In my A/B testing experience I've never seen aesthetics be a major factor, and only rarely have I heard of it mattering for anyone else. That experience has been echoed by other people I know who do A/B tests. Instead messaging, usability, clear calls to action, etc win consistently.
But if you build a site with those goals and without worrying about aesthetics, you're going to wind up with an ugly site.
However that said, there are a couple of amusing cases I know about where deliberately making something ugly was an A/B testing win. Photoshop found that by setting ugly default fonts they push people to poke around the application to change the defaults, which results in a lot more usage of other features. Similarly Freshbooks redid their invoicing some time ago, and found that people overwhelmingly preferred the new version, yet conversions to paying customer were better when people were given the old version and forced to figure out that they could change it on their own. (And yes, most of the paying customers actually did change it.)
But in general aesthetics simply don't matter. And any theory based on the assumption that they do is wrong.
My favorite usability story was at my expense. I was watching a demo for a tool crib software - essentially, it allows people to gain access to tools like a library. The one that impressed everyone the most
* was in Visual Basic 6
* was ugly as anything I had ever seen
* had an incredibly cluttered screen
but everyone raved about it when they left. Why? It was far and away the most usable software we had seen. Instead of thinking about the pleasantness of the screen, they were thinking about workflow. Every time we asked, "can it do X", it did so, and with far greater elegance than we had hoped.
Even with the clutter, it was easy to use. Now, when I look at designing a user interface, I think of it with completely different eyes.
What many of those examples really shows is that people will put up with amazing amount of shit to save money. Take the Ryan air site for example. Everybody I know hates that site with a passion and do everything they can to avoid both the site and the airline. Most people check every other available option before reconciling themselves to dealing with Ryan air.
However since they offer a competitive product at very competitive prices people put up with horrible usability and service. So no, ugly design doesn't sell, cheap sells, and if you're cheap enough people are willing to overlook ugly.
HN is a nice example. Imagine if you let a designer loose on it, the disaster that would ensue. It would be eye candy for sure, but we'd all be leaving in a hurry.
Especially for stuff that is used frequently less design is usually better, 'b(l)ing' is for kids with too much money to spend on their cars. Tools have to be functional first.
They can aesthetically pleasing, but only if the form follows from the function they have. Once you can't remove stuff any more the design is optimal.
>"Once you can't remove stuff any more the design is optimal."
That's when the design is minimal, not necessarily optimal.
Optimal is subjective: optimal for whom, performing what operation?
Having been on HN for a couple of years now (yes I've changed accounts - part of an experiment in aggressive commenting) I find that collapsible threads a la reddit seem to be really missing. There are other things I think reddit does more optimally too - like providing hints on formatting with one click (and no page change) of the post. I could go on (and on) but you get the idea.
I read HN despite its design and usability, not because of it. Sure a bad designer would make it worse, but I think a great designer could do wonders for HN.
That's only because you're created a straw-man definition of design. Don't buy the hype: design is about influencing the subjective state of the audience. Sure, beauty is one kind of experience, and maybe it gets a lot of attention, but it's certainly not the only thing designers care about.
I think the main skill that designers bring to the table is empathy. Unlike traditional artists, who can safely ignore what other people think, designers have to be interested in how something feels to their audience. This is important if you're building things for people who aren't like you and don't think like you, and you want to get their attention, get them to think what they're buying is useful and usable and yes, maybe beautiful if that matters, and the people who made it are trustworthy and reliable.
Design is about communication--it's a kind of language. Who could argue that words don't matter, only function matters?
This seems kind of obvious, but OK. How do you know what a customer needs? How do you know whether your idea meets that need? You have to look at the world from their perspective begin to answer those questions.
Follow-up question here, as I have just come from attending a focus group this evening. Is what you describe a typical part of the job description of a Web designer (what prompted my question above) or is it really the job of, say, a focus group facilitator?
A focus group facilitator doesn't try to influence the audience. And focus groups usually don't even empathize with the customer, they simply write down what the customer wants, if they prefer blue or red, not why they prefer it and what that says about their overall goals.
