We give people seven choices when they generate a demo video: narration voice, language, whether to burn in captions, caption style, caption size, industry, and whether to animate the cursor.
Across roughly 900 completed demos, here is how often each of those choices was actually changed from the value it arrives with.
How many people change the default settings?
Almost none. Every single option is left alone by the large majority:
| Option | Default | Kept the default | Share |
|---|---|---|---|
| Cursor animation | off | 839 of 846 | 99.2% |
| Language | English | 908 of 965 | 94.1% |
| Caption style | classic | 840 of 896 | 93.8% |
| Caption size | medium | 835 of 896 | 93.2% |
| Captions burned in | on | 810 of 903 | 89.7% |
| Industry | auto-detect | 634 of 903 | 70.2% |
| Narration voice | onyx | 604 of 963 | 62.7% |
The most-exercised control on the entire form is the voice picker, and even that is left untouched by nearly two thirds of people.
Which setting do people actually engage with?
The voice, and only the voice. When someone does change it, the choices spread out across the catalogue rather than collapsing onto one alternative:
| Voice | Demos |
|---|---|
| onyx (default) | 604 |
| nova | 107 |
| shimmer | 77 |
| echo | 70 |
| fable | 51 |
| alloy | 44 |
(A further 10 demos used voices added to the catalogue too recently to have meaningful volume.)
That is a real preference being expressed — 359 people listened to something and decided it was not right for their product. It is the one place on the form where the option is earning its keep.
What does a 0.8% adoption rate tell you?
Seven people out of 846 turned on cursor animation.
We built that feature. It renders a moving cursor across the page during the walkthrough, it took real work, and it is switched on in fewer than one in a hundred demos. There is no version of that number that means "users just have not discovered it yet" — it sits on the same form as every other option, in the same list, styled the same way.
The honest read is that it solves a problem people do not have. A demo video is not a tutorial; nobody watching needs to see where the mouse went.
What should you take from this if you build software?
A default is a decision you are making for almost everyone. If 93.8% of demos use classic captions, then the caption style picker is not really offering a choice — it is a way for a small minority to opt out of a decision we already made. The value of the control is not the flexibility; it is the escape hatch.
Options are not free. Every one of those seven controls is a row on a form, a branch in the render pipeline, a column of state, and a thing to keep working. Six of the seven are exercised by under 11% of users, and one by under 1%.
Engagement concentrates where the output is subjective. Voice is the only setting where people have a real opinion, because it is the only one where the "right" answer depends on their brand rather than on general good practice. Caption size does not have a brand-dependent answer. Voice does.
Methodology and limits
Figures are from every completed demo in our production database as of 19 August 2026, not a sample. Denominators differ per setting because the settings were introduced at different times and older demos do not carry all of them — each row above states its own denominator. Percentages are rounded to one decimal place.
Two caveats worth stating:
- "Kept the default" is not the same as "was happy with the default." We cannot distinguish deliberate agreement from not noticing the control. Both are real, and this data does not separate them.
- We deliberately excluded one field. Whether a demo pauses for script review looks like a setting, but it is assigned by the system — always on after a user's first demo, and split by an experiment before that. Reporting it as a user preference would have been wrong, so it is not here.