StudioDesign

How to Make a Poster With AI (No Design Skills)

Ridvay · June 11, 2026 · 6 min read

How to Make a Poster With AI (No Design Skills)

A friend asked me to make a poster for her pottery studio's open day. She'd already tried — a Word document with the studio's name in 28pt, a clip-art coffee cup, and the address crammed into a corner. Printed at A2, it looked like a parking ticket. Nobody walking past would stop.

The problem wasn't her taste. It was that a poster isn't a document with bigger text. It's a hierarchy: one thing your eye lands on first, a second thing it reads next, and a small detail it hunts for once it's interested. Get that order right and even a plain poster works. Get it wrong and no font in the world saves you.

Here's how I rebuilt hers with AI in about ten minutes, and the handful of rules that made it readable from across a room.

What a poster actually has to do

A flyer in someone's hand gets thirty seconds. A poster on a wall, a noticeboard, or a lamppost gets about three. Whoever's looking is walking, scrolling past a photo of it, or glancing up from their phone. So a poster has exactly one job at first glance: make a stranger think "what's this?" before they've looked away.

That means three tiers, in this order:

  1. The hook — the single biggest thing. A name, a date, a benefit. "Open Day." "Half Price Sundays." "Live Jazz."
  2. The what and when — the next layer down. What it is, the date, the place. Read in one or two seconds once someone's curious.
  3. The detail people hunt for — time, price, a website, a QR code. Small on purpose. Nobody reads this from across the street; they walk up to it.

Most bad posters fail because every line is fighting to be tier one. Everything is bold, everything is big, so nothing leads. Your eye bounces around and gives up. If you remember one thing from this post: pick what's biggest on purpose, and make everything else smaller than you're comfortable with.

Writing the prompt: describe the hierarchy, not the decoration

When you ask an AI design tool for a poster, the instinct is to describe how it should look — "modern, minimal, blue." That gives the AI nothing to organize around. Describe the content and its order instead, and the layout falls out of it.

Here's the prompt I used in Ridvay Studio:

Event poster, portrait. Headline: "Open Studio Day". Subhead: "Clay Lane Pottery · Saturday 12 July". Smaller details at the bottom: "10am–4pm · Free entry · Try the wheel · claylane.co". Warm, earthy feel — terracotta and cream. Big readable headline, lots of breathing room. No clip art.

Notice what's doing the work. I told it which line is the headline, which is the subhead, and which is the fine print — so the AI knows the hierarchy before it picks a single font size. I named a mood ("warm, earthy") and two colors instead of leaving it to guess. And I asked for breathing room, because the number one mistake AI design tools and humans both make is filling every pixel.

What came back: terracotta background, a fat cream headline taking up the top third, the date in a medium weight below it, and the details in a tidy row along the bottom. The hierarchy was already right because I'd handed it the hierarchy.

The squint test

Before you touch anything else, do this: shrink the poster until it's the size of a postage stamp, then squint at it. (On a phone, just hold it at arm's length.)

You should still be able to read the hook. If the biggest line turns to mush at thumbnail size, it's not big enough — or there's too much competing with it. This is the single fastest way to check a poster, and it's how the design will actually be experienced: small, in your peripheral vision, for a moment.

My friend's first attempt failed the squint test instantly — the studio name and the address were the same size, so neither won. The AI version passed: "Open Studio Day" was still legible as a blur. That's the whole game.

Editing what comes back

The first generation is a starting point, not the final poster. Mine needed three small fixes, and none required design skills — just nudges:

If you're new to nudging an AI design after it's generated, I wrote a whole post on editing an AI-generated design without starting over — the same moves apply here: change one thing, look, change the next.

The mistakes that ruin posters

A few traps worth naming, because they're the difference between "made it myself" looking charming and looking amateur:

Sizing it for where it'll live

One poster rarely lives in one place. The pottery open day needed an A2 print for the studio window, a square version for Instagram, and a tall one for stories. The content is identical — only the canvas changes.

This is where generating with AI pays off twice: instead of rebuilding from scratch, you resize the canvas and let the layout reflow, then nudge the headline and details back into their tiers. The hierarchy you set up once travels to every size. A print poster and a social post are the same three tiers wearing different proportions.

Try it

Posters reward a clear head more than a steady hand. Decide what the one big thing is, hand the AI the three tiers, and edit toward more breathing room — not more stuff.

Describe your event and Ridvay Studio will draft an editable poster you can refine line by line. Start with yours here:

Make a poster in Ridvay Studio →

Try Ridvay — the free AI design tool

Describe a poster, social post, flyer or slide and Ridvay generates a complete, editable design in seconds.

Open Ridvay Studio   ← All posts