Pretty much whoever you ask, folks will tell you that the lifespan of a website is about 2 to 3 years. After that it looks tired, the person who manages it can’t stand working in it, and somebody starts pricing out a redesign. Most of the web design world treats that number like mileage on a car, as if a site simply wears out from being driven.
However, I actually think about that number a little differently than most. Two to three years is about how long ‘design drift’ takes to wear a site down when every new page gets built as a design exercise, and it keeps such good time on the sites I see because nobody at the operations level is in charge of the structure once the site launches.
The one argument I want to make here is the simple fact that design drift is an operations problem. Until building a new page becomes a content job, a fresh design will always drift on the same schedule as the one it happened to replace.
What drift looks like from the inside
A lot of the people who come to me open with some version of the same sentence: “Hey, my site doesn’t look like it used to.” It’s often paired with the simple fact that the person who manages the site absolutely despises working on it.
Drift starts very small. Somebody wants a little extra emphasis on one line, so the font gets bigger, and then the rest of the site’s font sizes get nudged to match. Things get a little bolder. Then someone decides a button needs a different color to stand out. I see this a ton on sites that typically use a single color. Blue is one of the most common ones I see, where every accent becomes a different shade of blue, and therefore nothing stands out because you’re sitting in a sea of blue.
The early signs are easy to miss. A new page gets built by someone who doesn’t know how the image cutouts were made everywhere else, and every image on it ends up a plain square. The buttons come out a little rounder, or a little bigger. The phrase I listen for is “Is that good enough?” When someone looks at a page they just built and asks that, it usually means the page has wandered off brand and they can feel it. Give it a year or two and you can tell which pages the professionals built and which ones the admins built.
None of those edits is a disaster on its own. Stack enough of them and you end up with a website you’re not proud to show people, one you steer prospects away from when it should be the first place you send them to back up everything you’ve told them. And it happens on every platform (WordPress, Squarespace, Wix, Webflow….. it doesn’t matter), because every one of them is powerful enough to let you change font colors and sizes across the board. You can drift a site no matter how locked down it is.
Every new page is a design job
On most websites, building a new page means designing a new page. Somebody opens a blank white screen and makes a string of design decisions (which layout, which font size, which color, how the images sit) and then writes the content.
That’s a LOT to ask of the person who usually ends up doing it. A lot of website management gets handed to somebody without web experience, and that’s perfectly fine, because managing a website shouldn’t require it. The person who understands your customers and can write on-message words is exactly who should be building pages. They just shouldn’t need a designer’s eye to do it.
Designers wouldn’t agree with each other anyway. Everybody on a team has a slightly different design eye, whether they’re a designer or not, because all of us have different views of what actually looks good. Give three people a blank white screen and the same brief and you’ll get three different pages. Give them the same five, six or seven blocks from a library and ask them to start there, and the pages come out looking like one company built them. I walk through that sequence in the order of operations that keeps a new page on brand.
Most teams think about their site ONE page at a time, and every page gets treated like its own flyer. It’s designed completely from scratch to serve its own unique purposes. The sites that hold up get treated as one set of building blocks, the way you’d treat a box of Lego bricks. We design the bricks, and your team puts them together however the heck they want. When every page is a flyer, every page is a fresh round of design decisions, and enough small design decisions made by enough different people over a couple of years adds up to drift.
A brand style guide doesn’t catch this, by the way. We’re all human and we take the easy path. You can have the best brand guidelines in the world and every template already built, and someone will still go into Canva and make something from scratch. If the design system is harder to use than a blank page, the blank page wins every time, and I’ve written a longer version of that in why your style guide didn’t stop the drift.
Why I call it an operations problem
A website drifts when nobody at the operations level owns the structure it’s built from.
Whoever builds your site, agency or in-house, should be thinking 8, 12, 24 and 36 months out, so the site has a structure you can grow into over the years. And once it launches, somebody on your side has to be the one saying, “These are the structures we want for this site. This is what we need so that we can manage a lot of it on our own.” Without that person, a busy team uses the site a lot and uses it badly, because there’s no system underneath them to fall back on.
That produces a strange result when you think about it. The teams that use their website the most (publishing pages and running campaigns) are the ones who wear it out fastest, and the site nobody touches looks exactly the way it did at launch while doing very little for anybody. Heavy use is a good thing, and it’s heavy use with nobody owning the structure that ends in a rebuild every two to three years.
