Lou Rosenfeld: “Long ago I was probably one of the world’s greatest information architects. For, perhaps, a year or two. (It might have had something to do with the fact that there were only about two dozen of us who claimed the title at the time.) Then I got bored and more importantly, the homesteaders were better at it than I was. So I moved on. And that’s fine; the issue wasn’t one of competence and intelligence, but of personality type and attention span.”
This is what I like to see happen after a redesign of a mid-size informational site.
Visits:
Pageviews:
Traffic coming from search engines:
Guess at which point we launched?
The next thing I want to see happen is for those lines to start slowly but consistently going up (which I think they will). User satisfaction (as measured by feedback) is also way up.
(The little bumps are weeks.)
Webword was one of the first blogs out there, and it was all about usability, so I loved it. John Rhodes was the man behind it, and today he emailed me with a new book he wrote, which I’m reviewing here.
The book is about selling user experience, not from the top down (ie. convincing your CEO), but from the bottom up, which is how 99% of us have to sell it. It’s funny, it’s brilliant, if I could write like that I’d be writing my next book today. I love it.
If you’re doing UX work in a large organization, you should buy this book. And if you’re a UX consultant, you should too. It’s that simple. The book is worth it’s weight in gold: it gives you (as a UX person) insight in how to really get things done in large companies.
The first chapter starts off good (and I’m gonna put a lot of quotes in this review to give you an idea of the writing style and wisdom in the book):
"99% of the people in an organization are not thinking about UX and the other 1 % are thinking about women, fire and dangerous things. Most managers understand UX about as well as they understand the average airspeed velocity of an unladen swallow."
A wakeup call, but true. It’s a practical book, can’t emphasize this enough:
“This book is full of stealth. We’ve got guerilla attacks, end runs, and cloaking devices. These tactics are not conventional. I’m asking you to reject the frontal assault. We’ll be successful under the radar.”
In "2. The First Business of Business is Business", he explains what business is all about.
“How Do You Talk About UX? The advice I am going to give you next is worth the price of the book: Do not talk about user experience for at least a month. Instead, before you say or do anything regarding UX, think about what it means to the bottom line. Modify your language to be more in line with the true intentions of the business.”
Chapter 3: User Experience is an Ugly Baby
I didn’t know Donald Norman used the term “user experience” in 1998 and 1999.
Again, John puts his finger right on the problem:
“Most folks involved in UX do not have business or management experience. This means that few people can bridge the gap between the two worlds. There isn’t a common language available. This leaves UX at a disadvantage.”
In chapter 4: Understanding Your Role in the (Dis)Organization, he explains how companies *really* work. Forget about the org chart.
"Managers hate risk; they love people who can reduce it. In business, there’s nothing so valuable as a sure thing. Put that idea in your pocket and never let it go."
In the following chapters, John explains how to deal with managers, co-workers, designers, sales, CEOs and executives, teams, stakeholders and consultants. One chapter each. This part of the book is pure gold: for every group, John clearly explains how they think (and this is true in almost all organizations), and even more importantly, how to influence them).
More good quotes:
"A consultant has power nearly equal that of a customer. There isn’t quite as much juice flowing, but it can be pretty damn close, especially since your organization is probably paying this person hefty sums of cabbage."
"I like almost all designers and developers. The reason is pretty simple. Unlike so many workers, these men and women get real work done. "
"Sales people talk. They talk to a lot of people and they talk all the time, mostly to product managers, marketing, and of course customers. Although unusually biased, these workers have an exceptional grasp of what your company has to offer and what your customers want and need."
By the time we get to chapter 14, it’s back to you. How to use project momentum to your advantage. Here’s the first sentence of this chapter: "All projects are headed in some direction. You want to understand the vector of activity and inject UX along the way." Damn good stuff.
Now go buy this book.
I’ve only read half of the book this far, but I am wildly enthusiastic, so I’m going to go ahead and post this review right now. Buy this book. Order it for everyone in your consulting company. Really. It’s almost at the level of "Don’t make me think", which I think is the best book about usability ever written. And I only say "almost" coz it lacks the funky illustrations. Go order it! If you’re disappointed you can email me personally.
Cruxy was a smart, future-looking service with a great team and great technology, but it never took off hence it’s now shutting down. “The world has changed for the better, and we are glad for that, but at some point we have to admit, Cruxy is not needed or used by enough people for us to keep going. While we have had an amazing cloud-based business model since day #1 that actual made sense and worked, thanks to my brilliant, co-founder Jon Oakes, we were never able to scale our business up with enough volume to allow us to make an actual living.”
The Nielsman does it again (after years of nothingness): really big menu dropdowns work well (as opposed to regular “menu” dropdowns). I believe it.
I’m working on a large project in a global team, and yesterday I tried something new: making a screencast of my (clickable) wireframes. I did a search today and remember reading about it first here.
So far it’s working great. Team members like this as a way to communicate, the screencasts are fast and easy to make, and they also help me in getting to grips with my own work. I’ll post more details on the how-to later.
I like to watch Dries talking. He talks a bit about how the raised money for Acquia, and other good stuff.
Now Iyengar has published a new study showing that one way to combat the effects of excessive choice is to group items into categories. It turns out that even useless categories make people happier with their choices.
In other words: even “useless” categories can be better than no categories. The study is titled: “The Mere Categorization Effect: How the Presence of Categories Increases Choosers’ Perceptions of Assortment Variety and Outcome Satisfaction” (PDF)
The IA Summit is next week, get your tickets now if you haven’t yet. I won’t be able to make it unfortunately (I should be there next year). Have fun everyone.
Facebook seems even more aggressive about being international than Google was, check this out:
Rock and *fucking* roll, as they say.
Hahaha Pluto is a planet in Illinois.
I consulted with Myngle from initial idea to IA and into development, and now they got a million bucks in funding. Yey for European startups!
The public Pingdom stats of poorbuthappy.com http://www.pingdom.com/reports/ogy7t9gjlgs0/ show how since the move in January from Mediatemple (gs 20$ plus an extra MySQL container 150$) to Amazon EC, hosting is not only cheaper (the EC2 server plus S3 storage is about 100$/m – ok, the stats don’t show that but I’m telling you) but more importantly, the site is about 50% faster (response times have gone from about 1500 ms to about 1000ms). A 50% increase in response times is a big deal, and we haven’t even done much optimizing on that EC2 server.
We’ve open sourced the Poorbuthappy one-click deploy code. It’s still rough, but it’s a start. I’ll try to work on the documentation soon, which is probably the most important bit. http://code.google.com/p/one-click-deploy/
One-click deploy is one of the greatest things I’ve ever added (Chris added it really): it’s like versioning: once you start, you’ll never go back.
“The platform falls through the cloud and is smashed to earth like a plummeting stock price!”
Great playing with metaphors, reminds me of Peter Morville’s playing with the metaphors of tags as leaves and taxonomies as trees. (“But we all know what grows from piles of dead leaves right? Beautiful new trees!”)
I am considering buying a netbook, and I wonder if it’ll be good enough for work. Will report back.
I have been working with Visio for many years now for making mockups and wireframes, but a few weeks ago I bought Balsamiq mockups. It’s a tool that lets you make simple mockups that look simple. I am in the very early stages of a large project and I like the fact that I can easily make something that looks unfinished enough to be discussed as a possibility, not as a finished product (which happens often with more polished wireframes). I was making actual sketches on paper and scanning them before, but they are often hard to read and understand – doing it in Balsamiq is faster.
My conclusion: Balsamiq mockups is fun to use and makes the team focus on what matters by being sketchy, but it needs to evolve just a little more.
Features.
Balsamiq lets you drag & drop lots of UI elements on a page, and combine them into mockups. So far so good. You can save as PNG and paste that into Powerpoint for presentations.
Grouping. You can group elements together, and then move them as 1 element. This works as expected.
Backwards/Forwards. You can move elements to the back or front, this works as expected.
Text. You can type in text, for which you have to doubleclick the element. I would prefer to just be able to select an element and start typing, but that’s a minor problem.
Locking. You can lock elements in place, this works fine. However, you can not unlock 1 element (if you want to make a little change to it), there’s only an option to Unlock all (see screenshot below). That’s a little annoying, if I want to make a change to something that’s locked, I have to unlock all and the relock everything again.
No multiple screens.
One drawback: each screen in Balsamiq is 1 file. So no multiple screens per file. That makes it much harder to keep a project with many screens organized, or to share it. I had about 20 screens in this document and things were already become a little messy. It’s doable, but not great. The tool can have many screens open at once (switch with tabs at the bottom), so it’s workable, but not perfect. My projects usually have between 20 and 100 screens, so let me manage that.
No backgrounds.
Perhaps related to the fact that there are no multiple screens, there are no backgrounds. This is a big problem for me: reusable backgrounds (containing the basic website elements) are a HUGE efficiency win for me in Visio: if I want to change the logo, I just change it once in the background.
No concept of Masters.
Also related to not having multiple screens, there are no masters. To be honest, Visio doesn’t have a decent concept of masters either, but Axure has, and it rocks. Masters are complex (grouped) elements that you reuse on many pages. If you make a change to the master, the element is changed on all pages. For large projects and for efficiency, you really need this. For example, a master can be a search box with a submit button and some text. Change the text in one place and it’s adjusted everywhere this master is used.
No easy creation of new UI elements.
If you want to create a UI item in Visio, you just use the drawing tools and make it, then group it and save it as a new element, done. The only way to create a new element in Balsamiq is to import it as a picture (I had to check the manuals). I made a few sketches on paper of some elements, scanned them, then imported them, but they didn’t look good and it was too much work.
Conclusion.
For the first weeks of the project, Balsamiq was great. It kept people focused on the fact that these are still just sketches, and was fairly easy to use. For the rest of the project, I’m gonna switch to Axure though, which has the necessary features to efficiently manage dozens or hundreds of wireframe pages.
On the positive side, Balsamiq is working on a browser based version with collaboration and commenting, which will totally make it rock for quickly iterating through different designs. Add some power features like multiple pages, backgrounds and masters (you could even leave out backgrounds if you have masters), and you’re done! And even for these few weeks, it was worth the money.
In our series of funky categories that start with a C, here’s the South Park main navigation:
All pretty straightforward except for “Crap” :)
Hahahahaha.
OK, so if you want to visit the USA, you need to visit this webpage to get a visa waiver https://esta.cbp.dhs.gov/esta/, which doesn’t work in Google Chrome (it does work in IE).
If you require travelers to your country to visit an online application before you leave, you better make sure it’s fucking up, US customs!
We just added 1-click deploy from Subversion to poorbuthappy.com. It’s the first time I have that available, so no fudging with FTP-ing changed files or command-line crap. And it ROCKS!
The first time I read about the importance of 1-click deploy was at Joel’s website, (he calls it 1-click build) and then later in the Flickr Scaling Websites book. Now I finally have it. It rocks. Maybe we’ll open-source that. Here’s a screenshot of the admin screen:
Of Joel’s rules for good software, we now do many:
Amazon really kicked Google’s butt with its web services the past years. Today, Google launched a new “status” page of their services, and I kept thinking: ”I’ve seen that before”. On Amazon of course!
You gotta love scaling stories.
Of course they talk a lot about all the mongrel scaling crapp that Rails needs that we don’t have to worry about with PHP :)
If you use OmniGraffle, check out Michael Angeles’ Sketch Stencils, they rock. They’re only 10 US$, that’s like 3 coffees, so go get them while they’re hot.
How cool are these stencils?
These guys make very cool short videos explaining products.
Supplo is a marketplace for goods that I did some consulting with (interesting multilingual challenges!)
I started a new blog about raising kids in more than 1 language.
Paul Minty emailed me to let me know that a non-profit association has been formed in Australia to support the work of Information Architect’s in Australia. Check it out at www.iaia.org.au. You aussies!
OK, bite dust Peter, for getting it wrong.
This is from an email exchange with Steve Chen (Youtube founder who is now filthy rich) in June 2005. He pinged me about Youtube (I was running Mefeedia at the time) and I criticized their service for not being open about their business model.
Me: “Thanks for taking criticism so well :) You’re right, there are lots of new players. Most won’t last though. I’m looking forward to see how you guys do and what features you come up with. Let’s stay in touch.”
What’s more though, in the following emails we discussed syndication and I helped them figure out (together with Josh who was working on ANT then) how to support the mediaRSS extensions in their RSS feeds and specifically add a file extension (.swf) which made my work easier and to add RSS2.0 enclosures.
Me: “Both ANT and Mefeedia (and probably other future aggregators) support RSS2.0 enclosures, not the media:player element. So could you add this to the feed (with the mime type and whatever the length is)?”
Steve: (few emails later) “Yes! Finally! We went through a lot of iterations with getting this right. Now it’s moving on to tackling REST and MetaWeblog API. :)”
So a bit of history there, I think I got Steve Chen to add enclosures to the Youtube RSS feeds :) I had totally forgotten that until today Facebook asked if I wanted to be Steve Chen’s friend. Actually, I remember meeting him in a bar in NYC and disussing the option of going to London to work with Youtube, but that may be a false memory (may have been someone else).
As information architects, we always have loads of ways available to view information in lists. Show the latest, the most popular, the one with the latest comment, everything of one particular type, etc… We don’t always know which way users want to look at the information, so the solution is usually to offer different views of the same information. Choose the most likely popular default view, and users can choose other views as they like.
The problem with that approach is that we’re expecting users to switch between views, and to understand what the various views mean. And that’s often too much to ask - in practice, many users stick with the default view, and whatever that shows or, importantly, hides.
So there’s a new tendency I’ve noticed the past few years to mix multiple views into 1 definitive view, a mixed list. The algorithm that determines what goes in there can be more complex, it can show multiple types of information, etc.
Google’s evolution is a good example. At first, they just showed different options to see their search results: see results in news, in images, on the web, etc…
The "tabs" were there for a few years, but are gone now, as they move to 1 definitive serp, they call it universal search.
The definitive search result page shows thumbnails of pictures, movies, blog results and other elements, all mixed on 1 page. The seem to be experimenting a lot with this and they think it’s successful. Here’s a good example, showing images, videos, shopping etc… on 1 page.
Examples are everywhere, study Facebook for a good example. They also mix content types and stuff like crazy to make their mixed lists more useful:
More thoughts/examples on mixed lists?
I love funky categories, I can always imagine the meetings around those when I see them. Here’s an example:
Second category is "Crave" - it’s the only one that jumps out a little.
The name of the game is influence. “Al Jazeera Network today announced the world’s first repository of
broadcast-quality video footage released under the Creative Commons 3.0
Attribution license.”
I wish I could find a link to a specific Amazon AWS issue from their status page. PBH was down a few hours today, probably caused by a brief issue they reported with EC2.
Todd Warfel is sharing some templates for data-driven personas.
Tripadvisor has a good way of dealing with reviews that are written in multiple languages: the user can choose which language to display first, reviews in other languages get listed below those.
“Let’s have a big time” is the slogan of a new Belgian radio station. Sigh.
I like Google Chrome’s tabs in Options: "Basic", "Minor tweaks" and "Under the hood :)
Amazon AWS continues to innovate, requester pays is another great step forward. Incredible.