Coming from a user interface design background, I thought this article posited a couple of false dichotomies which were perplexingly echoed in the comments here. Their suggestions appeared to be targeted at either amateurs, ad professionals, or flash jockeys, because a good designer would likely be taking this advice already. The gist of the article is summed up nicely by a commenter here: "Usability trumps design every time." I find that perplexing, because my goal as a designer was always to find the most usable solution through design.
1. Value -- a good designer will take the time to understand the the target audiences. Many target audiences aren't looking for value, they're looking for status or something just as intangible. Implementing a visual design which communicates something the target audience isn't searching for is closely akin to pricing your product incorrectly. In fact, design approach is almost completely analogous to retail product price stratification.
2. Trust -- this again depends on audience. Sometimes "labor of love" is the most relevant trust mark, but in other cases it's "well-established" or "industry leader."
3. Accessibility -- this is kind of a 'duh' for any experienced web developer. Most design-oriented site you go to are built for modern browsers only, but that's because all their visitors are using Firefox. Any good web designer still knows how to fix an IE6 margin-doubling bug, for example.
4. Flexibility -- um, it really is possible to design sites with good UI and flexible content. Would anyone here make the argument that Craigslist would have failed with proper whitespacing, appropriate font selections, and logical visual hierarchies? Those can all be accomplished with a single stylesheet.
5. Function -- this really is the point of design. Most good designers would tell you that. Design without utility is bad art. I love user interface design because I can build things that look good and work well. The two go hand-in-hand.
I may be reading too much into the comment that "in general, aesthetics simply don't matter." If that's really true, do you have a good explanation for Apple's success? How about Nike?
"Would anyone here make the argument that Craigslist would have failed with proper whitespacing, appropriate font selections, and logical visual hierarchies? Those can all be accomplished with a single stylesheet."
Many successful websites (Google, craigslist, ebay, wikipedia, myspace etc) are both highly functional and divergent from "good design practices".
What I find troubling is that I don't believe any of those sites would have been constructed in their successful form by a design professional. We don't fully understand how people's perceptions are shaped by the presentation of information.
Having appropriate whitespace/fonts etc produces a clean looking site which conforms to a very particular aesthetic which has no proven connection to the success of a site and may substantially harm it.
There's no doubt that some of the most successful sites have been pretty ugly. But to imitate the ugliness smacks of cargo cultism -- reproducing side effects in hopes of receiving a similar result. To me, the commonality shared by the sites you mentioned isn't their aesthetics, but rather that they represented either a radical innovation or a radical improvement over existing services. I DO believe this was a result of their engineering focus in the early stages, a side effect of which was poor aesthetics, but I don't think it's a clear indicator that an engineer-driven product cannot also benefit from smart design. Google's move to break the Altavista portal mode of thinking and present the user with a large, isolated search box was first and foremost a design decision.
I think Mint is a good example of design driving success. Mint is largely built on the Yodlee platform, but they made the interface their primary focus. I don't get any information from Mint that I can't get on my bank's website. But since signing up for Mint, I check my accounts twice as frequently, classify all of my purchases, track my financial history, and set budgets. I didn't have to wrestle myself into these habits, either -- they came naturally because Mint made them easy and available. And honestly, I would never have trusted my bank account information to a site that looked like early-days eBay.
Surely Apple and Nike's successes aren't completely dependant on design. Apple in particular are at least masters of the product launch and PR in general. They also make an above average operating system.
Excellent article. The only thing I would quibble about is the word "ugly". I'm not a designer or a visual artist of any kind, but I don't find Amazon ugly. It simply doesn't approach either extreme of the beauty spectrum.
Excellent article. I agree that distracting elements on a website impact conversions and should be avoided. However, there is definitely a difference between a well planned "clean" website and an unprofessional "cluttered" website. The product being sold (if retail) is also a major factor in the design.
A more professional-looking (clean) website is also more important for a services business (accountant, health care, restaurant, etc.), where your website is a representation of your business.
How a site's look affects traffic patterns and revenue is an incredibly complex and well-researched nebula. For example, Marissa Mayer talks about how a couple pixels of whitespace and subtle differences in color shades drastically affect traffic on Google.
I don't buy that un-categorically "if your website looks BMW-fancy your visitor is going to assume BMW-pricing". Rather than opining one way or another about what look is "right" for sites, I'd be interested in the process of finding that look.
independent of this article making a very good point, i appreciated the grouping of 'java applets' with html5 as an example of 'cutting edge' tech on the web!