The teams that use their website the most (publishing pages and running campaigns) are the ones who wear it out fastest.
It’s also why a redesign on its own so often disappoints. A new design dropped onto the same operations goes right back to being a design job every time someone needs a page, and it drifts the same way, usually on the same clock. To be fair to the redesign, sometimes it’s the right answer. If a site is so disconnected that the pages barely relate to each other, trying to patch it inside the current system returns next to nothing, and starting over is the honest call (I’ve laid out what a website redesign costs if you’re already there). What I’d want you to avoid is paying for that rebuild every few years forever.
It happens on well-built sites too
I don’t want this to sound like drift only happens to sites that were built badly. A good example is a university we worked with on retainer for about three years.
When we started, they had an older WordPress site built on a page builder that isn’t very popular anymore. Shortly after, they sent over a new brand style guide that looked ten times better than anything they had. The site was more than 3,000 pages, so redoing all of it at once made no sense. We keyed in on the 30 to 40 primary pages most prospective students would visit, designed flexible building blocks for them from the new style guide (a two-column block that could flip left or right, an icon block that could run one to four columns), converted those pages, and slow-rolled the rest over the next year to a year and a half.
For a while, that worked well. A department would ask for a page, we’d suggest the content it needed, build a template with dummy content, and hand it over. Over time, people from different departments started asking for pages that looked different from everything else, and the feedback on perfectly on-brand layouts was that they weren’t “high quality enough.” Everybody feels their division is the most important one and wants to plant its own flag inside the larger organization. That’s well intended, and long term it doesn’t make sense, but sometimes we built those one-off pages anyway because that’s what they wanted.
The other half was the archive. On a site that size, some old blocks had to stay in the library so older pages kept working, and some team members building new pages grabbed blocks that were five or more years old instead of the current ones. Then the same team would notice pages that didn’t look on brand.
Nobody did anything wrong on purpose. Design drift happens right under people’s noses, no matter how well a site is built, because there are always humans in the loop who want one thing done a very particular way. Sometimes you break your own conventions to make those people happy, even when it isn’t the best use of time, money or brand equity.
Making a new page a content job
The fix is a structure that’s easy to stay inside. You want it to be harder to build something willy-nilly from a blank screen than to open the block library, grab a handful of blocks, and start writing.
A lot of that gets decided at the level of the smallest elements. On the sites we build, every header from H1 through H6 has an exact font family, size and styling, and every paragraph has stock styling, so a paragraph looks the same wherever you drop it. Font sizes come as names (small, large, extra large) from a dropdown, and nobody’s picking pixel values. Buttons come pre-designed as solid, outline or plain text, and you choose the style. When those micro-elements are locked in, the person building a page doesn’t have to think about any of it, and their attention goes to the words.
On a site that’s already drifted, the work starts with a count. With a client we took on recently, the first thing we did was go through every page of their site and screenshot every block we found, and we got to somewhere around 40 to 45. A lot of them did the same job; there were four different versions of a card with an icon at the top, all different enough that you couldn’t tell they were the same thing. Their general look is good, so we’re not doing a full redesign. For every block, we ask one question: how does this block work across several different page types and still give the team flexibility? They’ll end up with a library of about 15 to 16 blocks, each with a couple of layout and color options they can use across the whole site.
None of this means a site should stop changing. A change somebody chose on purpose, with a reason, is how a site grows, and I get into where that line sits in healthy evolution versus drift.
You can start on the habit this week without us. Create a page on your site called your block library (make it unindexable by Google if you can). Once or twice a week, find a pattern of blocks you like somewhere on your site, save it to that page, and give it a name. If your CMS lets you save it as a reusable pattern in the editor (WordPress calls these patterns, other platforms call them components or layouts), save it there too, under the exact same name.
Then ask yourself these, one at a time:
- Who on our team decides what structures the site is allowed to have?
- When someone builds a new page, do they start from a blank screen or from blocks?
- Could we list every block our site uses today?
- When did someone last add a block to the library on purpose?
If those were hard to answer, that’s OK. If your team is using its site hard enough to wear it out, that’s something to be proud of. It just needs someone running the structure so the next page is a writing job.
And if you’d like help putting that structure in place (or you’ve got a site with 45 blocks that should be 15), book a call with us. We’ll look at how your team builds pages today and tell you honestly whether it’s a cleanup or a rebuild.
