Hacker News new | comments | show | ask | jobs | submit login
Why do all websites look the same? (medium.com)
266 points by guzik 5 days ago | hide | past | web | favorite | 223 comments


1) Why are all buildings the same? Light switches are often in similar places and the space between the floor and ceiling is pretty standard.

2) Why are all vehicles the same? Mirrors are always in the same spots and seat belts all work the same.

3) Why are all laptops the same? Keyboard center on the bottom with a trackpad or nub near the center. Screen on top, ports and stuff on the sides.

There are components that are common in all facets of our lives that when different can cause problems or surprise which could be good or bad. We need to join two floors of a building. Use stairs! People understand stairs. We need to showcase a collection of clickable images. Use a grid! People understand link grids.

If you want to make your website usable you have to lean on expectations and those are pretty well defined nowadays. Imagine walking into a room and turning on the lights using a switch in the middle of the floor or plugging in your laptop's power cord at the top/back of the screen.

Most companies spending money on a website want them to feel fresh and creative and engaging but they also have to temper that with usability and expectations. That's why all websites "look the same" or at least why the author thinks they do.

Just because certain elements are in the same spot(s) or behave similarly doesn't mean things are the same. Or, at least, to me they aren't.

A closer example would be books.

For ~500 years, books look approximately the same: normally paged sideways (not top-down), with some margins for handling, with text in rows or columns (depending on the writing system), some chapter structure, and index / contents page, page numbers, covers of more durable material to protect the pages, with some kind of a title on the top cover, etc.

Those features don't just exist because of tradition or technical limitations. They mostly exist because they are convenient, useful, and logical.

But they also exist because people expect them, from times of handwritten books. They put the skills people already had to good use. They created a visual language which is easy to pick up and easy to use, both for readers and typesetters.

Most web sites are a logical continuation of books, magazines, newspapers, etc. No wonder they actively adopt the time-proven, well-working concepts from the print media.

Forms have a much shorter, and much less rich history outside web, and here experimentation was wild; a lot of sites do forms quite differently. Though some common language (like labels, placeholder text, pre-validation, etc) already has formed. OTOH even checkboxes are not yet a commonly accepted visual concept; some e.g. prefer "switches", iOS-style.

> They mostly exist because they are convenient, useful, and logical.

What happens, when, in the words of the article, you "Do not be constrained by questions of usability, legibility, and flexibility"?

The answer, it appears, is their re-design of HN: https://interface.fh-potsdam.de/future-retro/HN/ which I find unusable.

That's shockingly bad.

The expanded view with lines between comments and their parents which scroll independently is literally the worst website UI I have ever seen. And I recall some terrible flash abominations.

The pure white screen has a certain aesthetic appeal, but the total lack of content puts it considerably below the original. (I assume this must be an intended effect, given the complete absence of error messages.)

The medium redesign is also awful. Its inability to handle my portrait-oriented monitor did trigger a sense of childish amusement (truncating POOR to POO) but it's entirely impossible to read anything using it.

> Do not be constrained by questions of usability, legibility, and flexibility

That is, create a work of art (as opposed to utility). Okay, art is nice, too.

But mixing artistry and utility is quite hard. Designing a nice, livable house, or a store, is a craft. Building fancy sand castles is a (pastime) art. But mixing the fancy free-form with being actually usable by thousands takes lots of work, resources, and time, because what you'd have to build is a cathedral. (If you think modern technology makes it simple, look at the history of building the Sagrada Familia in Barcelona.)

I can't imagine what they think someone with disabilities is supposed to do with that.

They don't care about people with disabilities. They don't even care about people without disabilities. It's not usable for people who are not differently abled.

That was a pretty hilarious experience.

the day/night button on the bottom right is genius satire

Actually wouldn't be terrible if the comments appeared in a more orderly fashion. Right now, they spawn randomly and can even overlap which makes it actually unreadable.

The "redesign" is a joke, right?

Only an artist could come up with this: http://prntscr.com/lgk7x0

Is this brutalist design?

No, brutalist is HN, lobste.rs or the Oreilly CD Bookshelf. Or any Unix CS guide with those square Motif-like buttons.

That site could induce a seizure in David Parsons.

Full justified text exists because of tradition. It makes the text harder to read because the word spacing becomes uneven, the line ends become harder to distinguish, and more words are broken by hyphens. But it's higher status because when all books were hand-written it took more skill to produce.

Web sites have a similar popular anti-feature that's included for reasons of social status: low contrast text. I think this is popular because ability to easily read it signals both health (good eyes) and wealth (good screens).

Huh. I always thought non-justified text existed because people are lazy; I absolutely consider fully justified text to be easier to read, not harder. Non-justified text is distracting.

I'm sure it varies from person to person. I believe, though, the bulk of the studies have found that, while people generally find justified text to look nicer, reading speed and comprehension tend to suffer for it.

Same for contrast. If you're not actually trying to read, low-contrast is more appealing. Thus, the CEO who has read their marketing blurb a thousand times likes low-contrast, even though it's counter-productive.

> "I think this is popular because ability to easily read it signals both health (good eyes) and wealth (good screens)."

Perhaps. But I always presumed the good eyes, good screen, and lack of empathy belonged to the designer. These same site too often seem to have experiences based on an ultra-fast connection, as opposed to a wonky 4G (at best) connection.

You might be right. But I've sat in meetings and/or shared office space with low-UX-IQ designer / frontend types.

I think the low contrast thing is simply a symptom of a general trend twoards form over function, which though it may seem cool, in the end turns out to be cheap. Its the equivalent of putting dark plastic tint on your car windows. It happens in a lot of fields. The truth is most people don't know what they are doing (and we all didn't know what we were doing at one point).

I don't see why is this downvoted. This is quite valid.

Fortunately extensions like Stylus allow to tweak page styles to make them more usable.

> Full justified text exists because of tradition. It makes the text harder to read because the word spacing becomes uneven

That should barely be the case with a good layout and hyphenation algorithm (or a competent printer in the old days). What word processing software does is not necessarily the best possible way to produce an even justified layout.

> For ~500 years, books look approximately the same

I think you can even go back as far as 1000 years to find those same standards. https://en.wikipedia.org/wiki/Carolingian_minuscule

Yet most books have awful spines that prevent the pages lying flat when open. Not because it's useful or logical but because it's cheap and easy for the publisher.

> Most web sites are a logical continuation of books, magazines, newspapers, etc. No wonder they actively adopt the time-proven, well-working concepts from the print media.

True, but print media formats got even more simplified in web. The reason? Responsiveness constraints.

Many times I've seen great advanced "article-like" designs for desktop from ambitious designers. It turned out that making them easy to manage in CMS, stable in all possible combinations (or implementing good validators) and at the same time logical and good looking on mobile was extremely hard and not worth it.

This led to further simplifications...

... and then Medium looks like Medium.

There's not much place for creativity on mobile. And this is where most of the "fast content" is read.

I'm actually gonna have to go ahead and blame the demise of Flash for the current lack of interesting web design. I know there were technical reasons to get rid of it. Security. Mobile. Proprietary. etc. Whatever.

Some flash sites were horrible. But there were some real gems. That Hacker News design by Tyrion's brother is probably inspired by a bunch of old Flash sites that followed that template.

Flash was so easy to work with. Draw some neat stuff with their tools, animate it in the same program with their timeline, then make it all navigable and smart, still in the same program. You test it, in that same program. Then you publish it, and it works exactly how you made it work, in every browser, ever. One file. So easy for artists, designers, and non-technical people.

That's why we don't see cool shit on the web anymore- it's too hard/boring to make.

Out of 100 sites in flash, maybe 2 were the gems you describe. The remaining 98 were utter crap that should die in a fire.

The people who designed these gems would probably be able to design good sites no matter what tools they had at their disposal.

Except in that the people paying the designers would demand sites like the 98 horrible ones.

And you consider that different from the current generation of sites how?

All kidding aside I think the overall level of USABILITY of websites is much higher than it ever has been today. But in large part that’s because they are all following the same foundations.

> it works exactly how you made it work, in every browser, ever.

My experience of flash is that it mostly didn't work, and I had to spend an hour after every new OS install trying to get flash to work.

And copy and paste doesn't work, you can't right click to save images, etc. Unless you're trying to make an animation, or a game, flash gives a really lame user experience, I don't know why you like it so much.

The objective of most websites is not and should not be cool and interesting design, unless it happens to be a site about design.

At the end of the day a site is there to serve a purpose. And, generally speaking, that demands familiarity with the UI and ease of use. If every site demanded increased cognitive load from the user the internet would be a horrible place.

