Hacker News Reader

> A real shame. Google is losing some major talent - which appears to have only been squandered over the last three years.

When I started at Google Nov. '06 Doug had created some awesome Gmail mockups that really took the design and functionality above and beyond and -- having heard about how ground up Goog worked -- I was thinking those designs would be acted upon and built in the coming months and I was excited about using that Gmail.

When I left end of last year, the mockups were still being iterated on, and the only thing that had been built based on those mockups were the buttons. Gmail's just too big, as are search and ads, and UX is too disorganized and outnumbered to have any say.

Are any of those mockups public? I love seeing good design ideas.

No, unfortunately, not that I know of.

> A real shame. Google is losing some major talent - which appears to have only been squandered over the last three years.

And Doug Bowman's not the only one - Jeff Veen left recently too, who headed up their UX group after Google acquired Measure Map and applied the lessons learned to gAnalytics.

It sounds like he's a poor fit for Google. The data-driven decision making process really is central to Google (at least in Search; I dunno how it is in other departments), and if you're working at cross-purposes with that, the culture will and should reject you.

That doesn't mean he's a bad designer - I'm sure he's not. But Google occupies a particular place in the market ecosystem: many of its users want a UI that's clean, spartan, and utilitarian. There're undoubtably market niches for stuff that's more visually stunning (wasn't one of PG's ideas for startups "A form of search that depends upon design"), but people interested in that would be far more economically productive seeking out new markets of people that are underserved by Google's spartan design rather than trying to force something visually impressive upon Google's existing users.

Nowhere does his post imply that the reason he's leaving is because his design aesthetic clashes with Google's desire for "clean, spartan, and utilitarian." It's perfectly possible to meet those goals and not be ugly, like, e.g., iGoogle and Gmail are. I consider most of Apple's designs to be "clean, spartan, and utilitarian".

Requiring user testing for shades of blue and thicknesses of border is the height of silliness. I'm reminded of those room makeover shows they have on TV. At the beginning, the designer tells the couple one of the colors they've chosen. The couple is aghast. "What, I can't stand magenta. No way!" Sometimes, the wife looks like she's about to cry. At the end, magenta is used and the design is breathtaking.

There is no way user testing has gotten efficient enough to beat a designer at that level of detail. Moreover, users often don't know they want something until a designer shows them they do. You can never get to those designs by testing blues and borders.

Why is that silly? For Google it takes no time to implement and, because of their traffic, no time to test. What's the downside?

You just spent three days of a teams time over what the designer already knew.

Assuming the designer is correct, when you have that much traffic and easy access to real data you don't need to make assumptions.

The point of the blue-bar incident was that the designer was not correct, and a different shade of blue (which most people inside the Googleplex thought was inferior) generated measurably more revenue - more than the designer's salary, and certainly more than the time spent experimenting with different shades of blue.

The best thing about a data-driven approach is that it gives room to be wrong. And we're proven wrong all the time: basically every meeting, there's some measured data where we go "that can't be right" - and then we either discuss ways in which we might've misinterpreted it or say "Okay, assuming this is right, then what do we do?" It sounds idiotic, but really we're (humanity in general) the idiots.

I wonder if there is any research/reasoning in why one shade of blue would generate more revenue than another?

Malcolm Gladwell had an interesting podcast about people's ability/inability to express what they really mean and feel. He uses the Aeron chair as an example of a product that was despised/considered ugly but eventually embraced for it breakthrough functionality/aesthetic.

I can see how data-backed design decisions can produced locally optimized results, the "ideal" shade of blue, logo size, typography, etc.. But I wonder if on the whole, all those optimizations work together optimally, as a wholistic experience?

[edit] Gladwell's Podcast: http://itc.conversationsnetwork.org/shows/detail230.html

Well, that's the point of multi-level A/B testing. You try all of the combinations to isolate the impact of each varying feature, but you are always testing 'wholistically'. They don't just throw a green bar up there on its own and see who clicks on it.

The original article about the 41 shades of blue says that Google decided to test the 41 shades because another team pushed a slightly different shade and it performed better.

So, it looks like the designer didn't know.

I'd contend that's because most designers don't optimize for performance, they optimize for aesthetic. When those two are in conflict which should win?

It sounds like Bowman thinks the latter, while Google thinks the former. As a stockholder I'm thankful that's the company's attitude.

> Requiring user testing for shades of blue and thicknesses of border is the height of silliness.

No it isn't, if a designers ideas cannot withstand a little objective scientific testing, they aren't worth using.

> Moreover, users often don't know they want something until a designer shows them they do. You can never get to those designs by testing blues and borders.

Which is the whole point of objective testing, do the design, throw it in front of real users and see how it performs next to the original design.

