Web 2.0 Talk – Leveraging the Network
February 21st, 2006
Last Thursday I gave a talk about Web 2.0 at the Greater Boston Chapter of the ACM. The Chapter is a non-profit educational and scientific society of computer professionals in the Boston area. About 45 60 people were there and we had a great discussion around such interesting topics as recommendation systems and the privacy issues surrounding web-based services like Google’s Gmail and Amazon’s account information.
Here are my slides (added a few since the talk). Web 2.0 – Leveraging the Network (pdf)
Major talking points
(the slides contain a little history and examples/supporting information for each)
- The home page is no longer the most important page on your site.
- The information architecture that people use to find your content is, increasingly, not yours.
- Each feature added to an application is more to think about – for everyone.
- Folksonomies are a way for users to map their own, familiar vocabulary to your alien one.
- Words are the currency of the Web. Spend the most time on your words.
- Seducible moments are those increasingly rare moments when you can talk to your users in an appropriate context.
- Recommendation systems are a forced move.
- Users want control.
- Users appreciate tools that help them make their own well-informed decisions.
- The best software models human behavior.
- Links model how users value content, and are only the start…
- Sometimes it is easier to design for yourself than others.
- There is always an opportunity for a better interface to data.
- All things being equal, faster interfaces allow for more innovation.
- Most people are willing to trade their personal information for good service.
- As choices grow, so does the importance of learnability.
- Redesigns are dead.
- Network effects are rare, and killer.
- Network effects work in the opposite way for teams building software.
- Personal value precedes network value
- People rarely do things for the “good of the network”
- Del.icio.us, though providing very cool tagging features, is mostly about a single person remembering items for later.
- “The accretion of tiny marvels can numb us to the arrival of the stupendous”
March 9th, 2006 at 6:41 pm
“Users want control.”
“The best software models human behavior.”
IMHO these are both the same thing. Together maybe with “Each feature added …”
The web is every day more about tasks, as opposed to the original design oriented at displaying documents.
This brings an important difference: when somebody publishes a document, or gives a speech, it’s all about his thinking and the audience is willing to listen. Afterwards they can do what they want, it’s not usual that they stand up and walk over the speaker. The “users” are in a pasive attitude.
But when the user is supposed to be active, because he is performing a task, then he wants control, wants the thing to behave properly, read on.
First, control belongs to the user, period. By 1998 when “webmasters” talked about “eyeball counts” and “not leting the user leave the site” they were inmersed in control fantasies that seemed the outcome of massive caffeine overdoses. Users responded leaving the internet (against all pronostics) and provoking a crisis.
But what kind of control? Software developers imagined that users wanted to set their favorite colors and the like: “personalization”. This is not giving control to the user. It’s comparable to a marriage where the husband chooses and sets the home and gives “control” to the wife by letting her choose the color of the curtains. It’s a hoax.
Control means that the user wants things done “his way”. Here is when “… human behaviour” enters.
As you know better than me, he user has built a “conceptual model” for the task he’s going to perform.
On the other hand, the develpoers have published a “manifest model” for the user to operate, mainly thru the UI (it’s appearance but most important, it’s dynamic behaviour).
Well, if the manifest model is not designed in conformance with the user’s conceptual model there is a clash.
See this picture:
http://www.elearningtoys.com/Merchant2/graphics/00000001/sm_plan_shapesort150.jpg
manifest model that does not conform to the appropriate conceptual model is like trying to introduce a shape in the wrong hole.
Because the user is expecting the thing to behave the way he thinks, and the thing behaves in a different way which is not evident.
That’s the relation between user in control and modeling human behaviour: the user will feel he is in control as long as the software behaves the way he is used to think. No matter the font or background colors.