I get that designers want to design cool stuff. Love it. But when all the smoke and bull clears out the mission is to sell a widget, deliver a service or provide information. Craigslist and good black and white movies prove that good and useful content is what people are after, not award-winning design and cool web tricks.

> the mission is to sell a widget, deliver a service or provide information

But why? What about a site about art? What about a site about me? What if I don't want you to get "information" about me, but to see the world through my eyes, with the website as my lens? Sure, I guess I could post my non-interactive photos to Instagram, but that seems so... uniform.

Not every website has to make money. They can entertain just by the virtue of being interactive. And I'm not saying it's impossible to do in 2018. I'm just saying it was a much more creative-friendly experience 10 years ago, and when we killed Flash, we lost it.

We did gain apps though. So I guess there's that.

> Not every website has to make money. They can entertain just by the virtue of being interactive.

This is exactly how I remember the best Flash sites of the 2000s. They weren't creating interactives to drive some lead gen campaign, or sell their blockchain-based SaaS or whatever; it was made purely because it was cool and fun to use. Some portfolio websites for UI and FE developers still have this feel.

> We did gain apps though. So I guess there's that.

I get the sarcasm, but IMO, I'd rephrase and say "We did gain Codepen". Lots of cool visual stuff there, but I guess it would be mostly of interest to developers, and not the wider public. Still, I see stuff in Codepen daily that would never make it into a "production environment" because it didn't fit the CMS template, it wasn't responsive, or was using new tech like CSS Grid, or a multitude of the other reasons why websites stopped being cool.

not OP, but I think the fundamental difference is design vs art. Designing for use will drive you to create a site similar to what's out there. Creating art for expression leaves you plenty of room to communicate the aesthetic, mood and so on that you want to share.

I believe I did cover your scenario when said there was an exception if the site is about design. If showcasing art, design or a specific personality is the objective then usability might have to take secobd fiddle, and that’s fine.

"Did you click on the squid?" http://www.penny-arcade.com/comic/1999/03/03

Reminds me of the only flash site I really liked, jimcarrey.com. Shame it's down now. Though somewhat preserved in video eg https://www.youtube.com/watch?v=5QzhEZnOYE4

I'm not disagreeing that flash was easier but are you sure it isn't just you got jaded? The things you though were cool 20 years ago in Flash just aren't as cool?

I googled "cool websites" and these came up


don't know if any of those do anything for you. I'm too lazy to look for more but I feel like they are out there.

The nature of how we browse the web has fundamentally changed since the days when Flash was king. What is the point of investing in a gorgeous website if the audience has moved over to FB/Twitter/Insta and are now acclimatized to scrolling through endless content feeds for a slight hit of dopamine.

By Googling "cool sites", you've stumbled upon the crux of the issue. The site that ranks at the top is the SEO-optimized, but incredibly cookie-cutter, template-heavy Awwwards, which is a back-patting congratulatathon of creative directors at design agencies. So of course the websites featured are their own.

I'm sure at some point they wanted to feature real client websites, but chances are those were mini-sites created for specific campaigns, and were taken offline after the campaign ended.

There's a lot to be said for Flash. It's a good visual design environment. It certainly beats the horror of CSS/Javascript/WebAssembly/WebGL we have now.

(Giving page designers total control over scrolling was a huge mistake.)

I love flash for games and movies but never for webpage content or navigation

> I'm actually gonna have to go ahead and blame the demise of Flash for the current lack of interesting web design. I know there were technical reasons to get rid of it. Security. Mobile. Proprietary. etc. Whatever.

The thing is, all that's been replaced with JavaScript, which has all the downsides (well, it's not really proprietary, but JavaScript-heavy websites aren't really usable as something to learn from & modify, especially once minification comes into play) and has the great new flaw of not realistically being disablable, unlike Flash tended to be.

We even saw this week Google deciding not to support non-JavaScript browsers.

JavaScript is a boot stomping on a human face, forever.

I'm not even talking about the consumption side of Flash. I'm just talking about the tooling for creating Flash content.

Did you ever use Flash? It was incredible. Draw right there on the canvas with basically Illustrator-esque tools (or just File->Open whatever .jpg you want). You can add a keyframe in the timeline, and just drag the image to wherever you want. Then you hit play, it looks right, then you hit publish, and it would play exactly the same in any browser ever.

Also, ActionScript 3 was basically typed javascript, back in like 2004.

I'm with you. I used to be able to crank out a Flash game, with art, in a or two in my spare time, no big deal, not have to look anything up, I knew exactly how the whole system worked (I literally made games completely disconnected from the internet the entire time, my productivity was crazy high), never had to waste a bunch of time fighting frameworks or seeing how things interacted, and could get it to sing (although it had definite limitations. Everything existed in a movieclip object, for example, so when I was asked to port a C++ game that had 900+ objects on the screen at a time it really slowed down).

With Flash, I once ported a game to Flash, from scratch, with art (but no sound...although I could have added sound), in the span of 12 hours, and while it wasn't my most popular game or anything I still get people telling me how much they enjoyed it 15 years later.

In fact, the most popular game I ever designed and released, Proximity, I designed and released in a single week of work (in my spare time).

I have never been able to match the speed and flow of development that I got with Flash since with any game engine since then, and I've tried a whole bunch trying to find something.

The closest thing to it now seems to be Unity but that's a much bigger beast that I don't have a full grasp on and have to look things up or download various things from the asset store. It's possible to make games quickly, but they'll probably play and look like garbage if you don't take your time with them and hire a proper artist, judging by the flood of garbage Unity games released on Steam.

Pico-8 is also fun to program with (it includes a built in sprite editor, level editor, and music editor), but it's a little too limited for my tastes, since I can't really make a commercial product with it, and I'm not that great at doing 2d pixel art. But I still spent some time doing most of a port of Proximity while playing around with it and for people who can do pixel art it's a lot of fun to use.

* 12 hour game, Squarez: https://www.newgrounds.com/portal/view/91933 * Proximity: https://www.newgrounds.com/portal/view/183428

These same tools exist now for generating similar content sans Flash. Heck, the company behind Flash (well, the acquiring company behind Flash) even makes some of said tools.

Flash did use JavaScript. More precisely, ActionScript, which is the abandoned JavaScript 4 with classes etc. very much like Java. What Flash had was a usable scene graph + synced audio, though, rather than half-arsed attempts to use the DOM/CSS, canvas (immediate mode 2D), undermaintained SVG, or WebGL. I wonder why we were so fast to kill Flash (myself included); maybe it would have been worth to try and open-up Flash. Many, many more designers could author Flash and create valuable content compared to the schizophrenic web stack we have today.

> Flash was so easy to work with


I wasn't really around in its hayday but it had a very low barrier to entry - I remember lots of cartoons being spread years before youtube, you couldn't stream video, but you could wait a minute and then watch a flash animation. You didn't have to be a programmer to make content.

As an example, the old jkrowling.com website had a great non website interface.

You're confusing mass production with adoption of similar artistic concepts.

Buildings have light switches and ceiling heights standardized because of building codes which mandate these things, and the fact that many components are mass produced so they're extremely cheap (like light switches) and are used everywhere.

Vehicles have things in the same places because regulations require it, it makes sense to do so (mirrors aren't much help if they're behind your head), and because drivers expect similarity.

Laptops are similar because that's really the only way to make them work. A keyboard over the monitor isn't usable: your arms would block your view. This is just silly.

Websites may be built with the same components (python, PHP, web servers, HTML, CSS, etc.), but that in no way means they need to look the same. It's entirely possible to make them look very different, and you only have to go to the Wayback machine and look at how sites used to look 15+ years ago, and compare to modern ones, to see this.

Basically, most of it is a cargo-cult mentality: sites update to "newer" designs that are less useful because it makes them look "fresh" and "modern" even though they waste a ton of whitespace and make the site slower and less useful. Sites used to be much better in the mid-2000s.

It's not just a cargo-cult mentality.

There's a decent amount of research that has found that putting things in non-standard locations impairs most users' ability to accomplish what they want on your website.

The Nielsen Norman Group is a decent place to find some of this research. There's some decent information here about how conventional layouts tend to be more effective:


and how placing logos in the center instead of on the left often prevents users from accomplishing their goals:


and the reason all of this matters is that the computer skills of a typical person using your website are probably far, far worse than you think they are:


so if you deviate from the "standard" layout too much, you'll hurt your website's business value because people won't be able to use it effectively.

I think there's still lots of room for creativity within a standard layout. But too much creativity might result in a site that looks better from an aesthetic standpoint but is less effective at actually delivering business value.

