When I have an idea for an interface I used to open Figma. Now I often open a text file instead. Not because it is faster, though it is, but because of what a text file will not let me do.
There is no colour in a text file. No typeface, no shadow, no spacing to nudge until something feels right. What is left is where things sit and how large they are relative to each other. When I cannot reach for a gradient, hierarchy is the only decision on the table.
I redesigned my contact page this way. Three actions, one page. I opened a text file and roughed it out.
Three boxes, each with a small label, a headline, a line of body and a call to action. Eight lines, and not one of them is a decision about how it looks.
Then I asked for three structural directions off the same brief: one where the three doors carry equal weight, one asymmetric where the primary action dominates, one with a personal hero above the doors.
Variant A · Equal weight, three cards, no hierarchy. The default you'll get when you don't bias the layout.
Variant B · Asymmetric. The Cal door dominates with a "fastest way" tag; the other two stack smaller on the right. Hierarchy without removing choices.
Variant C · Warmer. Personal photo + greeting + credentials above three smaller doors. Says "real person on the other end."
The differences are visible rather than argued. A treats the three doors as equals. B makes one of them the obvious answer and stacks the other two beside it. C puts a person in front of you before you choose. Those are three different experiences, not three skins, and I could see which was which in under a minute.
I went with B, then rendered all three as working HTML to check.
The render confirmed what the sketch had suggested: the asymmetric hierarchy held at every size. That is the order I keep working in. The decision happens in letters. The pixels prove it.
What it makes me ask
Working this way puts the same questions in front of me every time. What is the dominant element. What comes first. What belongs together. What is the next thing someone does. What can come off the page entirely. In Figma I can defer all of those for an hour by making something look nice. In a text file there is nothing else to do.
There is a second effect I did not expect. Hand an AI a Figma file and it copies what it sees: my nav, my card style, my particular red. You get more of what you already have. An ASCII sketch is too coarse to leak preferences, so what comes back is not anchored to anything I have already done.
Where it does not help
It skips the feeling stage. Sometimes I need to play with colour and texture before I know what the structure should be, and letters cannot help me there. It also assumes I already have an opinion about hierarchy; if I genuinely do not know whether the headline or the image should lead, I am still in moodboard territory and need to leave it first. And for brand identity it is simply the wrong instrument. You cannot choose a typeface or a colour story in pipes and dashes.
Landing pages, contact pages, dashboards, internal tools, prototypes, posters, emails, anything where you mostly know what should be on the page and need to see what it could look like.
Brand identity work, illustration, photo-led pages, anything where the visual decision is the design. ASCII won't help you pick a typeface or a color story.
The rule I use: could a senior designer describe this on a napkin? If yes, letters will do. If they would need to show me a moodboard first, they will not.
Figma still comes in, for polish and the final screens. Just not first. Holding off on the polish is what keeps me honest about whether the structure underneath actually works.