I don't think the companies you mentioned should even be considered exceptions to the rule. Facebook and 37signals both practice self-centered design. 37s uses all their own products internally, and I imagine most Facebook devs are the right age to be on Facebook. By testing their designs against their own expectations, they can approximate the UCD process with less overhead and a much shorter turnaround time. Their success actually has a lot to do with their ability to internalize UCD, and is most definitely not a good example of why UCD is broken.

Their approach can only work when the target audience and the devs/designers are drawn from a relatively homogeneous group with shared core values around which the application can be built. In cases where the target audience isn't likely to include a lot of devs/designers, traditional UCD methods are still the best way to ensure that the needs of the audience are addressed.

Posted from the new ixda.org
http://www.ixda.org/discuss?post=30642

producing reliable, repeatable results, which is what most businesses need and want

How do we know this?

Take a popular field of contention: an endless coterie of businesses tried to follow their "tried and true, reliable and repeatable" processes to compete against the iPod, iTunes, iTunes Store, etc. The results have been not just minor disappointments, but colossal losses in the billions, forcing the vast majority of them exiting the business altogether.

Look at an anti-UCD company like Apple. When the chips were down, they bought SoundJam for iTunes; they went to Tony Fadell for the iPod; when everyone and his brother was telling Apple opening retail stores were an insane idea, they built and re-built a completely functional prototype; when everyone else was shipping customer support to India, Apple brought direct support and training right into the stores; when every pundit claimed devices without floppy drives, FM radios or physical keyboards were doomed they brought us the iMac, the iPod and the iPhone, etc. Non-trivial problems require contextual, problem-specific thinking.

"Repeatability" opens the door to copy-ability and with the ever-escalating speed of competition, if a company doesn't have sufficient differentiation, it's dead meat.

-- Kontra
http://counternotions.com

Correct. And they continually iterate their products, which are easily changed, based on user feedback. I think sometimes we may lose sight of the world of tangible electronic products with UI's that run on embedded software and maybe even no internet connection. It seems like we wouldn't want to narrowly define "interaction design" to mean "web page design." With that in mind, there's probably something to be said for the talented interaction designer taking a prudent route in non-web situations and testing their designs prior to launching a product with millions of dollars on the line and no way to put the genie back in the bottle.

-al

On Jun 23, 2008, at 9:13 PM, Spencer Nowak wrote:

I don't think the companies you mentioned should even be considered exceptions to the rule. Facebook and 37signals both practice self-centered design. 37s uses all their own products internally, and I imagine most Facebook devs are the right age to be on Facebook. By testing their designs against their own expectations, they can approximate the UCD process with less overhead and a much shorter turnaround time. Their success actually has a lot to do with their ability to internalize UCD, and is most definitely not a good example of why UCD is broken. Their approach can only work [trim]

On Tue, Jun 24, 2008 at 12:13 AM, Spencer Nowak spencer.nowak at gmail.com wrote:

I don't think the companies you mentioned should even be considered exceptions to the rule. Facebook and 37signals both practice self-centered design. 37s uses all their own products internally, and I imagine most Facebook devs are the right age to be on Facebook. By testing their designs against their own expectations, they can approximate the UCD process with less overhead and a much shorter turnaround time. Their success actually has a lot to do with their ability to internalize UCD, and is most definitely not a good example of why UCD is broken.

I was wondering about that myself, but I guess folks are dispair(ag) ing of the user research phase and focused usability testing in these cases. ?

As counter examples, I have personal experience with software companies focused on software devs who turn out, let's say, less than usable products. The engineering (and sales and marketing, to be fair) focus on function overrides any focus on other aspects of user experience like usability, findability, and desirability. They can be very successful in terms of the business despite that.

So I think these folks, mostly "Web 2" types, are doing something different, and I agree with you, Spencer, that they are using certain aspects of UCD (or is it SCD? : ) . I saw a talk by Ryan Singer that made it very clear that they followed UCD principles, even if they weren't out doing formal research and usability testing. I don't know about Facebook, and Google now, if not originally, is certainly using UCD principles as well. Heck, I think someone just recently was talking about their usability labs, which sounds pretty UCD to me.