> Buildings have light switches and ceiling heights standardized because of building codes which mandate these things, and the fact that many components are mass produced so they're extremely cheap (like light switches) and are used everywhere.

You mean similar to included youtube videos, Facebook like buttons, Google analytics, ...

> Vehicles have things in the same places because regulations require it, it makes sense to do so (mirrors aren't much help if they're behind your head), and because drivers expect similarity.

GDPR cookie settings? Imprint? Privacy statements?

> Laptops are similar because that's really the only way to make them work. A keyboard over the monitor isn't usable: your arms would block your view. This is just silly.

It would be silly to place the menu on the bottom, as you would always need to scroll down the whole page to see what options you have.

> Websites may be built with the same components (python, PHP, web servers, HTML, CSS, etc.), but that in no way means they need to look the same. It's entirely possible to make them look very different, and you only have to go to the Wayback machine and look at how sites used to look 15+ years ago, and compare to modern ones, to see this.

Houses may be built with the same components (bricks, wood, concrete, glas, metal), but that in no way means they need to look the same. It's entirely possible to make them look very different, and you only have to go out in the world and look at all the different implementations.

Sorry ;-)

Menu on the bottom is obviously pointless, but sticky menu on the bottom would be so cool, most of the mobile apps moved to tabbar.

Why it's not used extensively in web? (it would make it at least a tiny bit less boring).

On iPhone when you scroll down the page, bottom Safari bar collapses and your sticky menu would be still sticky at the bottom. But clicking any button on the sticky bottom menu wouldn't invoke click on the menu. It would make iPhone Safari menu uncollapse -> huge confusion -> potentially less engagement -> retention.

Web today is full of similar constraints. That's why there are usually one or two options that make sense and everything looks the same.

Because on mobile the primary interaction point is your right (or left if you're a leftie) thumb, which naturally falls over a tab bar at the bottom of a screen. Whereas on the Web the primary click zone where your cursor focus is, is at the top of the browser window where the navigation elements are.

There's also the obvious stuff about top-to-bottom reading order making such a layout making the navigation much more obvious when you first look at a site (which is probably want you want for most things, although not all, which is reflected in things like blogs often putting ancillary navigation elements in the page footer).

I don't think this is a broken thing that needs fixing.

I'll also note that it has taken mobile UX designers YEARS to start moving navigation elements to the bottom of their apps, driven by larger and larger screens making elements at the top of the screen quite uncomfortable for one handed use. It's required that to become sufficiently cumbersome that the trade-off in immediate discoverability becomes worth it. And you'll note that the most popular mobile apps such as WhatsApp still have all their navigation elements at the top, regardless.

It's not a cargo-cult mentality. It's the convergence on a set of best practices "illuminated" by analysis of user-behavior data, as well as the application of basic graphic design principles (like the use of white space to create a visual hierarchy).

It's not mass production that causes this though, it's the constraints (the things that make the process "design" instead of "art") that come from how they are used -- and while some of those have solidified into technical or legal constraints, many are simply matters of convention.

Similarly, any user experience study of a website isn't going to find that you should make your menu some diagonal shimmering nonsense, or try to convey information on the side of a spinning cube or any of that sort of thing.

Thank you!

Design != Art !!!

If designer is more interested in fulfilling his artistic/creative ambitions than business value for client (making things great looking, stable and reasonably easy to implement -> in budget), he shouldn't even touch digital today. It's too complex and filled with too many constraints.

But if same designer wants to make decent money, well...

> Laptops are similar because that's really the only way to make them work. A keyboard over the monitor isn't usable: your arms would block your view. This is just silly.

With tablets like the iPad, the monitor is the keyboard, and your arm does block your view. Maybe it is 'silly' but a whole lot of people buy and use them.

There's been lots of other laptop designs, and they seemed much less silly to me than tablets. Compaq used to put the trackball on the right side of the case (even lefties I know mouse right-handed). HP made a laptop with a little pop-out mouse, which didn't even require a surface to place it on. IBM made a keyboard wider than the case, which unfolded when the lid was opened.

Everyone I know who tried or owned these laptops loved them. Why did they die out?

We were actually talking about something else in the office today that seems similar: Apparently around 2009-2011, several companies were coming out with dual-monitor laptops that folded out when open... Which then died out and disappeared.

Are we talking like Razer's concept from a couple years ago at one of the tech shows (CES? I don't remember, honestly) or something like Acer's Iconia dual screen tablet/notebook?

I had the Thinkpad W700 with the pop out side screen, not to mention the wacom tablet built into the right of the touchpad. It was a great parlor trick, and genuinely useful for running a terminal or a music player to the side of your main workspace.

[0] https://www.engadget.com/2009/01/26/lenovo-w700ds-laptop-rev...

Microsoft also had a foldable dual-screen design in the works called Courier. Apparently they're trying to revive it: https://m.windowscentral.com/microsoft-bring-back-courier-20...

Acer Iconia was actually one of the models we found online, I don't know what Razer concept you're referring to.

Just had to look it up: https://www.razer.com/project-valerie

They demoed it and apparently the unit they demoed was stolen when they were leaving the show.

Not going to lie, I’d love an Acer Iconia style portable at this point. Wouldn’t use it for desktop publishing or probably anything with heavy typing, but just seems like a fun form factor.

Bob Lutz, the car guy (senior positions at BMW, GM, Ford and Chrysler), once commented on this. He noted that his designers liked Bang and Olafson design, with black on black buttons and unusual form factors. This was all wrong for a car interior.

make the site slower

Websites are far faster than they've ever been in the past, even with the bloat of frameworks and fonts. Most sites load in < 5s these days. That was not the case 15 years ago.

I have to disagree with this. Websites loaded slower because our connection was much slower. You’re comparing network speeds to page load times, which are different.

With today's speeds, the older internet is so much faster than “modern” sites. Some sites even have a loading bar (ex: gmail). Compared to sites with minimal/no Javascript, they’re much slower. 5s is not a good load time for a webpage especially with today's internet speeds.

5s is not a good load time for a webpage especially with today's internet speeds.

I completely agree. I work to a perf budget of 200ms to first paint on the things I build. 5s is a huge amount of time to wait. It's the upper limit of what I think is acceptable.

What you're describing is not a web site, but an application delivered through the web browser. Sites in the past were smaller and faster because they were just web sites. Today we have a whole host of applications delivered through the browser, with the Internet as the hard drive they're loaded from.

It's not a valid comparison at all. Compare the time it takes to download and install Thunderbird versus the time it takes to log into Gmail, and there you'll have a valid comparison. Compare the time it takes to install Office versus the time it takes to log into Excel Online and there you'll have a valid comparison. And in comparison, it's a hell of a lot faster.

Gmail in 2018 feels like I'm downloading Thunderbird every time it opens. Ideally I'd only need to load it all once and then it'll be cached for faster load next time. But maybe that's too much to ask when changes must be deployed multiple times a day.

And yet the Gmail experience is not significantly better than non-JS alternatives, at least for m my use cases.

Irrelevant. When more resources become available(like bandwidth) we use them. The fact is pages are loading faster than 15 years ago.

Yes they could be faster, but most are still useful at a 5 second load time.

Which ones?

In what is most likely a parallel universe that I inhabit, websites have been getting consistently slower for a while. No amount of edge black magic can compensate for the growing bloat of fonts, frameworks, multiple-dozen-megabytes background videos, and half a thousand requests to ad networks.

> Websites are far faster than they've ever been in the past, even with the bloat of frameworks and fonts. Most sites load in < 5s these days. That was not the case 15 years ago.

I had a cable modem 15 years ago. Pages loaded faster, even though I had 1/10th the bandwidth that I do now.

Could be that web servers don't have much more bandwidth per client. Or that they're actually cold starting VMs or lambda functions instead of running constantly. Still, I agree the overall feel of the web had gotten shower and jankier over time.


A while ago most people didn't realise to what extend loading times affect engagement.

If they knew, flash would be dead much quicker.

This reminds me of how disappointed I was in my adolescence by how all airports look the same. Only much later I realized that almost any kind of building has a common design and relies on familiarity and expectations in its potential inhabitants.

> Why are all buildings the same? Light switches are often in similar places...

That's not true at all. Buildings are often very different from one to the other in both exterior design and internal layout.

In my experience, even light switches in commercial buildings are often in very different spots to where you'd expect, relative to a normal residential house. This is partly due to lights being switched on from one central area, and not being something normal visitors or workers of the building need to use. These are fittings anyway, not "the building".

Actual building architecture and interior design is very diverse, so your analogy here using buildings is not a good one to compare with the often identical website layouts seen everywhere.