I found the quote from Scoble ironic - "We trust things more when they look like they were done for the love of it rather than the sheer commercial value of it." - meanwhile, Scoble has gone quite stylized, what with his new logo and all.
aesthetics aren't meant and shouldn't be expected to improve your usability. aesthetics have a role in cases where appearance is important (the example used was bmw).
"We’ve battled designers and CMO’s day in and day out for nearly a decade but overwhelmingly following the 5-rules laid out above drive results that simply win."
aesthetics are not bad, although the tone of the article makes it seem like ui designers are bashed when they shouldn't be. both practices serve their own purposes, and usability should not be confused with ui design.
And, don't forget that simple and ugly is not arrived at by not designing. It is designed to convert, designed to elicit a certain visceral response. The definition of design is not "beautiful."
For some situations maybe you want to say "this is really expensive" and in others you want to say "this is really useful." In either case you are designing.
Despite the title, there's not much here about "ugly" sites ... perhaps it a bad choice of words. In the examples he gives, I wouldn't consider all of them ugly.
Plus, some of them (Drudge and MySpace) don't have ecommerce "conversions" which is what the rest of the article appears to talk about.
From my experience with PPC landers, this is almost certainly the case. Getting someone to click an ad works better if the site is ugly. I am not sure on why but my guess is less distractions and perhaps 'get me the hell out of here.'
their example with the 3 variations for ordering flowers is pretty biased. the other two have the "order now" button likely below the fold of most email clients. consequently it has little to do with the overall design, and more so about the placement of one button which should have been obvious if the email was actually designed by a seasoned designer. some things are elementary.
bad way to close a biased post that doesn't acknowledge the value of design; design is form AND function. the real problem here is that designers have been shelved by inexperienced (read: affordable) i-downloaded-photoshop-and-forget-to-learn-the-fundamentals.
Wow this article is incredibly stupid. My favorite moment is this :
"They scaled to huge numbers of visitors with tiny staffs – keeping your site flexible enough so the CEO can change the homepage content may not be ascetically appealing, but it sure does beat a static beautiful website.
A website that’s easy to change, update, and experiment on is better than one that relies heavily on advanced CSS, Flash, images etc that you can’t change quickly. "
Since when are good design and flexibility exclusive ? Where do you see any sign of such a trend on the internet ?
And sometimes bad usability is what sells too. That's why you have TVs that come with remotes with a million buttons on them, it makes the buyer feel like he's got something powerful, and by extension, makes him feel powerful. If your target audience is interested in feeling powerful and achieving a sense of mastery, there's a way to design for that. If you want the audience to feel like they're getting something for cheap (Amazon, Craigslist, Ebay), there's a design for that.
Ugly design is still design though, it works in one context and fails in another. But that doesn't mean you've freed yourself from the tyranny of aesthetics, you're just using the aesthetic that works.
There's a great temptation for programmers to conclude that if they can't do something by themselves, it must not matter, and anyone who claims otherwise is probably a corporate drone or something. Maybe this works in some cases, but that doesn't mean it's not also a mistake.
It's like that joke about the guy who lost his keys in the alley behind a bar one night. His friend finds him on the sidewalk looking under a street light, and asks him why he's not looking in the alley. The guy says "It's too dark in the alley, I can't see a thing!" The moral of the story: when faced with a difficult real-world problem that you don't know how to solve with your current skillset, it doesn't make a whole lot of sense to try to change the problem into a different one that you do know how to solve. Maybe you don't need good design, but if you do, you should get a designer, and ignore blog posts about how pretty things are for girls.
But if you build a site with those goals and without worrying about aesthetics, you're going to wind up with an ugly site.
However that said, there are a couple of amusing cases I know about where deliberately making something ugly was an A/B testing win. Photoshop found that by setting ugly default fonts they push people to poke around the application to change the defaults, which results in a lot more usage of other features. Similarly Freshbooks redid their invoicing some time ago, and found that people overwhelmingly preferred the new version, yet conversions to paying customer were better when people were given the old version and forced to figure out that they could change it on their own. (And yes, most of the paying customers actually did change it.)
But in general aesthetics simply don't matter. And any theory based on the assumption that they do is wrong.