Organizing global websites.
At the European IA Summit in Barcelona, I'll give a talk about information architecture for global websites. The slides without the talk aren't that great, but here they are anyway, maybe it'll help you decide wether to attend this talk or one of the other great ones. I'm also giving a Drupalized version of this talk at Drupalcon the day before, by the way. See you there!
IA for beginners: take your time
OK this is smart: "Jared Spool is interviewed about UIE's Inherent Value Tests. It's a test designed to reveal why new users struggle to see the purpose and value of some product or service when a large body of loyal users is complete devotees.The test is a modified usability test broken
into two pieces. First, loyal customers are asked to give you a tour of
the site and tell you what they find valuable. Second, new users are
given tasks to see if they come up with the same values. By comparing
the two tests, we can see what it is that the new users don't get and
why."
Simple approaches that solve real problems like that, I like it. The yes/no 10-questions game to reveal facets (for faceted classification) is another great simple approach that gives real value. I should make a list of these one day.
Search engine optimized or search engine friendly? I like search engine friendly. Actually, link friendly is arguably even more important.
UX services for startups
Since it turns out, for some strange reason that I can't really fathom, that I've been doing information architecture almost exclusively for startups since I started consulting independently a few years ago, I was wondering: perhaps I should start some specific UX services aimed at startups. We're in the middle of a startup bonanza, after all.
So, dear reader, any ideas for UX services that startups might like?
Scribefire test.
Customer acquisition for startups, and a bunch of fairly random numbers.
I'm often surprised by how customer acquisition is left to the "marketers" (that haven't been hired yet) at startups. When really, it affects how you build the website, which features you add, the information architecture of the site, everything.
I'm also sometimes surprised by the naivete surrounding customer acquisition. Regardless of some success stories that apparently just became popular just "because", most startups have to work hard at this.
If you're trying to get paying users, a customer acquisition cost of 10 US$ / user is normal. 20$ would be normal too.
(Figures vary widely from site to site of course, but I'm adding some that have been fairly typical for sites I've worked on just because this information is so hard to get.)
That means, for example, that if your business model assumes you will only make 10$ per paying user over time, you're not going anywhere.
There is also often little understanding about the amount of customers you can reasonably expect to get, and what that means.
For example, if you have 10,000 signed-up users for a free service, that most likely means squat, nada, zilch, nothing. Your 10,000 first users are rather easy to get. If you have 100,000 users, that is a good start, 100,000 users is a respectable start. A million users, that's doing well, someone might want to buy you.
Again a disclaimer: numbers are just numbers, and your situation will always be different. I'm just putting them out here as a kind of rough guideline.
As another example, if you're getting 100 new signups a day, and it's not trending upwards, that doesn't mean much either. If it *is* trending upwards, then you might be going somewhere. At 1,000 new users a day, *if* those users actually use and enjoy the site, you're doing good, going somewhere. You might get an investor interested with those numbers.
Another example. If you are hoping to get 10,000 paying users for your app, that's certainly doable. If you then think you'll get to a million just as easily, you're mistaken. A million paying users means a *lot* of work doing purely customer acquisition. Years of work. And yes, that means marketing, amongst other things.
A much more useful metric than users are active users. Those are people that have used your site in the past week or month. You can accumulate 100s of 1000s of users, but if they signed up and then forgot about your site, there's not much happening there. If you have even a few 100 very active users, that can be enough to jumpstart your site.
Digg only recently hit the million users point, and most of those users don't post stories. Using Digg is very easy - just visit the homepage and click some links. That simpleness of the action is one of the reasons it's been so successfull I believe. Next level, you can sign up and vote for stuff. Voting is super easy too: there's only 1 button. No thinking required. Then the next level (and a very small % of users does this) is submitting articles.
So Digg has relatively few users for being such a powerhouse. And even when they only had a few 10,000 users they were considered popular.
Anyways, this post is mostly for testing Windows Livewriter, an excellent product from Redmond.
Testing Livewriter's glossary feature: the UI is a bit weird, but it'll help me adding links like this: (via Simon) by pre-populating the links. Cool!
OK I upgraded my Wordpress install and now I can use Livewriter again. Yey!
Facebook is booooring!
Facebook is boring. I just realized that. For a social network that as LOADS of my contacts already on it to be boring, that means something. But every time I visit it, there's nothing interesting to see but apps my friends added and empty discussion forums, nothing to do but add some friends. Booooring!
And I'm not alone.
What happened to podcasting?
Podcasting seems to have dissappeared from the radar of the blogs I read entirely. There's a lot of value there though, so what happened? And since no leader has emerged in the podcasting space, now *could* be a good time to start a killer podcasting company.
Peter Van Dijck’s Guide to Ease » Blog Archive » A brief and personal history of videoblogging
Peter Van Dijck’s Guide to Ease » Blog Archive » A brief and personal history of videoblogging: "Since memory is fickle, and these days, and blogs - not news - are the first draft of history, I wanted to note down my personal history of videoblogging before I forget most of it."
