone only, so a tutorial about software that makes sound needs macOS 27 or another tool. One or two lessons a quarter: stay here. A library you maintain as the product changes: the manual splitting adds up.

1. Press Shift-Command-5 and choose Record Selected Portion; drag the frame over your app.
2. Open Options: choose your microphone, turn on Show Mouse Clicks, and set a Timer if you want a delay to get the pointer in place.
3. Click Record and work through your checkpoints. Stop with the menu bar Stop button or Command-Control-Esc.
4. The file saves to the desktop as a .mov. Open it in QuickTime Player and use Edit > Trim to remove the start and end.
5. For a mistake in the middle, split the movie into clips around it and remove that clip, or record a replacement and add it as a clip.

Review narration and fix single steps in Qreata

Tutorial mistakes are usually small and verbal: you said “File menu” while opening Edit, or took three attempts at one sentence. Qreata is built around repairing exactly those: for the first, Record voiceover · keep video in the Segment menu lets you say the sentence again over the same picture.

Local transcription turns your narration into text on your Mac, so you can read a ten-minute lesson in a minute and spot the wrong instruction. Word-level cuts remove a false start by deleting the words, Remove silences takes out the dead air, Remove filler words takes out “um” and “uh” (words such as “like” or “you know” only if you choose), and Captions in video burns word-highlighted captions into the MP4. You can copy the transcript text as the draft of the written steps for the help article.

When a whole step is wrong, Replace Take (the Rerecord tool in the editor) re-records only that range. Say checkpoint 2 runs from 3:10 to 3:45 and you chose the wrong formula. Shift-drag across that range on the timeline, put the spreadsheet back as it was at 3:10, record the step again, and checkpoints 1 and 3 stay untouched. The range has to sit inside one segment and can’t cross a cut. A retake shorter than the 35 seconds is rejected, and a longer one is trimmed. Projects are non-destructive, so when the product’s UI changes next quarter you replace the affected step and export again.

Qreata runs on Apple Silicon Macs with macOS 14 or later and costs $49 once, with a 7-day trial. A fair ten-minute test: record a three-checkpoint lesson, read the transcript, cut one false start, turn on captions, and replace one step.

Questions about this workflow