Why you can't just click and edit

If you've ever used Wix, Squarespace, Webflow, or any of the drag-and-drop builders from the last fifteen years, the first ten minutes in Ullbek feel strange. You look at your site, you see a headline you'd like to reword, and your hand moves to click on it — and nothing happens. The cursor doesn't change. There's no text box. There's no "Edit" button hovering above the section.

That absence feels like a missing feature. We hear about it constantly. It usually comes phrased as a question, sometimes as a complaint: "Why can't I just click on this and change the text?" And then almost immediately after, the same person asks a follow-up: "And why can't I just drag the corner of this image to resize it?"

Both questions feel like they're asking for the same thing. They're not. They're asking for two very different things, and one of them — the click-to-edit cursor — is something we left out on purpose. Here's why.

The promise of click-to-edit

To be clear, the click-to-edit pattern isn't a bad idea. It's a great idea — for a particular kind of website. One built out of fixed, isolated blocks. One where the text on screen is the text in the code, in exactly one place. One where changing a word doesn't have any second-order effects, because nothing else on the page is connected to that word.

For a site like that, the click-to-edit cursor is the most direct, most intuitive interface ever invented. You see it, you click it, you type, you move on. The whole mental model is "what I see is what I edit." That promise — WYSIWYG, in the literal sense — is what made the drag-and-drop era feel magical when it arrived.

Ullbek isn't building that kind of site. So the cursor would lie to you.

The change that breaks three other things

Let's take a concrete example. You're a coffee shop. Your hero says "Welcome to my coffee shop." You want to change it to "Hi, we're The Daily Roast."

In the click-to-edit model, you click the hero, you type, you press enter. Done. The headline now says what you wanted.

Except it isn't done. The string "Welcome to my coffee shop" almost certainly appears in at least four other places on your site — the call-to-action in your navigation, the alt text on your hero image, the meta description in your page's head, the line in your footer that says "© 2026 my coffee shop". None of those change when you click and type. They go stale, silently, the moment you save. You didn't break them on purpose. You didn't even know they were there. The cursor just made it easy to miss.

And that's only the text. The image is the same image. The font size is the same. The line height is the same. On a desktop it looks great. But your new headline is longer, and at 390px — the iPhone size — it now wraps onto three lines instead of two, and pushes your "Order now" button below the fold.

You didn't resize anything. You didn't move anything. You just changed a sentence. And now there are five small problems on your site that you can't see from the screen you were editing on.

This is the part the click-to-edit cursor can't help you with. It edits the one thing in front of you. It has no idea about the other four things that depended on it.

The point of the loop above isn't to dunk on the old way. The old way works fine if your site is a brochure. The point is that on a real site — the kind people actually run, with a nav and a hero and a footer and a contact form and a blog and a SEO strategy — a single line of text is connected to a dozen other lines of text. The click-to-edit cursor knows about the one in front of you. The AI knows about all of them.

What the AI is actually aware of

When you describe a change in plain language — "change the hero, update it everywhere" — the agent doesn't just run a find-and-replace. It does a few things at once, none of which a cursor can do, and most of which a human editor would forget to do.

The same is true of the question that usually follows. "Why can't I just drag the corner of this image to resize it?" For the same reason: a single image lives in a layout, and the layout around it changes when the image's shape changes. A bigger image pushes the text below it down. A taller image breaks the grid on mobile. Resizing an image isn't editing an image — it's editing the page that contains it. The agent knows that. A handle on a corner doesn't.

The workaround that doesn't feel like one

Now the practical question: if I can't click on the thing I want to change, how do I tell the AI which thing I want to change?

This is the part we made as frictionless as we could. You don't have to describe the element in words. You can right-click on it in the preview, choose "Add to chat," and the chat pre-fills with the element's context. Then you just type the change you want.

It feels slower than clicking for the first day. By the end of the first week it feels faster. By the end of the first month, going back to a drag-and-drop tool feels like work — because every change is one thing in front of you, and you'll spend the next twenty minutes walking through every other place the change should have gone.

A temporary feeling

If you've just started using Ullbek and you're reading this, I want to say one thing clearly: the feeling that you've lost something is real, and it will pass.

It passes the first time you change a word and the navigation, the footer, the meta tags, the image alt text, and the mobile layout all quietly update themselves. It passes the first time you ask the agent to "make this CTA sound more confident" and it rewrites it across all three of your landing pages at once. It passes the first time you resize a hero image and the section below it breathes to accommodate it, instead of overlapping.

What you're not used to yet is the part where you stop being the person who has to remember every place a change is supposed to go. The agent remembers. The agent is, frankly, much more reliable at it than you are. And unlike a cursor, it tells you what it did, so you can check the work.

The first time I used a text editor that didn't show me line numbers, I missed them. The first time I used a phone without a home button, I missed the click. The first time I used a search engine that didn't have an "advanced search" page, I missed the controls. The newer tool wasn't missing a feature. It was a different tool, doing a more interesting job. The familiar interface was a way of coping with the older tool's limitations. When the limitations went away, so did the coping mechanism.

The click-to-edit cursor is a coping mechanism. It's a beautiful one. It made the web possible for a lot of people for a long time. But the thing it was helping you cope with — having to manually keep every related thing in sync, by yourself, across every breakpoint — is exactly the thing the agent does for you now.

You're not being handicapped. You're being promoted.

When you can't click to edit, you are not being pushed down a level. You are being pushed up one. The drag-and-drop interface put you in the role of the person who has to type the text. The chat interface puts you in the role of the person who decides what the text should say. The cursor is for typists. The agent is for editors. The agent is for you.

You don't open Photoshop, click on a text layer, and edit the characters yourself when you want to change your tagline. You tell a designer, "Change the tagline to X, update it everywhere it appears, and make sure it still looks good on mobile." That's what you're doing here. You're the designer. The agent is the typist.

And the typist — for what it's worth — is very fast, very thorough, and never charges by the hour.