And indeed, I have forgotten most of it. So that was my 2007 draft of videoblogging history, here's my 2005 draft of videoblogging history which really lays out things month by month.
Today, I could write another one but it'd be totally off and consist mostly of personal memories that didn't really happen that way, like those people that, asked about their meeting Bugs Bunny at Disneyworld, recalled vividly the details of that meeting (Bugs Bunny isn't a Disney character). That's memory for you.
Vagabox
Vagabox "Burn This Book" 1 dreamslaughter Selected Quotes:
Work like you don't need the money. Love like you've never been hurt. Dance like nobody's watching. ~ Satchel Paige
Hey PBH is number one for the query "what a boring site".
winterson.com: episode iii, the backstroke of the west
winterson.com: episode iii, the backstroke of the west: what happens when you cross-translate.
Protoscript - Home
Protoscript - Home (via Simon): I don't understand why they say it's "for prototyping"? Is there something wrong with this javascript library that makes it unsuited for real use?
Myngle.com
Some link juice for Myngle, a language learning marketplace startup that I helped with.
apophenia: confused by Facebook
apophenia: confused by Facebook: "But what I don't understand is why so much of the tech crowd who lament Walled Gardens worship Facebook. What am I missing here? Why is the tech crowd so entranced with Facebook?"
It's like TV: it's summer, there's nothing else on. :) That's the simple answer. That and a feeling of relief that we finally have someone attacking Myspace's dominance, because admit it, nobody over 21 likes Myspace, especially not the web 2.0 crowd that was in love with Flickr's superclean design.
I really want the same user interface goodness of Google Analytics (it's the best UI out of all of Google's products) to be applied to Google Adsense. Come on guys!
Projectpage.info: share deliverables on a page with a password. Simplicity itself, because I needed it. Let me know if you want an invite.
Peter Van Dijck's Guide to Ease - video blogs, vlog, or podcast on Mefeedia
Peter Van Dijck's Guide to Ease - video blogs, vlog, or podcast on Mefeedia - Most of my old videoblog stuff is there. Even though I sold it, Mefeedia is still going strong, and the *archive* aspect of it is becoming more important every day.
Some damn innovative navigation.
OK so now THIS is some innovative navigation :) Talking heads for each major section that tell you about it. Really, it's innovative. You really have to visit the site to get the full effect, with audio turned on.
What the hell is a chupacabra? (aka the elmendorf beast)
ABC news, in the town of Cuero, Texas, USA (only in the USA!): "Phylis Canion lived in Africa for four years. She's been a hunter all her life and has the mounted heads of a zebra and other exotic animals in her house to prove it. But the roadkill she found last month outside her ranch was a new one even for her, worth putting in a freezer hidden from curious onlookers: Canion believes she may have the head of the mythical, bloodsucking chupacabra."
That's one nasty animal - nobody is going to do that any favors. And the name isn't helping either, and calling it "elmendorf beast" still doesn't help - what kind of name is that?
Wikipedia says "Eyewitness sightings have been claimed as early as 1990 in Puerto Rico, and have since been reported as far north as Maine, and as far south as Chile."
1990 - that's not early, that's just a few years ago.
This was supposed to be an urban legend - can an animal like this really exist without us knowing about it until now? I'm not buying it. I'd like to buy it, I'd like to think the world is that cool, but I just don't think so. Then again, before you know it, we'll have chupacabras all over the place. But then again again, it's Texas! They'll buy anything.
And for your continued enjoyment of the chupacabra: some more pictures.
Finally, here's a video that doesn't help much either. But hey:
[youtube http://www.youtube.com/watch?v=SmHmnfgON4A]
ps: "chupa" is Spanish for "suck", "cabra" is "goat".
Mosoci is a (new?) consultancy with Dina Mehta that provides 2.0 strategy based on deep ethnographic-style research (at least, that's what I understand.)
DonnaM: Free pass to Oz-IA 2007
DonnaM is giving away a free pass to the Ozzie IA conference.
Drupalcon and IA Summit both in Barcelona
Drupalcon is at the same time in Barcelona as the Euro IA Summit, and I'll be giving a talk there too :)
If you're going to the IA Summit, you could come a day early and we could have a beer at Drupalcon.
Long pages work continued
My long pages work post (a quickie really) got picked up: Dion Almaer over at Ajaxian says: "The iPhone is also showing that scrolling is a nice UI tool". Sean Kane provides some "it depends" input: "A long page with important messaging below the fold in this case could hurt sign-up rates." I have to disagree with this example though: long pages can be *great* for converting users to sign up for something.
Oh my god the Belgacom site sucks. I've been trying for hours to get online billing to work. Conclusion: it is broken, not to the point that you could persevere for a few hours, or get lucky, and actually get it to work, but it's broken to the point that, after you jump dozens of usability hurdles that'd stop an angry African elephant in its tracks and make it go back home shaking its head, that is, if you *do* persevere and use a few tricks only geeks like me would know about, and you do finally get access, you get a "this feature is temporarily not available" message. And all that after they send me a letter saying they set me up for online billing and please pay like that in the future.
They must really not want my money.
A new movie by Brian De Palma: "A new film about the real-life rape and killing of a 14-year-old Iraqi girl by U.S. soldiers who also murdered her family stunned the Venice festival, with shocking images that left some viewers in tears. Abeer Qasim Hamza al-Janabi was gang raped, killed and burnt by American soldiers in Mahmudiya, south of Baghdad, in March 2006."
shtikl.com » You don’t need a plan, you need skills and a problem.
shtikl.com » You don’t need a plan, you need skills and a problem."
1. Don’t start out with big plans.
2. Work on skills.
3. Apply your skills to a problem.
"Most plans are rubbish. Business plans are fake. Life plans don’t work. The big plan is unknown to mankind anyway. "
http://www.vlogeurope.com/ is this weekend!
I moved the XFML specification here, because I lost the domain to squatters.
IA for beginners: long pages work.
Long pages work. Just look at Wikipedia. In the 90s, we had a fashion where websites would cut up an article into 6 pages, 2 paragraphs per page, to get more page impressions. Luckily these days advertising runs on Pay Per Click more than Pay Per View (thanks Google!), so that practice is going away. But still, people seem to have a tendency to cut up good content, and often, they shouldn't.
Look at this page for example. Over 100 comments, and what's wrong with showing them all on one page? Sure, the page text is 116 KB, but that's not too bad. And it's compressed, so it's really just 21 KB, which is fine and fast even for a dialup.
What it does is that it lets people scan over the conversation, scrolling, and get to the end and contribute their comment. If I were to cut those long comment strings in pieces, the experience would be worse.
Scrolling works. Long pages work. Anything else tends to be informationarchitecturitis (which I define as "the practice of adding too much structure when it's not needed and even gets in the way").
Wiki travel guides
My poorbuthappy travel project is going well. The idea is to have a social site where we create wiki backpacker travel guides. It's in progress (I consider it alpha), but it's going well. The best proof I got is when I got sent a picture of this 71 year old backpacker traveling alone in Colombia, using a printed version of the Colombia travel guide. This makes my day.
Craigslist's mistake
Craigslist rocks (and for reasons similar to why eBay rocks), but their internationalization approach has a serious flaw, which I think is responsible for their limited success in many international markets: they don't localize their taxonomies. I'm not gonna write a long post, but here's a quick analysis I did of their Dubai site:
As you can see, many categories just sit there empty (the red dots), and only a few categories are active (the green ones). For the original Sanfran Craigslist, the screenshot would be full of green dots.
The result of this approach is that the site feels empty and inactive. The solution would be to remove most categories and build local specific categories. For example, "woman seeking woman" isn't particularly appropriate in Dubai society. Construction is booming on the other hand (foreigners cannot own property in Dubai itself, but clever businessmen as they are they have started building multiple artificial islands in the sea where property *can* be owned by foreigners), so one of the few active categories is "real estate for sale", and there could easily be more real estate categories (like "offshore real estate"), specific to Dubai.
IA for beginners: ordering stuff.
Digg is a smart ordering algorithm to show you interesting new stuff. Facebook's newsfeed is a fairly smart ordering system to show you interesting new stuff. Google is a very smart ordering algorithm to show you interesting stuff related to your keywords.
Conclusion: ordering stuff is really powerful.
Checking your analytics daily when working on a website is like weighing yourself daily when dieting: don't do it. It'll just make you obsess about the wrong things.
It seems Livejournal is still going ok (in red) and going up a bit, while xanga is on the way down. Both are somewhat similar community sites.
Mmm the new Bloglines beta isn't bad. Ajaxy and stuff. But ugly though.
What I wish they'd done instead of the ajaxy stuff was to make it more social and make it easier to find new blogs.
Kawasaki: "As a venture capitalist, I have to listen to hundreds of entrepreneurs
pitch their companies. Most of these pitches are crap: sixty slides
about a “patent pending,” “first mover advantage,” “all we have to do
is get 1% of the people in China to buy our product” startup. These
pitches are so lousy that I’m losing my hearing, there’s a constant
ringing in my ear, and every once in while the world starts spinning."
This week is the last of my self-imposed new-father-holiday, so from next Monday I'm back in business. Let me know if you need some information architecture-itis done.
Which brings me to this question: why hasn't anybody tackled the problem of managing mailing list subscriptions easily? It's not that hard, I think it's just a lack of imagination.
- What does a true, next-generation Internet word processor look like, one that deeply embeds the browser and community? -> very good point.
If you're not subscribed to Simon Willison's blog, btw, now's a good time.