A website that is almost finished is a commercial leak

I like opening the websites other designers proudly share. Within thirty seconds I know whether it wins me over or loses me.

It usually goes like this. A web designer shares a post from client A. Client A posted it themselves, full of pride: the new website, built by that designer. Makes sense. They see the surface, and the surface is usually fine. Then I open the site, because I enjoy seeing what others build and whether I can learn something from it. And within half a minute I see the cracks.

The page is not locked down. Scroll a bit too hard and it jumps a few pixels to the left, with a white strip appearing on the right. On mobile, the footer items are packed so tightly you cannot tell where one link ends and the next begins. The hamburger menu is taller than the screen and overflows. And sometimes, the most annoying of all, there is simply a typo in there.

This is not about taste

The designs are rarely bad. It is about something else: settling for almost. The final ten percent, where a design needs to become a working system, gets skipped. Sometimes by a deadline, sometimes by budget, sometimes because there simply is no eye for it. The result is the same. The site has been delivered and it does not hold.

The applause comes from people who never opened the site on their phone

Below the post, fellow designers react positively. Understandable. They see the same render the designer saw: on a large screen, in the format it was presented in. Nobody picks up their phone, scrolls hard and watches for that strip on the right. The appreciation is about the picture. The execution stays out of view. That is how almost stays invisible to everyone except the visitor who actually has to click.

For the business owner, this is a commercial leak

You pay a substantial amount for a website, and your potential customer lands on a page that shifts, a menu that is impossible to scan and copy with a mistake in it. That customer does not know why it feels off. They only feel that it does, and they leave before reaching your offering. That costs revenue before the conversation even starts. I am that visitor who clicks away myself. If you cannot fix that page on your own, a second leak appears: depending on a single agency for every correction.

The design was approved as an image, not tested as a system

That is the core of it. A static mockup shows one state, at one screen width, without interaction. A real website is a hundred states: every screen size, every scroll position, every movement. What looks fixed in the mockup starts shifting the moment the build is not anchored. A page that jumps to the left with a white strip is almost always an element wider than the screen. The layout was not locked to the available space. Google even has a metric for it, Cumulative Layout Shift, part of the Core Web Vitals. What you experience as "it moves a little" is a number that can be measured and recorded.

The footer that runs together on mobile comes from the same principle. It was styled to the width of the design, not on a grid that also works at a narrow screen. And the typo survives because proofreading happens in the design file, not on the live site. Two different things, two different checks. It is the same principle I apply when building a logo on fixed proportions: lock down the measurements, and it holds at every scale and with every supplier.

Four checks that take two minutes

• Open the site on your own phone and scroll through it firmly. Does anything jump, or does a strip appear that should not be there.
• Look at the footer and the menu on mobile. Is everything clearly separated and scannable at a glance.
• Click through every important link on the live site, and read every word along the way. Does every button work, does every menu item lead somewhere, and is the copy correct in headings, buttons and error messages. A dead link costs you a customer at the exact moment they wanted to continue.
• Narrow the browser window on your laptop to its narrowest. A page that is locked down stays intact. A page that is not falls apart here.

If any of these checks finds something, the site is not finished. Not because it looks bad. Because it is not locked down.

I still enjoy opening the sites others build, precisely because I learn from them, including from the things that are not quite right. If you are unsure about your own site, I am happy to take a look. No sales pitch, just a second pair of eyes that knows where things start to shift. If you want it structured, start with a short baseline measurement of your own brand.

Demy Rink, Brand Systems Architect