Who does Lorem Ipsum leave out?
I have filled thousands of text boxes with fake Latin and never once wondered what it sounds like out loud.
We have all used it. Drop a text block into a frame, fill it with a paragraph of fake Latin, get on with the layout. It is so normal that it does not register as a decision at all.
That is exactly why I keep looking at it. The defaults that do damage are rarely the ones we argue about in review. They are the ones we stopped seeing.
What actually happens when the room reads it?
Put a Lorem Ipsum prototype in front of eight people and seven of them will do something reasonable. They will skim past the gray blocks, register that the words are not the point, and give useful feedback on hierarchy and flow.
The eighth person is using a screen reader. What reaches them is not a gray block they can skim past. It is a voice reading Lorem ipsum dolor sit amet, consectetur adipiscing elit at their listening speed, in full, before anything useful arrives.
They cannot glance. Glancing is a visual affordance, and we built the entire convention on top of it without ever noticing we had.
So the likely outcome is that this person says less in that session than everyone else, and we write that down as a quiet participant rather than as a prototype that would not let them in. The exclusion never arrives as a complaint. It arrives as an absence in the data, which is the hardest kind to notice and the easiest kind to repeat next quarter.
Why does the habit survive?
Because everything around us affords it.
Figma offers it. Plugins generate it. Every tutorial we learned from used it. Don Norman's point about affordances applies to our own tools as much as to the things we make with them. A tool that makes one action effortless has made a decision for us, and we will mostly take it.
What our tools do not afford is the question. Nothing in the interface asks who is going to hear this. There is no friction at the moment the choice is made, and all the friction arrives later, attached to someone else's job title.
We both believe UX professionals do not intentionally exclude people from their work. Nobody reaches for placeholder text in order to shut a participant out. We reach for it because it is fast, because the tool suggests it, and because it is inherited, which is the shape almost every accessibility failure takes once you look closely.
Does it only matter in research sessions?
No, and this is the part I underestimated for years.
Placeholder text hides the things that break real interfaces. Latin has an even texture, no long words, no numbers, no names, no diacritics, and a length you chose deliberately. Real content has all of those, which is why layouts that looked comfortable in the file start truncating the week after launch.
It also hides meaning, which matters more. A label reading Lorem ipsum beside a field cannot be evaluated for clarity, and clarity is an accessibility property, for people with cognitive disabilities, for anyone reading in a second language, for anyone doing this while tired. A design cannot be tested for comprehension using text that has none.
What goes in the box instead?
The fix is not a checklist item at the end. It is a different default at the start.
- Write real sentences, badly. A rough honest sentence ("Your application was received on 3 March") tests more than a perfect paragraph of Latin ever will. It reveals length, tone, and whether the label above it makes sense.
- Use content that could plausibly be true. Real names of realistic length. Real error messages. Real empty states, which are where placeholder habits survive longest because nobody designs them twice.
- Read the screen aloud before the session. Not necessarily with a screen reader, just aloud. Anything unbearable to say is unbearable to hear.
- If you need volume, generate structure, not noise. Repeated realistic records beat a wall of Latin and surface truncation bugs while they are still cheap.
- Ask what the first thirty seconds sound like. For a sighted participant the question is odd. For everyone else it is the whole experience.
- Recruit for the session you claim to be running. A study that never includes assistive technology users is not measuring the product, it is measuring the product for people who work the way we do.
I built No Latin (opens in a new tab), a web app that turns content blocks into an accessible HTML prototype, out of this. But the tool is downstream of the habit, and the habit is the point. Any team can change the habit this afternoon with no tooling at all.
The wider pattern
Placeholder text is a small instance of something large. Our defaults encode an assumption about who is in the room, and the assumption is nearly always someone who works the way I do.
Microsoft's inclusive design principles put the sequence better than I can: recognize exclusion, learn from diversity, solve for one and extend to many. Lorem Ipsum is what happens when we skip the first step, not from malice, but because recognizing exclusion means examining the parts of our practice we have agreed are settled.
So examine one settled thing. The gray blocks in your next file are a good place to start, and the cost is an afternoon and a few honest sentences.
Who is in your next research session, and what will the first thirty seconds sound like to them? Let's find out before the session, not after.