It's fun to see how <fields> start wild, people explore things as they feel. It's a jungle. Then <field> become a thing, optimization/laziness kicks in, people expect some concepts to be expressed in a certain way so they're predictable. No more jungle.

> Why are all buildings the same? Light switches are often in similar places and the space between the floor and ceiling is pretty standard.

You may not notice, but ceiling height varies regularly from 7.5 to ~10.25 feet, with absolutely everything in between represented. Only in a relatively small subset of large, wood-framed houses is it somewhat standarized to 10-and-a-bit feet.

Lightswitches are crammed in wherever and are very often a case of "yeah, looks good." Often a contractor will have a height he has people put them in at, but it varies from person to person. 4' is the most common but there is no code for it like there is for walls or railings.

Light switches themselves are all dictated by standardized form factors because different companies make the boxes, switches and faceplates and all need to agree on where the screws go for anything to work. Software, particularly web dev, is hardly limited by that. Design rigidity and similarity is dictated by convention, not necessity.

> Why are all vehicles the same? Mirrors are always in the same spots and seat belts all work the same.

Those two are laws. However specific shapes, measurements and ratios like hood height, hood length/windsheild size, and body shape are all extremely similar in order to very aggressively optimize aerodynamics and crash/pedestrian safety. That's a reasonable comparison to web development: webpages are made to understood principles of UI and UX design, and well understood design patterns.

> Why are all laptops the same? Keyboard center on the bottom with a trackpad or nub near the center. Screen on top, ports and stuff on the sides.

As an electrical engineer, I personally hate this uniformity. I hate the experience of laptops in general. I would very much like to make a laptop with no (well, one) moving parts, entirely glass, plastic and carbon fiber. The only moving part would be a single, extremely robust and stiff 180 degree hinge. Both sides of the clamshell are low-bezel 1440p touchscreens with localized tactile feedback, matte finish, and the keyboard and mousepad light up with a border wherever. Reconfigure the UI components with pinch and drag. Tilt the keyboard 45 degrees to type in bed. Use it like a book, a tablet, a newspaper, a laptop, whatever.

Problems include typing fatigue, touch-typing, breaking the damn thing, battery life, yadda yadda. But someone could try something creative. Laptops are in a hellish halfway of standardized consumerism and unstandardized technology- so identical, but so unaccessable and so unexchangeable. So uncreative. Why are slow/fast chargers and external batteries so hard to find? Why is Microsoft making the most creative devices, like the surface? Ugh. Laptops suck.

You may be getting your wish, or very close to it: https://www.zdnet.com/article/windows-10-dual-screen-laptops...

I think the redesigns in that article demonstrate why most websites look the same.

Let's do a test. Go to a bookshop and buy a novel. Now go to a different bookshop, and buy a different novel from a different writer and a different publisher. Read them. Chances are they're going to look very similar. Different cover maybe, but inside, mostly black letters in an unobtrusive font on a white page with sensible margins. No decorations or illustrations, nothing at weird angles or sizes. They all use the form that's efficient at transferring the content, the story, to your head.

Some web sites are about their design, but most are about the content they present, and they want to present that content in an efficient manner. A design that distracts from the content is not practical.

    t l i sya f t a 
    hoo n-tlm r-e t
    rd anddsnak tbe
    ens dn  tel's g
    aeiyeuwgsryc yi
    stmed in   ialb
    o ih otioomtrsi
     ea o oi  si oi
    wtrt yuuetttgtt
    ei abstqfr rriy.
    bsihea e-o aau
      st e r f '-t
I think that makes the point about why avant-garde style can be difficult to comprehend, no?

(read the columns up-and-down …)

>I think the redesigns in that article demonstrate why most websites look the same.

I actually went to each of them, they may "look cool" but they were utterly unusable. The hacker news one was horrific.

That's exactly what I mean. They're cool but impractical. In the end, most websites have actual content they want you to access, and overly cool innovative design can get in the way.

Imagine you logged into Hacker news tomorrow and they switched over to that design. I would assume they were being held hostage by hackers.

Like Hacker News. Most of us are on this website right now via the browser. It's the same interface it's been for years, and that familiarity increases the efficiency of understanding. Re-imaginging everything is a great opportunity for designers to showcase their skills, but wouldn't really improve users' efficiency.

Geocities and MySpace comes to mind when seeing the redesigns. Websites needed some type of convention for the average user to feel comfortable using.

The designs offered really remind me more of graphic designers' websites from the 1990s—confusing, impenetrable, but vaguely artsy.

Some people need to be reminded that design is not just being artsy. Design is for accomplishing purpose beyond screenshots to put on one's portfolio.

A website shouldn't be a book.

It also shouldn't be seizure inducing, retina burning vomit of random colors over a page. A website should be readable with the least amount of mind power given to stuff other than the main content.

Look at the gmail redesign. The modernized it. Used cool new material design and other modern goodies. And yet there is a gmail redesign hate thread on HN every other day.

When I am reading or doing anything, I want to spend my focus and time on the content. Not on the side distractions that serve no purpose.

A website whose purpose is to convey textual information and to be readable should resemble a book, or more likely, a magazine, because it has a similar function to the printed page.

"Convey textual information" is not the sole purpose of any website.

Websites have many purposes and functions, including of course interactive functions and connections with other elements both within the site and from other places.

Completely different to any book.

> "Convey textual information" is not the sole purpose of any website.

It doesn't have to be the sole purpose of any website for it to be the primary purpose of many.

For instance the purpose of this forum, along with many forums, is to convey primarily textual information in the form of comments arranged in a hierarchy.

The purpose of medium.com is to convey textual information in the form of blog posts.

The purpose of a site like Wired, newspaper sites and magazine sites, like their analog counterparts, is to convey textual information (albeit with a heavy graphical element) in the form of articles.

Websites expand upon the possibilities inherent to the medium of the printed page but HTML is a hypertext format, and its purpose (and therefore the purpose of most webpages) is to convey information in the form of hypertext. Interactivity and embedded elements don't detract from the textual nature of the web or its origin in concept or the relationship in the styles evolved with the printed page.

>Completely different to any book.

Different in structure, but not different in purpose, and therefore not completely divorced from the same design considerations.

Yes, we could design every website to look the way the redesigns in the article do, but would you really rather use the hacker news redesign presented in the article than this one? Would you rather use any of those redesigns?

Yup. It should use the power of the medium (the "magic ink") to become even more efficient at its function.

Which is the exact opposite of what this article tries to recommend.

A webside definitely shouldn't be a poster. The "nice" redesigns offered in the article "work" as visually interesting posters, but utterly fail as websites - they become unusable, they sacrifice function to improve the visuals. A book-like website may be suboptimal but is usable, but a fancy poster-like website is much worse than that.

A website is not a book, but just like novels have found an effective form, so websites have also found an effective form.

I'm sure there will be improvements on that form in the future, but when that happens, most websites will adopt it, and most websites will still look very similar.

The book analogy doesn't work. There is no point mentioning books.

A website is nothing like a book in form, function, purpose, appearance, use, lifespan, the list goes on.

You're confusing the equivalence. A "novel" is a type of book. A "website" is not a type of "something".

A website can be anything the owner wants it to be, with different functions and purpose. It doesn't need to conform to the other "novels" because of some perceived "effective form". That's the point I take from the OP's article.

Websites have not "found their form", that's just the reason you've chosen to believe based on so many websites using the same template.

It was never an analogy. I'm saying that a website should be like a website. Just like a book is like a book. And like books, websites have a purpose that determines their design.

A website can indeed be many different things, but a lot of websites have very similar purposes, so it makes every kind of sense that websites with the same purposes end up looking similar. Just like cars look alike. Or houses.

Of course websites with a different purpose are going to look very different.

I wonder if this is in a way tangentially connected to McLuhan's notion of 'the media is the message'. That is, what we really, often, seem to be trying to communicate is the form. The form is the primary. Notice how even the most blase and normative and familiar story is rendered anxiety generative when paired with a non-normative form (most David Lynch movies, for example). Whereas, even the most seemingly brutally intrusive content, is eventually normalized by what we might call normative story-telling structures (most so-called transgressive horror films, for example). This, perhaps, is why we react so negatively to form, rather than content, as we instinctively know that form will neutralize any potentially-revolutionary content. And, of course, this spills over into every other area. After all, the major reason sited by those on the analytic side of the philosophy divide most often cite the seemingly apparent obscurantism (form and the obsession with form among the likes of Heidegger and Adorno and Hegel and Kant) of the construction of continental philosophy as its major detractant.