I think what's really going on is that more often than not, designers want what they like rather than what works best. Engineers want what works best, and Google has an engineering culture. Form does and rightfully should take a back seat to function.

I think that's right, but a little overgenerous to engineers. We have our hangups, too.

Say, which do you think is better, emacs or vim?

Even in a data-driven environment oftentimes the argument gets pushed back from "what to do" to "what to measure."

You can't measure emacs vs vim like you can the effect of color or layout on conversion of an app. When measuring is trivial to do and results are easily seen to directly impact revenue, there's just no excuse to not do it.

I agree. I'm just saying this isn't an "engineer" versus "designer" issue. Engineers have just as many "irrational" opinions as designers.

Emacs vs. Vim, Ruby vs. Python, or whatever.

And those things are quantifiable to some extent. Take a random sample of engineers, give them identical tasks to complete, and vary the editor.

From the way I understood it, he was suggesting that testing for shades of blue and border thickness is nitpicking. An equivalent nitpick more familiar to us would be testing whether using tabs or spaces for code indenting improves coding efficiency.

If it affects conversion, then it's not nitpicking, tabs vs spaces isn't a valid comparison because it can't be so easily measured.

Google search may be clean, spartan, and utilitarian, but I just don't see those adjectives applying to Gmail. OK, I'm not sure, maybe Gmail is utilitarian, but it sure doesn't seem either clean or spartan to me.

> It's probably a big reason why Google hasn't had a truly innovative design in a long time.

I would argue that they do have a truly innovative design. The fact that it doesn't look new and challenging doesn't preclude that. They have a design that's nuanced in thousands of ways to improve their business. Their design is reminiscient of Victorian architecture, every last detail polished off until it was perfect. Having your business fine tuned to that amount of detail is not something most companies can claim.

"Without split-testing, your product tends to get prettier over time. With split-testing, it tends to get more effective." - Eric Ries

True -- but design isn't a science... Or it's only part-science anyway...

You might as well say you can take code and with minor increments make it perfect over time.... Sounds logical, but we all know it doesn't work like that.

Sure, tuning will take you a long, long way -- but occasionally you need to break the rules to leap forward.

I don't disagree that sometimes you need to make radical changes to make true progress. But I think those radical changes need to be put through the same testing that small changes are put through.

I would much rather say "We are sacrificing 1% conversion to take a bold new direction with our design" than to say "We are redesigning beacuse our old site looks old and that's bad."

I think you are confusing functionality with design - even though two are closely related. Google is a landmark in both functionality and design - but former was the one that hauled latter onto the scene. IMO google has had one of the worst designs/branding ever before gmail era, which is in stark contrast with it's peak in functionality. This makes it a unique blend that stands out.

I honestly think a lot of their 3px, 4px, or 5px decisions are actually an exercise of finding local maxima for their current design.

If they're looking for interfaces that are truly better, they would do far more testing on innovative/educated directions and iterate each of those to see how good they can get. Instead you just make one version as business-helpful as possible, which may be much weaker than your true potential.

I like the idea of data, but there's only so much progress you can make with the same designs in the same mold, tested pixel by pixel.

> "Yes, it’s true that a team at Google couldn’t decide between two blues, so they’re testing 41 shades between each blue to see which one performs better. "

Actually the evolutionist in me thinks this is pretty cool but I can still understand why a designer would want to quit.

I guess it depends on if you trust the majority making the right decision. I think often that just leads to mediocracy.

41 shades does seem a bit of a gimmick. I doubt it really matters that much, or that most people can even tell the difference.

It's impractical for most businesses, but when you're at Google scale, a 0.1% difference in conversion can be huge.

I guess it depends on if you trust the majority making the right decision.

When you're running a business and the decision is whether or not to convert, trusting the majority is the only decision. If others see it as mediocrity, so be it.

I think the point he's making isn't about the correct way to choose a shade of blue, it's that real design work encompasses far more than picking colors.

That's true. Design is not about shades of color and pixel widths. But it also sounds like he resists having his designs put through the ringer of empirical analysis.

It's Wal-Mart vs. Whole Foods. Which would you rather be?

Both answers are fine, of course, the important thing is the "why."

If you can test A and B, you'd test 2 colors. Or maybe 3. But when you have to test many things (not just shades of blue) and you run tests of A, B, C, D, and so on, all the time... I'm sure they could test HUNDREDS of shades if blue just as easily as they can test 3.

From a designer's perspective, this is an example of engineers over-thinking a problem in a space they don't belong. It's absolutely stupid and a complete waste of time for the simple reason that electronic color is not consistent. A majority of the users will see a different shade of blue depending on their display settings. A very small percentage of the population will see the color that's intended.

And if he's just being sarcastic, my opinion still applies. :)

This lack of consistent colour presentation is actually an advantage for Google's approach over the designed based one. Their approach is optimising across the population of 'end colour presentations'. Even if 50% of users happened to see it shades of aqua then the best 'aqua' presentation would also be incorporated into the result with a 50% weighting. A designer would find it very difficult to weight all the of presentations proportionately even if he knew about them whereas Google's method doesn't even need to know about the variations. Let's face it you never really know what peoples internal representations of colours are.

Of course this type of testing only works where you can measure the results and the number of options to be tested are low compared with the number of tests. If you look at the whole layout of a page then then there are just too many dimensions for this type of testing to be practical. A designer could perhaps argue that changing the background colour, the border colour and using a bigger red button would be the best option. A designer could also argue that they are optimising for 'overall experience' rather than some measurable metric like clicks.

I think this all goes to show why both designers and 'measurement geeks' need to work together on these things.

You make good points that I partially agree with, however I still don't believe this is optimal for a few reasons.

End-user design is about creating a usable experience and often deriving an emotion to make them want to do it again, a process that relies on color combinations, layout, typography, etc. -- the complete package. I'm less attached to GOOG webapps because they lack subtle things such as sound feedback. I've found sound to be critical in software usability, far more important than very minor color discrepancies. That is, when software goes outside of very basic operations, and requires more attention from the end-user.

(Fighting about the shade of blue makes me puke in my mouth at the thought of such a corporate cliche).

Furthermore, this type of color testing would need to be extended to cover cultural differences to be truly effective. Colors will trigger completely different responses in China, Japan, France, etc, often the exact opposite of American counterparts.

Honestly, I think the engineers have a place AFTER a design has been released and there is data to mine and analyze that takes into account the entire product presentation as a whole. Interjecting them in the design process too soon and giving their opinion overriding power is a mistake. It's akin to slapping memcache on your back-end before you've done an ounce of optimization on your queries.

> I've found sound to be critical in software usability

Very little is worse than web pages that use sound. It's no accident that most sites don't use sound, your tastes are just way outside the mainstream on this.

> Interjecting them in the design process too soon and giving their opinion overriding power is a mistake.

The success of Google clearly shows it's not a mistake, numbers don't lie, but designers are full of bullshit they can't justify.

Desktop apps are used by more people on a more regular basis than any web app to date. Web apps are getting there, but there not there yet. Sound, when used appropriately, is a very critical component to user experience and more importantly, usability — hence why most operating systems use sound feedback. Apple's new Nano uses sound to extend its usability tremendously, and other examples are endless. (By no means do I assume sound should be forced, always a user-defined option).

Agree or Disagree: Twitter clients are an example of sound feedback hooking users attention (among other things), where the web was failing to do so.

Success of Google and 'design sensibilities' is not really an argument worth correlating; there are companies like Apple that master both engineering and design... tremendously successful. Furthermore, one could argue that Google Calendar is a near exact rip off of iCal, no credit to Google except putting it on the cloud and syncing it with their own software suite. I think influence is great, but goog engineers are not necessarily to credit for design and usability in this and other products.

You're moving the goal post. No one's talking about desktop apps, you said web apps. Web apps are a different space and sound is just not appropriate there, and it likely won't ever be.

There are many cases that justify when to use and when not use sound on the web...

I would say it's the goal of most web-app builders to mimic desktop apps b/c the result is a familiar environment for the end user. Software/UI design is an established practice, with many studied and proven methodologies for handling complex user interaction that has been unfamiliar to the web in former years.

That said, I believe it's inevitable that web-apps will mirror their desktop counter-parts in time (many are doing so already)... the exception being, highly-specialized web apps.

> It's probably a big reason why Google hasn't had a truly innovative design in a long time.

I think I'm on the other side of this. In the past, I would have said that Google had the best designed interfaces on the Web. Lately, I find myself wishing there was a way to opt-out of their 'upgrades'.

Most of my complaints are about their recent attempts to improve search results by automatically 'fixing' the queries I enter, but I'm guessing the border in question is the bold blue box surrounding the Gmail message list. I hate it.

Presuming he was the one arguing for making it the way it is (visually bold but distracting and unnecessary), I would side strongly with those who wanted to actually test it on users instead of trusting his judgment.

But therein lies a big danger. Design, why a given design works better, has _also_ to do with the culture and the taste of the many. The taste here not so much in the PG sense, but more in what people are used to. Similar like how the many think PCs cause a lot of problems because Windows has a lot of instabilities (Please, just trying to make a point. You know what I mean, that's enough.). This taste will change with the time, what will Google have as a response? Or: when will PG write of the death of Google's design?