And I tend to think Allen's point about less-easily-changeable products is important as well. Perhaps it's not UCD that's a problem but an unwillingness to adapt it to one's environment and needs, e.g., excising aspects that are less necessary because you, say, already deeply know your audience. ?

Are there any advocates of the idea that UCD is broken out there that can clarify what they see the problem with UCD to be?

--Ambrose

I'm sorry but one of the main goals of UCD is to avoid "self-centered" design. Once you go that route in any way, you are not applying UCD methods. Unless you consider users beyond your 4 walls, it ain't UCD. Simple.

Ignoring our best examples of successful design feels like bad business. It is standard Business practice to study case studies to learn from them. Avoiding these seems impractical.

Why the heck doesn't anyone talk about how Google does (today) and Amazon has for a really long time DO use research. Google does mini AB tests out in the wild all the time, and so does Amazon.

Now, I would like to hear from the host of UCD practitioners here about their top of market efforts they work on and how UCD was the champion that got them there? and which processes they used.

-- dave

Posted from the new ixda.org
http://www.ixda.org/discuss?post=30642

I'm sorry but one of the main goals of UCD is to avoid "self-centered" design. Once you go that route in any way, you are not applying UCD methods. Unless you consider users beyond your 4 walls, it ain't UCD. Simple.

If 37signals really is designing exclusively to "scratch their own itch", then they are the users, and it could easily be argued they are indeed practicing UCD principles.

That said, I don't believe for a second that they can truly say they only design for themselves. Maybe it was true a long time ago, but now that their livelihoods depend on their customers, it would be foolish to stick to that mantra.

Why the heck doesn't anyone talk about how Google does (today) and Amazon has for a really long time DO use research. Google does mini AB tests out in the wild all the time, and so does Amazon.

So, running AB tests means they're practicing UCD? I would hope they do a whole lot more than that.

Now, I would like to hear from the host of UCD practitioners here about their top of market efforts they work on and how UCD was the champion that got them there? and which processes they used.

What about people using other approaches on top-of-market products and how those methods got them there? Are you interested in those stories, or do you only care about UCD successes?

-r-

As counter examples, I have personal experience with software companies focused on software devs who turn out, let's say, less than usable products.

This is hardly a counter-example. There are approaches other than UCD that are every bit as valid and every bit as successful. Not practicing UCD does not necessarily mean you're letting engineers make all the design decisions.

-r-

Are there any advocates of the idea that UCD is broken out there that can clarify what they see the problem with UCD to be?

I suppose that would be me.

UCD is filled with problems. Here's a list.

I could go on.

Even those who often advocate the practices included within UCD, such as Jared/UIE (who researched the best web teams to see what they have in common), fully admit that most of the time these things are done wrong, and done poorly.

Usability testing has flaws as well, but should still definitely be done when it makes sense.

For the record, none of this means a non-UCD designer has no understanding of human behavior on the web. I've personally sat through hundreds of hours of usability tests, study psychology in my spare time, constantly engage in conversations with people about their computing experiences, etc. I just think that it's far better to focus on activities rather than people.

Yes, this can mean talking to people who perform the activities you're designing to support, but in many cases, you can become a SME on an activity without ever talking to another person—particularly when designing web apps.

From there, I rely heavily on metrics—click paths, click stats, etc.—combined with an iterative design process, to determine how people are using an app and improve upon it. No method is perfect, but in my experience, UCD has far too many flaws to come even close. And considering there are so many great apps out there designed and built without an ounce of app-specific user research, I think it's safe to say you can indeed create something wonderful by throwing traditional UCD methods entirely out the window. -r-

I don't have time for a long response, but some quick comments:

It's interesting that some of the same people saying "UCD is dead" are the same ones saying they don't know what it is and are constantly trying to define it. How can it be dead if we don't know what it is? Seems like a lot of the argument is around semantics and granular definitions instead of the broad concept of what UCD is.

For me, all this simply means there is not a standard way of doing things that is alway repeatable - sometimes UCD is not the right approach and more a visionary design strategy is best. Apple, some Google products, the other products mentioned in discussions here are all examples.

However, when the problem is more complex than simple text based search UI, or playing music on a device, then clearly it can be really advantageous to pick the appropriate UCD tools and techniques and use them to one's advantage. It's contextual and making blanket statements about any approach always working or not working seems unwise. You don't always have to do personas to do UCD. And UCD research doesn't have to be this monolithic gigantor of a time and money hog. You can use super cheap and lightweight methods and get great results.

Industry leaders who use UCD: Salesforce.com, Autodesk, Google, Amazon.com. Cooper's products, Adaptive Path's clients and products, AA Razorfish. Those firms I believe are examples of great design, and each use some flavor of UCD, as far as I know.

I don't think any method should be wholly discarded - as designers we should have a big 'ole toolbox of approaches, and know when and why to use certain ones.

UCD is not broken in my opinion. Just my two cents. Great discussions here recently.

Jeff

On Tue, Jun 24, 2008 at 9:50 AM, Robert Hoekman Jr robert at rhjr.net wrote:

Are there any advocates of the idea that UCD is broken out there that can clarify what they see the problem with UCD to be? I suppose that would be me. UCD is filled with problems. Here's a list. User research is horribly unreliable, either because it's done poorly, not done to a great enough extent, or focused on the wrong people. Focusing on niche groups can mean alienating audiences that could otherwise use the product but weren't part of the research. User research can be horribly expensive and time-consuming, and I have [trim]

Hi All:

In a reply on the original thread (IxDA), David Malouf said:

' UCD is a collection of methods, not the act of "thinking of users".'

I think that is the core of why this discussion goes on and on.

If all UCD is, is a collection of techniques then of course they will become antiquated in time as the profession moves on.

However, I do not think of UCD as "a collection of techniques" or even the 'act of "thinking of users." To me it is a philosophy that grew out of the dissatisfaction that many felt with the way software was being developed in the early days of computing. Much software was (and sadly still is) designed by programmers who were not successful in producing usable or desirable products. Much design was also mandated by business people who made decisions based on what pleased them or would forward their specific business goals. Sadly, this too often happens.

UCD grew out of dissatisfaction with the outcomes of these development practices and was much more than simply a collection of techniques. It was, and is, a philosophy that argued that we need to focus on users' needs tasks and activities, their mental models, minimizing their learning curve and similar issues. The techniques that were developed over the years are ways to implement this philosophy.

You would think that caring about the user would be a no brainer but that was not, and still is often not, the case. Corporations are not relationship oriented. They are not benevolent. They exist to make profit and pleasing their customers and employees is a secondary consideration at best. So getting attention for UCD has been a difficult process.

Today the web and the availability of mobile devices have fundamentally changed things. As the web has become a major channel for connecting with prospects and customers, there is much more awareness that you need to please your users to succeed. That's a good thing.

The evolution of the web has also altered the way we think about user interactions. It is no longer about one user in front of one computer consuming the information parceled out by a centralized IT command and control structure. We are much more about community, user generated information, and complex social interactions. In that environment, there is no doubt that we should rethink the techniques of UCD which are often cumbersome and may not yield as much as we would like.

So, why is this all an issue?

We still have a long way to go in convincing the world of the importance of what we do. We are finally getting some traction as the business world sees advantage. We need to present a simple and comprehensible face to the external world and focus on developing the field. Whatever differences we may see between approaches like UCD, ACD, Ix, IA, Ux are only valuable when they lead to clarity and common understanding, not when they lead to confusion and hairsplitting.

In my opinion, every interactive design should be useful, usable and desirable. Whatever techniques produce that result are worth understanding and using.

So taking the position that UCD is just a collection of techniques and not a philosophy about what's important to creating superb interactive products will surely lead you to discount it over and over. Personally, I find that a bit boring.

Charlie

Charles B. Kreitzberg, Ph.D.
CEO, Cognetics Corporation

Reading this conversation is making my head hurt. It seems like a waste of time debating the definition and importance of such a vague term as UCD.