So it's not what you say, it's how it's said. Which would go far in explaining why traditional liberal democratic practitioners always seems ultimately to fail in arguing against nazism, facism, etc. (That is, not only are they fighting with different weapons, but are on different fields on different days.)

So it might be possible to say that it is not design that distracts from the content. It is the design which is the content. The practicality is the reinforcement of an already-established form of cognitive mapping. In this way, it can be viewed as having parity with religion in society (though in some sense subordinate, as it is form, again, that underlies the function of the religious experience), in the sense that perhaps a given society is propagative-force for those forms on a kind of meta level, so, then, to change the forms is to change the society, the potential injection of a type of medical retro-virus (if attempted consciously), if you will, which, of course, as in the case of any external invasion, would likely trigger a sort of immune response.

True, there may be some absorption and incorporation of a rival form, but only rarely, and possibly, in parallel with biology, the more complex the organism, the less likely such would be possible. And, in any regard, absorption and incorporation would either have to be done is such a fashion as that only those aspects that do not create fundamental change are incorporated, which, in effect, means the intrusion is still neutralized, in the way a dominant culture an import cultural artifacts / aspects from another subordinate culture, a biological organism can, etc, or such absorption would give rise to a new form, meaning the effective annihilation of prior contestant forms.

(incorporation of pagan myth could be an interesting example of this phenomenon. Notice how the content can remain the same when encoded into the new Christianized form, and the same content ceases to be of the pagan culture, ceases to propagate the pagan culture.)

It would be interesting to see if form could be separated from content that it wouldn't then be possible to reconstruct, recognize, and/or recreate any given culture.

Websites are not art*

They are communication mediums. Either delivering communication effectively to the user, or to let the user communicate effectively with others, or do so bi-directionally.

Or they're tools to get work done - maybe even make other art.

In all cases functionality is most important - which is not incompatible with aesthetics. Aesthetics and usability are vital too.

But it does constrain somewhat the variety of your medium. And that's okay.

As LitFan said - it's the same reason cars and smartphones tend to look the same as well.

* Except when they are, and when they are they should be appreciated as such and graded as such, and there are such websites, but they are by far the exception.

> As LitFan said - it's the same reason cars and smartphones tend to look the same as well.

I agree, but for a different reason: they're all following the same design fads and/or using the same frameworks, which make certain designs "easier" than others. The content of the fads or the defaults aren't necessarily the most effective or functional for the particular use-case.

I would tend to agree on those points but I also want to believe that websites are an art form. While content strategy and user experience are now in a pre-thought context rather than an after thought, it obviously is about displaying something useful. But it almost stagnates new ideas from coming to the forefront it leads to us all getting swept into the game of 'creating the same layout,' all sites become a boring shell with different words.

> Websites are not art* They are communication mediums.

I'm not sure what you think "art" is, but art is communication. That's the whole point. That's all it is.

As the article (mis-)quotes, "Don't mistake legibility for communication."

Writing a text-only webpage isn't somehow opting out of art, any more than a brutalist building is opting out of architecture.

Why is everyone comparing websites with physical objects like cars, phones, laptops, and buildings? As soon as you do that, you've broken whatever conclusion you put forward next...

"...and because of the way cement is poured into the foundations of modern office buildings, that's why everyone is using the same website template".

Sorry, but you'll need to try a lot harder with your analogies.

They don't look the same; you can make an argument they look similar though.

"Fabian and Florian turned Hacker News into an interactive visualization. The social media site is a news aggregator, focusing on computer science and information technology. Its design is bare-bones but it has complex functionality for voting and discussions. Fabian and Florian have taken the existing structure and turned it into a typographic space of timelines and networks. The visual presentation is based on the sequence and connections of news and comments. They also connected their design to the API of Hacker News, so you can actually use it to read the site. View the redesign of Hacker News."

And the redesign is a usability nightmare. You cannot easily skip a topic you're not interested in, and you cannot even skip the comments easily which are emphasized. In normal HN, you can browse news.ycombinator.com and click on the links from there. You can open the original links and the HN discussions in tabs. Not in this redesign. Perhaps it is God's greatest gift on mobile (I tried it on desktop), I don't know. Examples of doing it different should be objectively better; this one doesn't appear to be such an example, so don't mention it [1]. Instead, go back to the drawing board.

[1] There's also a bug that clicking on No. and Pts. gets you to that user's hompeage. Which could be 1 or 76, but it has nothing to do with that userID.

I’ve tried it on mobile. (iPhone 8) The whole thing takes more than five seconds to load, and the UI is basically unusuable. Tried clicking at an article, comment text is almost impossible to read due to the broken layout.

Turns out, 95% of the Internet is used for practical purposes, and certain best practices arise...a balance between the content and the ads.

The other 5% is a self-expressive art project.

Part of me wonders if the author is trolling everyone by writing an article about the need for more unique design on Medium.com.

This person takes the time to write an infuriating rant but can’t take the time to understand human pattern recognition, the reason why the web is the way it is.

All websites looking "the same" is effectively evolution at work. Many website designs and layouts have been tried, and the "best" of them have stuck around.

This can be seen in any product (cars, smartphones, food - the list goes on).

From time to time there are designs that don't look like others, and those may or may not be met with approval. If they're liked, those designs (or aspects of them) get blended with the standard.

> All websites looking "the same" is effectively evolution at work. Many website designs and layouts have been tried, and the "best" of them have stuck around.

Yes. This is quite similar to asking why all newspapers look the same.

One of them is pink. It's quite distinctive, but has not spawned a flock of imitators in blue, yellow, etc.

The USA Today was the first to use color in the masthead (and on every daily paper's front page), and that has been copied by some.

Attribute improvements are adopted when they make broad sense (including economic). The same with color television, or improvements in book printing quality.

Comprehensive macro changes usually only occur when there is a rather extreme inflection point in technology that makes a true leap possible. Something that was previously impossible not only becomes possible, but makes the experience a lot better for the end user. That's why the iPhone approach murdered the old flip phone industry so rapidly, that technology inflection made a lot of new things possible that were very beneficial / desirable to end users. It justified a large macro change, which users had to adjust to.

I don't think this is true. I think designs get copied from websites which are successful (in all ways, not just design), because people have positive feelings with those designs, and also because using them requires a learning effort. It doesn't mean, however, that the designs themselves are best.

For a casual user, a design that requires extra learning effort to use it is objectively worse than a familiar design.

So for a generic website which cares about first-time visitors (as opposed to a specialized tool that people use often) following established standards and familiar UX paradigms is good usability design.

There are local path-dependent optima. There may be no -intrinsic reason for favouring these initially, but once established, their familiarity becomes their most valuable aspect.

Chirality, side of the road, alphabet, notational conventions, AC/DC, power frequency, mains socket design, measurement standards, sugar and DNA handedness, etc., etc.

Best was in quotes because it's subjective. If the people in charge of designing a product perceive a similar product's design to be the best, then they will copy it.

Similarly, if the people paying the people in charge of designing a product perceive a similar product's design to be the best, then they will pay those people to copy it.

I run a front-end agency that specializes in helping top-league branding agencies and designers code their websites.

Designers try to make sth fresh and out of the box all the time. The problem with these solutions are:

1. UX sucks

2. Don't scale to other resolutions (especially mobile / tablet).

3. EXTREMELY expensive to code comparing to "boring" web solutions (multipliers might be as big as 5-10x)

So most of the work we do is to try to make designers think with patterns ("boring" but the only way to fit in budget) and try to get as much as they can from STATIC design. Play with typography, key visuals, content etc. This is where their work brings most value for reasonable money.

The author of the article is a professor, so wouldn't expect realistic business thinking from him. He might be sad because web is boring, but usually no one wants to spend huge amount of money for prototypes where there's 95% probability that UX and stability will be worse than standard.

And there's a HUGE value in familiarity for users. That's why most of "creative projects" end up looking pretty similar at the end. I've seen many e-commerce sites which had "fancy design" and after months/years ended up with standard e-comm layout.

Also, web today is super hard comparing to 10 years ago. Number of things to think about multiplied by A LOT. Coding great static website with all things stable and looking great is a big task already.

It's like asking for a car with steering wheel in the back or with 3 wheels and paying for it 5x more. Do you want one?


Not saying "creative projects" are bad or sth, they sometimes are super coool. But I treat them more like "web art installation". Nice if you like them but not really practical.

None of the examples the author shared are usable. They're different just for the sake of being different. This sort of stuff only works for marketing campaigns that have a short shelf-life or maybe art museums.

I'm surprised a bunch of people who call themselves interaction designers put little thought into how people would actually interact with these "web-art" masterpieces.

