Design early. Earlier than that. | Hacker News Reader

Design early. Earlier than that.

helloform.com

36 pointsfredoliveira29 comments

"I’m not going to remember you use a sharded MongoDB infrastructure as your database"

But you're sure as hell going to remember if the site is down for 11 hours.

Part of good, early design is working with engineers to make solid, appropriate technical choices that will affect user experience.

Do you think it's a serious blow for a consumer facing site to be down a day? Enterprise, sure.

I do actually, more for the effect on credibility than for anything else. Just as users have "itchy back buttons," I think they also have "meh, this is the last straw, I'm going to try out the competition" moments.

Can you cite a case where this has actually transpired? Obviously in the extreme example where the site is never available I can see it having an effect. But spotty downtime I've seen actually have the opposite effect - users become more loyal. On one forum I frequent the "500 errors" are sometimes looked forward to since they usually signify a big news event is happening. Reddit and twitter are other sites that don't seem to have been negatively affected by spotty downtimes. I do think that regularly (excruciatingly) slow performance can have an impact though - one of the reasons I left digg with V3.

I don't think companies are going to publicize cases of losing customers due to downtime. But just to turn it around a bit, think of someone coming to a site for the first page. What's the user experience for them if the site is down?

i recently quit quickbooks online because the site went down for over 24 hours within the first week of my signing up

I tend to think no. For a small site with few users, no one will care. Once a site has become large enough, downtime with press coverage seems to create a virtuous cycle of social proof, e.g. Twitter, Facebook, etc.

I agree with the majority of the article. I think that design is very important and is essential to user experience on your site, however, i dont think design is more important than engineering like the author states.

You can have a functional site with no design, but you cant have no function and just a design. I'm not saying you have to choose between the two here, just saying, engineering is the meat of all websites, without it, you have nothing.

You can have a functional site with no design

I would disagree with this. Without design of some sort, you don't have an interface at all, and thus no functionality; all you have is inaccessible backend code. Regardless of whether there was intent to design or not, if there's a user interface, there's design.

I think that's actually where a lot of design failures happen: They start with the misconception that something was not a design process and thus didn't need design thinking. By the time design is brought into the picture, there are often underlying inflexibilities and assumptions about use patterns that have real effects on user experience, and the design process is working at too high a level to solve them.

Is design is more important than engineering? Neither good design nor good engineering can succeed without the other.

The difference is that any engineer can build an interface, even if it sucks, but the reverse is not true.

^^ this was basically my point.

I don't disagree with the point you're trying to make. Obviously you have to have a product. But there is no such thing as 'no design.' The opposite of good design isn't 'no design,' it's bad design. The problem with trying to differentiate a product's functionality from its design, is that from a user's perspective, they're the same thing.

>You can have a functional site with no design

That's the thing. You may have all kinds of killer functionality, but if your design gets in the way of my ability to use it, you effectively don't _have_ a functional site.

You're right. I was particularly agressive in my wording to get my point across. I will say, however, that if you come up with a static prototype based on proper design thinking, you'll be in much better shape to go around and ask people "would this make sense to you?".

We both know that the key here is balancing these 2 things out, so we're in agreement ;-)

Whenever I read a blog post like this, my reaction is "can't we just agree that both are crucial and have that be our answer?"

Even if one day someone establishes a sound mathematical proof that design is more important that engineering or vice versa, what will we have gained from that? Reminds me of the joke about the most important organ in the body...

I used to think the brain was the most important organ in the body.

Then I thought, look what's telling me that.

(emo philips)

What's the joke?

Something like "the one that just failed".

I quite agree with the suggestion that a startup be founded with a designer. IMO, founding engineers should ideally be able to wear product and design hats, but a dedicated designer probably doesn't probably doesn't make sense in all cases.

As TamDenholm said below (paraphrased), design is important, but no engineering == no product.

well said ... a designer isn't going to be utilized 24/7 in my mind but the engineers will.

I'd be curious to find out how a company like 37signals keeps its designers completely occupied though.

They operate in teams of one designer, two developers: http://37signals.com/svn/posts/2099-2010-the-year-of-the-pro...

I understand that, I was just wondering how they find enough stuff for them to do everyday, I'm a little ignorant about how you would keep a designer occupied full time working on just one product.

Much like programming, design is never finished. I can imagine a designer working full time on improving the experience of an application like Basecamp. Even more if the designer works on stuff like SEO, analytics, A/B tests, etc.

Just like how there are a million lines of code that could be improved, there are always a million design tweaks that can happen. Nothing is ever perfect the first time.

As a designer, I often have just as much work as the engineer. I also implement a lot of the front end including css, html, and JS so it's not just pure photoshop.

"I’ll go on a limb and say that design is more important than engineering"

No it isn't. Let me explain.

Design is incredibly important. No doubt about it. But I hire designers by the week, perhaps by the month. I hire them by the design. I have longterm relationships with them. They are a core part of the project but they are interchangeable, if push comes to shove. And I don't need them engaged every day. I really don't.

Most companies are not Apple. Apple has design at its core because it creates tangible products. I loath Jobs and all he stands for -- I detest bullies -- but I saw an Air for the first time (years ago) and had to touch it. Shiny. Hardware design is not software design.

Personally, I don't like the metaphor of "experience" that is currently en vogue. I have never had an "experience" with a site or application, but I have used them. They are functional things, designed for purpose, to fulfil an objective. They are not an orgasm. I wish. They are do not evoke a tangible emotional response. I have never been moved to tears, even close, by viewing a webapp. I do not weep at Google's minimalism YMMV. Graphical designs, as applied to computer application, are not an experience. They really aren't.

I can buy that interface design. And I love to. I love the process, the passion of the artist, and the attention to detail. But that's all it is. And I can get it in many places.

But the skill required to build the infrastructure and networks of complexity that satiates all the requirements of my business plan, and things I have not thought of yet, and things I don't even know I need yet, now that takes skill and experience that cannot be defined, let alone hired by the week. That takes immersion and involvement. That takes a large degree of dedication, foresight, and yet more dedication. It almost take faith in the belief that unforeseen obstacle can be overcome. All of them.

Rewinding, design is incredibly important, a central part of any initiative that's outward facing. But top of the pile it ain't. And it never will be. I can buy it by the hour. I can buy it off the shelf. I hope I never have to.

" I’ll go on a limb and say that design is more important than engineering or ops today"

So, a designer thinks design is most important. Oddly enough as an engineer, I think engineering is most important. Anyone know an ops person who thinks that ops is most important?

Seriously though, this article has a good point. It's a shame the author ruined it with the above. Is it really a good idea to start these kinds of arguments? Trying to convince people your department is better than theirs is not only childish, it's bad for the whole company.

Instead of having a single person working on UX/design everybody should be thinking about it. While eating your own dog food (you are aren't you?) think about how many mouse clicks you can remove, about how many settings options you can remove, etc.

This is important in more than just startups and software/websites. This is important is pretty much everything. Designing for human factors should be incorporated as early as possible in pretty much everything, from software, to physical user interfaces (with actual toggles and such), to boring old registration procedures (done with paper!). If you accept the premise that you should be designing for human factors at some point, then you should be applying it as early as possible to save you the most money, time, and generate the largest return. Your minimal viable product should have undergone some sort of "design" when it leaves the door. It doesn't have to implement everything, or even most of it. It just has to be ready to grow into something that does.

I've always wondered how ycombinator handles design. Do they have designers on site , do teams hire designers to get it right or do they count on the teams to do it on their own. I'm no good at design and I don't know anybody who's good at it. I just browse Themeforest and buy one if I need one.

Reply on news.ycombinator.com