In my workplace people through our 'user centered' and 'user experience' largely without any real thought or concern for the user. Frankly, most of our initiatives are built with great user indifference.

Some better questions to debate might be:

1 How do we accurately identify our users?

2 How do we get better, faster and more relevant data regarding what will work for users?

3 To what extent should we consider the user? Certainly technical constraints and business models must remain a concern. How to we blend our user data with the other two and maintain singular vision for product?

Lastly, and I guess I will put Robert on the spot here, if you don't believe the current state of identifying, researching and applying user data is up to snuff, then why not try to improve methods? An end around the entire issue does not seem like a real solution. Given the choice, is getting into shape not a better route than just waiting for the double bypass?

Mark

On Tue, Jun 24, 2008 at 1:53 PM, Charles B. Kreitzberg charlie at cognetics.com wrote:
Hi All: In a reply on the original thread (IxDA) , David Malouf said: ' UCD is a collection of methods, not the act of "thinking of users".' I think that is the core of why this discussion goes on and on. If all UCD is, is a collection of techniques then of course they will become antiquated in time as the profession moves on. However, I do not think of UCD as "a collection of techniques" or even the 'act of "thinking of users." To me it is a philosophy that grew out of the dissatisfaction [trim]

Lastly, and I guess I will put Robert on the spot here, if you don't believe the current state of identifying, researching and applying user data is up to snuff, then why not try to improve methods?

What makes you think I would do all this complaining without offering an alternative?

http://www.peachpit.com/guides /content.aspx?g=webdesign &seqNum=352

Pick it apart all you want, but this approach has worked remarkably well for me, my employers, and my clients, in project after project, for several years now.

-r-

To play Amen chorus to Charlie, with whom I strong agree, let me add one more important thing UCD is, perhaps the most important thing:

UCD is a political stance, a position of political advocacy on behalf of what some may call at worst an oppressed or often overlooked group or class or people: users, co-authors, navigators, creators, etc.

And whenever you talk about political advocacy, POWER is part of the equation. Add power to the mix, and the stakes go up, cuz you can give nominal lip service to just about anything, but transferring REAL power, well, that's a much more radical act. A collection of methods can be easily de-fanged, made innocuous to any of the other existing power groups anxious to hang on to their turf (engineers, advertisers, stakeholders, what have you) . A collection of methods can be hidden behind, like "Hey, we do UCD, we did our due diligence!"

And UCD must then necessarily have, underlying everything, a position of political advocacy, to find ways to give users voice, to bring users and their empowered social groups into the conversations, to allow them to build virtual worlds in their own image, and to advocate always on their behalf. And more importantly, it must put its money where its mouth is, and make a difference for users.

In the immortal words of Dr. Seuss from The Lorax:

"I am the Lorax, I speak for the Trees! [users!]"

Chris

On Tue, Jun 24, 2008 at 1:53 PM, Charles B. Kreitzberg charlie at cognetics.com wrote:

Hi All: In a reply on the original thread (IxDA) , David Malouf said: ' UCD is a collection of methods, not the act of "thinking of users".' I think that is the core of why this discussion goes on and on. If all UCD is, is a collection of techniques then of course they will become antiquated in time as the profession moves on. However, I do not think of UCD as "a collection of techniques" or even the 'act of "thinking of users." To me it is a philosophy that grew out of the [trim]

In my view, when we talk about UCD, we are talking about tools and as such can be used for good or for bad. UCD makes sense when it comes to interactive products, so we better know who will use the product and what it really needs is essential, after all we are not working on something for his own use. But UCD is just one of the possible approaches, can not be seen as a rule.

Everything is relative.

-- Eduardo Loureiro
Interaction designer
http://eduardoloureiro.com
t: 55 31 8836.6520

Mapa Digital
http://mapadigital.net
t: 55 31 3291.5811

Nice article, but it does not really address the primary problem with the current state of design research.

You can not effectively 'gear up' and do user research in the web world (or any other evolving moving market) . It moves to damn fast. And on one of the only issues that I will side with the agile folks, big design up front is too slow when research is involved. I understand that research is best when purposed and targeted to answer specific problems. But business, across the board, does a terrible job of knowing it's users.

No accounting department worth its crust waits until the end of the reporting period to gather records.

User research should be an ongoing and exhaustive effort on the part of any company introducing products into the marketplace.

1 That it seems expensive is not an excuse. User research is not a cost. When done right it is an investment geared towards reducing risk and increasing value - both are near and dear to most any MBA's sense of self preservation.

2 That others don't do it does not fly either. User research leads directly to opportunities for marketplace differentiation... which leads to larger margins. Again, increased profit can be a pretty good motivator for MBA's.

3 Most market leaders fall (when they fall) at the hands of disruptive model upstarts. User research can expose features that users do not have an interest in.

Mark

On Tue, Jun 24, 2008 at 2:12 PM, Robert Hoekman Jr robert at rhjr.net wrote: Lastly, and I guess I will put Robert on the spot here, if you don't believe the current state of identifying, researching and applying user data is up to snuff, then why not try to improve methods? What makes you think I would do all this complaining without offering an alternative? http://www.peachpit.com/guides /content.aspx?g=webdesign &seqNum=352 Pick it apart all you want, but this approach has worked remarkably well for me, my employers, and my clients, in project after project, for several years now. -r-

Nice article, but it does not really address the primary problem with the current state of design research.

And that would be?

You can not effectively 'gear up' and do user research in the web world (or any other evolving moving market). It moves to damn fast.

Agreed. Hence the discussion in Part 1 of the series about it taking too much time.

1 That it seems expensive is not an excuse. User research is not a cost. When done right it is an investment Hmm. I'd rephrase that. Designing a high quality experience for a valuable product is an investment, regardless of how you achieve it. User research by itself is not an investment—it's a cost. If it's a cost that contributes to the high quality user experience, then your right on, but it often doesn't. In fact, it can have a negative affect on the user experience for people outside the researched audience, and even those within it. 2 That others don't do it does not fly either. User research leads directly to opportunities for marketplace differentiation... which leads to larger [trim]

Sure, but there are other ways to achieve market differentiation. User research is one way, not the only way.

3 Most market leaders fall (when they fall) at the hands of disruptive model upstarts. User research can expose features that users do not have an interest in.

Then how is it that these disruptive upstarts, who often have no time or money for user research, are creating products so great that they're causing market leaders to fall?

-r-

I've written a post awhile back about UCD, citing some issues and how it's more important to consider UCD as an overarching philosophy balancing multiple centers, driven by the simple yet powerful premise that people matter (it's quite shocking how many companies don't truly get this) . And hinting that there needs to be deeper conversation by ixd folks around "(digital) product development", not just around UCD in and of itself, but part of a much larger and more complex activity and environment of practice.

http://www.ghostinthepixel.com/?p=95

More fuel for the fire... :-)

Uday Gajendar
Sr. Interaction Designer
Voice Technology Group
Cisco | San Jose

It's worth nothing that a near majority of breakthroughs in what we consider good interface design were driven by people with none of the trainings or backgrounds we obsess about here. They would almost all describe themselves primarily as programmers, engineers or entrepreneurs.

Bravo, Scott. Your entire post is dead on.

I often wonder if some of these UCD practices exist primarily because there aren't enough visionaries-slash-revolutionaries in the world. As in, these practices are an attempt to provide a recipe for people who are not necessarily chefs—so that a lot of people can produce good designs and quality products instead of only those who are naturally great at it (Jonathan Ives, for example).

There's a big difference between a "drummer" and "someone who plays drums". There's also a big difference between a "master chef" and "someone who cooks well". I'm sure you can see the analogy.

Scott also said, "We should be studying how people with little of our training and expertise are able to achieve what they did".

I agree 100%, but I also wonder if this is how UCD, GDD, and others came about in the first place

-r-

Mark:

Your comment...

"Yes, programmers and business people can mess things up - I do agree with you Charles. But they're also on the list of people that made the big leaps in design and UI happen."

...is completely on target. I started as a programmer and migrated to UX because I was so enthralled by computers that I wanted to bring them to everyone.