Why is there layout on the web at all?

In almost every case, I care about the content, not the particular web site designer’s vision of the “experience”. I don’t want an experience, I want information. I would prefer the content be delivered to my browser with metadata describing what it is, e.g. “news article” or “comment” or “image of purchasable item”, and choose my own experience- that is best for me.

Browsers seem like a silly vanity for me- 90% of the functionality is about allowing a remote party to control my experience. I’d rather control my own.

> Why is there layout on the web at all?

For the same reason text in a book is arranged in paragraphs, and why magazines separate articles into headers, body text, insets, etc, rather than having all printed media be simply unformatted text.

Having navigational elements and content groups on a site (a "layout") with a predictable visual style allows users to more easily differentiate between the different parts of a site, their different functions, and to navigate those sections more easily. Typographic elements like headers, paragraphs, font styling (bold, underline, etc.) make text easier to read and communicate concepts like hierarchy and emphasis. Flow is a necessary consideration when incorporating non-textual elements such as images into a document alongside text. Modern sites must also consider how they will appear on multiple screen sizes down to mobile.

>Browsers seem like a silly vanity for me- 90% of the functionality is about allowing a remote party to control my experience. I’d rather control my own.

A remote party isn't controlling your experience, they're sending you documents in a markup language, and your browser is interpreting that markup. If you want to download and view raw HTML then you can wget the pages and open them in the text editor of your choice (I think RMS does something like that to avoid executing nonfree code), but you should realize that most people don't want to view raw metadata and have to write their own stylesheet and schema for each site, they just want to read the document.

I want to decouple layout from content, and control layout myself locally. There is too much active content and too much commingling of style directives and content in most modern web sites. I want the markup to be higher level and focus on describing the content rather than specifying style.

Looking at this and then the 4 examples they give, wow, I appreciate Bootstrap more (saying Bootstrap as it tends to be blamed for conformity of websites now).

The reason why Bootstrap is nice relative to what they give as examples is that while (usually) it looks pretty, the content is what you focus on and not the dissonance that the UI creates. In the case of the art ones, it makes sense as I imagine that was the intent, to flip the focus, so as an art exhibition, that's perfectly fine and good. But as a website, I wouldn't use it.

It reminds me of early days (around 1995. I remember buying No Doubt's Tragic Kingdom CD around this time and their band's URL wasn't a custom domain but something hosted by one of their labels, I believe. Different times). A lot of bands, with Nine Inch Nails being a bigger one, would go for these UI nightmare sites with, my guess, the intent that they're essentially a game of sorts seeking out hidden stuff. Those types of things turned me off (still liked the music though, just not the sites), but I know there are those who loved them so more power to them.

"Users spend most of their time on other sites. This means that users prefer your site to work the same way as all the other sites they already know."


For a website that's about UX, the actual UX of trying to read each of their articles is actually infuriating. I don't want to scroll. Why isn't the definition next to the name?

By that logic, all websites should look like Facebook.

You must've thought you were being contrary, but websites already do. They have a top navbar. They have sidebars. They have "edit/delete" dropdowns. They feed you content in a scrollable column.

Facebook is as vanilla-standard as it gets.

I'm not sure that's wrong. Look at the (temporary) outrage whenever Facebook does a redesign, even Facebook is allowed to not look like Facebook.

FWIW, from what i remember facebook's redesigns were never just a visual thing. they always added and pushed some functionality, made some things much harder to find, or quietly removed things entirely.

A similar thing happened when Last.fm did their huge redesign. The new website was all js now, and they quit all support on the forums, messaging, pretty much all the old ways of communication between users. And for me, that was awful, since i was a part of a music collective that communicated primarily on those forums. After that everybody dispersed.

So outrage at a redesign is not always unfounded or silly. Web companies tend to hide things in them.

> they always added and pushed some functionality

Most of the outrage I remember was around hiding and removing functionality: Chronological newsfeed, page customizations, network pages, etc

Most platforms already do.

Here's the concept of HN: https://interface.fh-potsdam.de/future-retro/HN/ if someone missed it

What the fuck is that? It certainly looks cool, but I have no idea what it is, what it's for, or how to use it.

Being usable and looking really cool often don't go together very well. I suppose the few cases where they do, are the mark of a truly great designer.

Not to sound mean, but I'd prefer all websites looking generic (as uninventive as it might be), if this is supposed to represent the alternative.

I am not the author, but I do agree.

This is completely unusable -is it supposed to be a parody?

I see this more as an example of what can be done. It's a bit like concept cars - they aren't supposed to be driven so designers take the opporuntiy to push the limits of what can be done. Perhaps that website, once the usability is improved, could be the beginning of a good idea, or perhaps it should be dropped altogether but it doesn't hurt to try new things.

It's the 4th example in the article.

Superimposing the text of one comment over another may look cool, but doing so means you gave no thought to the experience of the user.

Which itself is a perfect showcase for why most sites look the same. That redesign is virtually unusable, and goes so far into being 'art' that it just doesn't work as a discussion forum.

Still, maybe it's more the article writer's concepts of a unique website that don't work more than all sites being the same being a bad thing. Their redesign for Medium.com is nearly as unusable too:


I get the creativity in letting people adjust the length of words/negativity of words/number of most common words in an article, but like the Hacker News redesign, it just creates an unreadable mess.

You can make something different without making it unusable, and most of these redesigns fail miserably at that.

This is very impressive from a technical perspective, but for actual using it, it's truly horrible.

I can certainly imagine all websites looking like this in the near future however. Website designers all seem to love jumping on idiotic bandwagons that favor looking cool over usability.

This looks awesome, but it's completely unusable.

I think HK has one of the best (if not the best) design of any news aggregator sites out there. It’s perfect and I hope it doesn’t ever change. I can’t think of many websites that scale so well from a mobile phone to a desktop computer.

I'm surprised he even categorized HN as looking the same. This place is very streamlined compared to other news aggregator type sites, even pre-redesign Reddit.

HN is basically a vBulletin forum from 15 years ago with less functionality. Metafilter and Slashdot are similar and their looks haven't changed much since they started.

The HN Redesign probably happened because the API was easy to use by students and they wanted to experiment with "cyberpunk" style interface, for which a text heavy site like HN was perfeect. It isn't at all the type of site

Nice, but too cool

"Why do all websites look the same?" and it's a Medium post.

More seriously, I think the Internet has the reverse issue. Every website looks and is used differently. You don't have to figure out how to turn pages when reading a new book.

They don't.

Trivial example: Hacker News looks nothing like Medium. All blogs on Medium look the same though.

The mainstream web looks the same because it's all gone through the same funnel.

The web is split. You have the oldschool developer part (mostly textmode with the odd showoff project), and the mainstream part (full of GDPR banners, advertisements, cookies, tracking stuff, 10MB page loads, etc).

The GDPR pop-ups are incredibly annoying. The particular website has no obligation to comply with it. Most websites don't need to even give it a thought. It seems to just be a knee jerk reaction yet it is irrelevant.

I normally don't complain about downvotes, because honestly who cares. But I am curious why I do get consistently downvoted for simply stating that the GDPR is not relevant for a significant portion of the internet. EU users might not like it. But it's true. Most websites do not need a banner. They can absolutely flaunt compliance without worries.

Man, all of those models look like flash websites. That’s the problem. The unconstrained mess of deconstructed html, the Skeuomorph movement and reconstruction of architecture in the browser, the relationship diagram... these are the same explorations of early flash and dhtml websites.

It’s challenging to rebel against the dominant paradigm because everyone ends up with tattoos and blue hair and alternative becomes mainstream all over again.

I'd say these re-designs are inaccessible but people would get the wrong idea; inaccessible usually means not accessible to people who have handicaps of some variety, and that doesn't accurately capture the problems with these re-designed websites: They're not accessible to anyone, anywhere, regardless of how able they are, unless they're the designer themself.

They not only blindly assume you have the latest web browser technology and are willing to use it, heedless of security concerns and basic propriety, on relatively minor sites, they assume you've crawled up inside the designer's head and therefore see the site the way the designer sees it.

Which brings me to why sites "All Look The Same" (kind of a stupid thing to say, but I'll not fight it): It's a design language, something where designers and users can agree which elements do what, what affordances they have, and, therefore, how to use the site to get what they want from it.

None of the re-designs follow that.

None of the re-designs are respectful of anyone else.

These re-designs are a bad parody of design work, and make designers look bad by association.

I was with the author, until I saw the HN redesign. Looked beautiful, but not functional. Made me realize why HN / reddit / quora / etc use the format they do: it works.

