On-screen text that answers a question
Text on screen has one job: to leave nobody holding a question.
There is a moment in most tutorials where the audience quietly divides. Somebody says a word, or does a step without naming it, or moves on from a number that was on screen for a second and a half. Half the viewers follow. The other half are now a step behind, and from that point on they are watching rather than learning.
The person who cannot see that moment is you. Not because you were careless. Because you know the answer, and knowing the answer makes the gap invisible.
Text on screen is the cheapest instrument for closing those gaps, and it is almost always used for something else.
Why can you not see where your own viewers get lost?
Because you compressed the steps years ago and cannot uncompress them on demand.
When you learned your craft, the thing you now do in one movement was four things. Over time those four collapsed into one, the intermediate steps stopped having names in your head, and the whole sequence became something you do rather than something you describe. That collapse is what expertise is, and it is also why teaching is difficult.
So when you say just fold it back on itself, you are naming an action that contains three decisions. You made all three. You did not mention any of them, because to you they are not decisions any more.
This is worth naming plainly, because it gets mistaken for vagueness and the instinctive repair is to talk more. Talking more does not surface a step you cannot see. A method does.
How do you find the question marks?
Watch your own video with a pen, at normal speed, from the beginning, and mark the timecode every time one of three things happens.
A word arrives that you have never defined. Any word you would not have known five years before you started. It does not need a lecture. It needs three words on screen while you carry on talking.
A number goes past. Quantities, temperatures, measurements, settings, anything the viewer might have to hold in their head while doing something else. Spoken numbers evaporate. Written ones can be paused.
A step happens without a name. You did something and moved on. If you cannot say what that something is called, the viewer has no chance, and they now have no way to search for it.
Do this once with your own footage and it will feel unpleasant, which is a good sign. The list you end up with is your annotation plan, and it is specific to this video rather than to videos in general.
What should text on screen actually do?
Answer the question at the moment it forms, and then get out of the way.
That is the whole specification. A word appears when the word is used. A number stays on screen while the number is being acted on. A step gets its name at the point somebody would want to write it down. The text is not summarising the video, it is not reinforcing your point, and it is not there to keep the frame interesting. It is answering something.
Which gives you a test for every piece of text you are about to add. What question does this answer, and who is asking it? If there is no answer, the text is decoration, and decoration on a demonstration is a cost paid by the viewer.
When is text worse than nothing?
Three cases, and all three are common.
When it repeats what you are saying out loud. A whole sentence spoken and simultaneously written is two channels carrying one message, and the viewer has to reconcile them, which is usually wasted work. The exception is narrow: a single keyword, a name or a number held on screen earns its place when the material is complex and the viewer sets the pace, because there the text is labelling rather than duplicating. Captions are a different matter entirely, because they exist for people who cannot use the audio and for anybody watching with the sound off.
When it competes with the demonstration. If your hands are doing the important thing and a word is flying in from the side, the eye goes to the word. You have just moved attention off the lesson, at the exact moment the lesson was happening.
When it arrives with a flourish. An animated entrance takes time, and during that time the text is not readable and the frame is not calm. A word that appears, holds, and disappears does the job. The elaborate version mostly demonstrates that somebody owns the software.
Where should it sit in the frame?
Away from the hands, away from the bottom, and gone before the next question.
Put text where nothing is happening, which is usually the upper half or the side opposite whatever is being demonstrated. Keep it out of the lower strip, because on a phone the platform's own buttons and captions live there. Leave it up long enough for a slow reader to finish it twice, then take it away, because text that outstays the moment becomes part of the wallpaper and stops being read at all.
One more piece of restraint worth having: pick one look and keep it. A single typeface, one size for labels and one for headings, one position on the left and one on the right. Consistency here reads as care, and it is free, whereas a new style for every note reads as improvisation, which it is.
Go back through the last tutorial you published, pen in hand, and write down every timecode where a beginner would have stopped. Then compare that list with where you actually put text. Wherever the two lists fail to meet is the next version of the video.
Keep reading
This is a service, and the method is written down.
Everything above came out of doing the work rather than writing about it. If you want the method instead of the story, it runs in order on one page.
Start with a free audit
Tell us where your content is now. We will come back with what we would change and what result to expect.
A person reads the channel and writes the audit by hand: a considered read typically takes three working days. That is the usual shape, not a promised turnaround. We use these details only to reply to you: no lists, no lurking.
What you will get
A fit snapshot: where your channel stands, and whether we are a match.
Two to three opportunities: specific, prioritised, yours to keep.
A recommended next step, even if that step is not us.
The audit is free and commits you to nothing: nobody follows up with a call you did not ask for.