Sometimes I get frustrated by developers and by business people too. But being married to a business thinker, I've also learned how much depth there is to business as well.

I believe that the most progress will be made when business, technology and user experience design are aligned and synergistic. That's why we need to share our vision with others in a constructive way and learn from them as well.

One of my professional goals is to help developers understand the UCD (or whatever we want to call it) design approach. What I've realized is that to teach developers what I know, I need to learn from them as well. It's been a long time since I created code and the field has shifted so much that I've become almost technically illiterate where I once was competent.

Like many on this list I value innovation. But I have learned, is that not everyone is an innovator. There are many competent people who want to understand how to do good design. They may not be breakthrough thinkers nor do they aspire to be. A lot of business people fit into that category. They want to do their job well and want to know how to measure success. Their passion may lie elsewhere.

If you are a creative, passionate individual it is easy to discount these people as uninspired. But that's not really fair. As Ambrose Little said earlier in this thread:

"As for innovation...businesses can't bank on that, and even most who aspire to that will fail....I don't think it's wise to...toss out process—the point of which is to provide some repeatable consistency, even if imperfect and not particularly sexy."

While I push back on those who say "UCD is Broken, " I also listen to them very carefully. I think that there is merit in their criticism and it does spur me to rethink ideas that I may have held uncritically for some time. What I don't agree with is the idea that you must discard the past to move forward. For me, that's too cheap and easy an approach.

In the US, many of our folk heroes are mavericks who, without formal schooling in the status quo, come up with a breakthrough idea that turns everything on its head. Sometimes that model works and produces exciting results. But what we forget is that most innovations come from people who studied the process and were insightful enough to see beyond it. Beethoven studied Bach. Picasso studied El Greco and Gauguin. Like Newton, they "saw father than other because they stood on the shoulders of giants."

We need both rich process and breakthrough thinking. And understanding the former is one of the best ways to achieve the latter.

Charlie

Charles B. Kreitzberg, Ph.D.
CEO, Cognetics Corporation

On Jun 24, 2008, at 1:57 PM, Robert Hoekman Jr wrote:

I agree 100%, but I also wonder if this is how UCD, GDD, and others came about in the first place

http://en.wikipedia.org/wiki/User-centered_design

It's interesting to note how little there is about design in an entry titled "user centered design." I will also note, that a statement like:

"In broad terms, user-centered design (UCD) is a design philosophy and a process in which the needs, wants, and limitations of the end user of an interface or document are given extensive attention at each stage of the design process."

Could only come from a group of people with little industrial design or graphic design training. Both ID and GD are rooted in things like communication, context and audience. To ignore the audience or the customer is inherently bad design process. In other words... UCD at best is redundant as a concept and at worst, allows an opening to engage in poor design practice by ignoring the balance that is need in creating products between use, technology and commerce.

My experience has been that UCD came into being because too many people without design backgrounds were thrust into an environment in software and interface design where they had to go toe to toe with hard-nosed engineers or business types. Since too many people lacked design training or lacked the ability to work through the means to communicate with those types of people who think in terms of pure business or pure technology, early folks in this industry fell back to what they knew since many of them came from backgrounds that were more rooted in research or process: They invented a "process" that forces the team back to consider the user at every stage to create the balance.

Good design assumes an understanding of audience, context and aims to fully communicate and converse. And even so, good design is ultimately a balancing act to address a large set of concerns, of which the user is but one piece.

-- Andrei Herasimchuk

Principal, Involution Studios
innovating the digital world

e. andrei at involutionstudios.com
c. +1 408 306 6422

What I don't agree with is the idea that you must discard the past to move forward. For me, that's too cheap and easy an approach.

I greatly respect your willingness to pay close attention to and consider opposing arguments, but this particular point bothers me.

"The past" is filled with far more examples of products, innovative thinking, and success stories based on activity-centered research, magic, genius design, and just plain luck than UCD can claim even on its best day.

What's cheap and easy is the idea that we can dissect a chef's work and call it a recipe. That we can simply analyze genius and come out with a one-size-fits-all plan for success.

-r-