> David Carson once said, “don’t confuse communication with legibility.” We should apply this advice to the current state of web design.

Granted I didn't go to design school, but efficiency and the ability to be understood are pretty high on my list.

Take example 2, not only does it scroll jack for a square pattern, there are things that highlight but aren't clickable, there's a gif that's chewing up CPU cycles, and this seemingly functionless page requires 58 requests, 48MB and somehow 1.2 minutes to load according to chrome dev tools, compare to loading the HN homepage in 1.1 seconds.

I'm not an advocate for turning the internet into a server of pure text files, but when art and usability are head to head, usability must win every time or your users will go somewhere else.

Carson was saying that as a rejection of minimalism. It was an embrace of experimental graphic design. He was specifically saying that "efficiency and the ability to be understood" were not always the most important factors in a design.

Please don't encourage these kind of interfaces to websites. A website is not a piece of art. It's a thing I interface with! Your fellow users and I, thank you.

At no point did the author argue that your local bank's website should resemble one of the examples. It is an article about unconventional designs on the web, of course he is going to show the most out-there examples from his students.

Personally I loved the cyberpunk feel of the Hacker News prototype, even though it's a usability nightmare. I like to view this stuff to engage my imagination and think about what's possible, not whether this design is ready to scale to millions of users.

> Why do all websites look the same?

This question puts two things in the spotlight:

1) the all too common software developer's ignorance for the art of design

2) the all too common designer's ignorance for the art of frontend software (which is usually data retrieval and entry, using common and learned UX functionality)

There often is a lack of humbleness regarding putting what I (designer) fancy behind what users need.

OTOH devs admire things like material design today or bootstrap in the old days as a way to "solve" design. Yet, those frameworks only touch half of what design is needed for (common UX patterns, don't come across as unprofessional) and fail at the other. I.e. there is no uniqueness, subliminal mood, branding, emotion or even recognizability in most pages/apps that use those frameworks.

The idea of letting A/B tests guide your design decisions gave us the many ugly things Google did in the past. And while they certainly improved on design, material design breathes the same spirit. The spirit of design as something "to solve".

Luckily with material design's CDK this spirit seems to be solved too :-)

The attempt at redesigning the web ends in unreadable nonsense, so I think that answers the question of "why" with "readability"?

Why shouldn't they look the same? They are tools. Usually for communication, and you usually want your visitors to intuitively know how to navigate your content. By all means, feel free to break conventions if you have a purpose in doing so, but for most brochure sites or information-based sites, you have no reason for such things.

Quite the contrary, when I set up a site for someone, I tell them to make it match their brand, but not to get overly creative other aspects of the design. A creative design may be interesting, but is that the goal of the site? To convince the audience that you found an interesting designer? Or do you want to communicate something about your product/project?

Simple, effective designs have evolved over the last couple decades. Designers should experiment, as that is their purpose, and it is how we avoid stagnation. Everyone else should use what works.

As a hacker with no design background, reading Design for Hackers [1] was life changing. I finally understood that design is just a set of rules and that really modelled well on my engineering mindset.

Add to that frameworks such as Bootstrap CSS and opinionated website builders such as Weebly and I finally overcame a major limitation in my skillset. I could now actually build what I wanted to build, knowing that it looks decent and I can focus on what is under the hood.

If the web all looks the same, it is probably because it is built by people like me who need a rigid framework to work within.

[1] https://www.amazon.co.uk/Design-Hackers-Reverse-Engineering-...

There's some irony to having this posted on Medium which is about the most generic looking site on the internet.

My thought exactly, even when I clicked on the link.

I think publishing this in Medium rather than on one of those designs of their students, kinda proves why most websites prefer to look boring.

Indeed -- made it hard to read the actual article knowing it was posted on Medium

Interesting article and there is some truth to it. Although to be fair, the counter-examples provided, while interesting and nice, are almost all unpractical, illegible or hard to navigate. A lot of work went there, sure, but the results kind of weaken his otherwise good points.

No, just no. I understand that to push the current state further, you need to think outside of the box, but what you call bland, I call essential to communication.

I find the David Carson quote (confusing communication with legibility) to be ridiculous on it's face. If your intention is to communicate information, it needs to be legible, clear, and understandable. If your goal is to make beautiful designs, fine, but understand and state your goal.

Not a single website redesign here would ever work, and would be offensive as a user trying to consume information.

Design is about embracing constraints for the medium to solve a goal. This is about removing any constraint to make a pretty picture. Great, looks nice, but I won't ever use it. So what was the real goal?

The answer to this as of the last 5-10 years seems to be to simply be 'responsive design'.

Since we need to lay out a web page in several different ways as-is, it seems as though evolution has taken care of what seems to be a very 'natural' design pattern at this point.

I think it's because web design used to be a branch of graphic design, and now it's not - it's now a branch of web development.

Designs for websites were made in photoshop by creatives. Now designs for websites are made by front end engineers, or UX consultants.

Exactly. Making an artistic and unique website is simply too expensive. Think of all the lost revenue. Web design has become a science of analytics crunching, conversion optimization and retention. The author seems to show total ignorance for human pattern recognition.

"We have the capability to implement almost every conceivable idea and layout. We can create radical, surprising, and evocative websites. We can combine experimental typography with generative images and interactive experiences.”

Depends on the type of website, for example, if it’s a banking or airline ticketing website, my tolerance for creativity for the sake of differentiation is non-existent. I am not passively consuming media, I came to the site to complete a task and want to get it done with the least friction and advertising shoved in my face as possible.

Similarly, this is why I’ll never stop using ad-blockers and keep mentally noting which sites employ scummy UX dark patterns so I can spread awareness of alternatives to friends and family-- cause I consider 'dark patterns' both psychological opportunistic and predatory.

People use technology as a means to an end, whether to check their savings account balance, chat with friends, or look up information, introducing friction in the form of 'avant-garde' creativity without UX research/data on the impact to usability, is a recipe for disaster on user engagement metrics and revenue stream intimately tracked by upper management to justify continuing to pay the salaries of designers, developers, product/project managers/BSAs, etc. and to keep the lights on at a majority of tech companies (in my experience). Thoughts?

Ironically, this is a medium post. Why didn’t the author make their own website to demonstrate how creative they can be?

The real answer why sites are homogenous is because everything has to be mobile friendly. There are only a few options that work, unless you’re going to completely redo your layout and css for mobile. The truth is, modern front end is hard, so that’s not a good option. Ui designers also save a ton of work by keeping the mobile design similar to the desktop one.

Because Designers don't know how to build them without using frameworks. Because Customers don't know what they need, but very much what they want.

Most frameworks are built around readability and a few standard layout configurations based on the principles of magazine page design and typography.

What do customers need that modern web frameworks don't provide, and why is it wrong for customers to want the web to be familiar and unobtrusive?

I still think there's a lot of subtle UX innovation going around. You just need to know where to look. Some more well known sites include:

- Medium

- Atlas Obscura

- Techcrunch

- New York Times (visual / interactive story articles)

- https://baymard.com/research - ecommerce UX research

Not all UX design content mediums have been invented yet, there's billions of permutations, and different content requirements. Different color, typography, color psychology, shadows, icons, animations, etc all change the way a story is told. A website is just a content medium, much like a video or image.

However, at the end of the day, content will mostly be the same. It's just CSS and JS changing the look and feel of the site. Example, one static HTML file and multiple CSS implementations: http://csszengarden.com.

To the untrained eye, every site does look the same. My definition of good UX is this - minimize the total distance your eyes have to adjust, and maximize the amount of content captured. Without being overly fatiguing (usage of whitespace, eye relief, etc).

These stylistic experiments on breaking away from the 'by the numbers' routines immediately reminded me of Vonnegut's "pity the readers" advice, acknowledging the inevitable limitations of artistic liberty of a writer.

7. Pity the readers

They have to identify thousands of little marks on paper, and make sense of them immediately. They have to read, an art so difficult that most people don't really master it even after having studied it all through grade school and high school --- twelve long years.

So this discussion must finally acknowledge that our stylistic options as writers are neither numerous nor glamorous, since our readers are bound to be such imperfect artists. Our audience requires us to be sympathetic and patient readers, ever willing to simplify and clarify --- whereas we would rather soar high above the crowd, singing like nightingales.

Pity your average web surfer.

Because rectilinear layout is well-supported ... but completely non-intuitive and artless. Options to do otherwise are painful and time-consuming and immediately run into browser incompatibilities.

Just try laying out a series of iconic buttons in a swooping curve. Or try laying out some text inside a circle ... or just making a circle, a triangle, a leaf-shape... or fitting an image between two columns of text ...

The tools to do all of this were available 25 years ago in desktop-publishing apps. Creating a beautiful page took design skills, but the layout was not done with CSS but by drawing on the page and making adjustments with handles.

Not by endless hours of fussing with CSS version xyz and grids and flexes and compromises by sacrificing original intent.

HTML layout is absurdly painful. Most people with deadlines don't have time to spend hours to make it work in one browser, switching to another browser and SCREAMING, and tossing it all out and returning to rectilinear boxes.

That's why.

This is a case where an Artist has confused themselves with a designer. Designers solve problems, often by using existing conventions to help an interface/tool/object/room be instantly usable and understandable.

An artist flexes their creativity and communicates thoughts through imagery/experience/sound.

I also recommend the author's follow up article "Balancing Creativity and Usability" [0].

[0]: https://medium.com/@borism/balancing-creativity-and-usabilit...

If the search area is not on top right, everybody will loose time looking for it. Web is not art anymore. It's mainstream. It might be sad, but well...

Just look at all the rant there are when an app change a bit of it's design. You want adoption by users ? Make it easy for them.

Is it frustrating for designers ? I bet so. But can you imagine living somewhere where every shop have its own design and way of sorting things ? Imagine a supermarket where products will be stores in alphabetic order ?

Seems a good idea, you won't have to search for the good line any more. But it'll have to survive long enough to customer to adapt.

Well, now imagine every supermarket having those disruptives ideas.

Maybe in a future where we'll all have time for it, if we won't have to work more than 5-10 hours a week. Why not ? Until then, let's keep it simple.

I don't understand what the author was aiming for with this. "Designers are lazy and unimaginative; why don't they all make their websites look like these unusable, ugly explosions in a type foundry?"

I wouldn't even ~~be able to~~ want to read a static document that looked like those examples.

I was going to blame Bootstrap for this but then found this on the site: https://expo.getbootstrap.com. Some really good examples of innovative design even using good old Bootstrap.

I think he left SEO as a reason out of this article. We're now building websites as much for humans as for google crawlers to read them, which means meeting standards of accordance in order to even participate in the SEO game. H1, H2, H3 tags, layout, and content headers don't have to look the same or be layed out or structured the same, but many sites have deviated from these formats at their own peril only to return to the warm embrace of conformity. Not to mention that Amp is basically a design standard as much as a content standard.




Every time I try to do something original I go back to something generic because of the following reasons:

- Supporting all devices makes custom design really hard, with need to design for phones, tablets, PC in all kind of browser window size/ratio. Different screen size, dpi and input type. - Accessibility is really hard with out of the ordinary designs. - A website must "give out its content" in seconds. You cannot establish a relation with the user like, say, in games.

Truth be told, when a website is totally unconventional, it has my attention for a short time, generally on account of novelty.

Once that has worn off, unless the site is exceedingly usable, memorable or memorable, I'm out.

It's taxing to have to visually search, analyze, scaffold, and navigate a new mental model of how I expect a website to behave.

And therein lies the problem, people have expectations of websites, and it is generally in their best interest to conform to existing standards.

In the past: holy grail layout

Now: masonry layout / infinite scrolling

I don't think anyone has mentioned it apart from the fact most people would find these confusing, but these types of layouts look like accessibility nightmares, and for what benefit? Making something look a bit nicer?

I mean, it's fine if the main goal of the website is to be artistic, but as others have pointed out, that's not the main goal of most websites, it's communicating information.

Do they look the same? Or does the differentiation just take place on a lower level of detail that is hidden to the author of the article?

I prefer same type design as opposed to going off piste..

In fact, if I want to leard about something or somebody I will often choose Wikipedia primarily over something else because I know it's format and can get straight to digesting the information I came for.

(The issue of me using a crowdsourced platform as a prioritised source of something else for a different discussion)

Donald Norman does a good job explaining why in The Design of Everyday Things. I am sure most of you have probably read it, but if you haven't... https://www.amazon.com/Design-Everyday-Things-Donald-Norman/...

Swiss style design solved much of bad design.

If you divert from that you better have a reason and be good at it or just layer something unique on top.

All webdevs use the same tools that push certain directions. No webdev has the time to be creative and try out new things so they build each website the same way.

Diversity is found near exclusively in the domain of homepages built by people not in the business that didn't make their website to put on a resume to proof they can use whatever tool is currently in.

Depends on the purpose of the website. If the purpose is to convey information, then there will be quite some similarities between websites that share that purpose.

If the purpose is to use the web as a medium of artistic expression, then you can see some variety. But that's likely a small percentage of websites by purpose.

"Type Primer" was a revealing and relevant read for this topic. Design is about functionality, and good typography and alignment can often-times get you there. Colors, shapes, etc. are interesting for art; but when someone needs to use what you've built, ensuring functionality is clear is crucial.

Matthew Butterick wrote about this 5 years ago:


Even publications that produce a beautiful paper product seem to be willing to accept a garbage design when they post it online.

If you compare web sites that are hand crafted to web sites created with a CMS and template, of course the hand crafted sites will be more unique. If you want to master design and user interfaces, you must first learn all the rules, but then you must break them!

Off topic but these days I load every website in reader view. It’s amazing how annoying custom fonts, sizes and colors become once you get used to reader view for everything. Doesn’t matter what font, font size, color or layout you use - I won’t see it and I love it.

I do this too. I scan the first paragraph or two and if I decide it is worth reading, I hit the reader button.

I really couldn't care less about modern web design -- Just show me the text.

Modern design is just so tiring. A wide range of font sizes from tiny to huge, impossible contrast ratios, font choices that stain the mind, media that distracts from the message - all add up to tiring me out mentally.

You can set every page to open in reader view automatically by pressing and holding that icon I think.

"Somewhere between $20m and $100m ARR,

All marketing sites converge on an identical design"


Because of Bootstrap. Everyone uses it. Every old website that looked different in the past, now does a UI refresh (by just using Bootstrap out of the box with minimal customisation) and becomes another one of those the same looking websites.

Outside of the HN comment bubble, nobody uses Bootstrap. And it doesn't explain the examples in TFA.

Speaking anecdotally, up to around 2016 every other website used Bootstrap or Foundation, often with minimal customization, just because it looks and works well out of the box on all devices. Now it seems like it's getting better, but generic UI components are an extremely relevant aspect.

Promethease is a good example. Years later a bootstrap website signals to me "it's about the functionality, not the design". https://promethease.com/

Before that everything was jQuery UI. Government websites still use a ton of it. All of Colorado's web portals are built with it. And now things are shifting towards UI frameworks "plus more", like Ionic, which adds a bunch of design/devops workflow and cross platform functionality.

Lord help them when they realize that all designed things tend to look the same.

Within a particular generation, all things tend toward a familiar design. There is a herd mentality to fit in with the crowd, allowing for only minor differences.

All of the real content lives outside the context. http://sebastian.io/monte-fischer-design-school/

Those examples make me very, very glad all the websites 'look the same'.

Indeed. The web is more consistent than its ever been, and I think that's a very good thing - it means that I'm not often surprised when I scroll down and the page moves to the right (I'm looking at you, Apple...)

I can't think of another website that even remotely resembles superbad [0] and that's quite okay with me. [0]: http://superbad.com

Man, all of those models look like flash websites. That’s the problem. The unconstrained mess of deconstructed html, the Skeuomorph movement and reconstruction of architecture in the browser,

"People spend most of their time browsing other websites"

Yes, it's a truism, but it explains why site designs converge. Fewer people will be willing to learn your site's custom layout and behavior.

None of these “redesigned web” examples are balanced or readable. While I’m not convinced today’s design is the most optimal, I’d prefer them over any of these futuristic examples

"Generic fonts"

The web had ~ 5 system fonts before Web Fonts were available in 2008 / 2009.

Now the web has many thousands of the best fonts ever created by the world's best foundries.

Yes, another "cool" thing slowing down loading/rendering the site. "Waiting for fonts.googleapis.com"...

Designers: "Let's make it avant garde"

Business People: "Let's make it fast"

Developers: "Choose one"

why do all cars look the same? why do all magazine covers look the same? why do all the graphics on cable news channels look the same?

These are questions that do not require an article worth of words for analysis..

Says the guy writing on medium.

They don't, you're reading the wrong sites.

It's a broader problem. Why do all phones look the same?

Why do all cars look the same? Houses? I could go on.

Readability and legibility are important, too.

Because everyone decided to use a framework.

It helped standardize design.

I'm against this but I can see the benefits of the same layout.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact