Amazon’s already built an implicit web app


In preparation for Defrag, I’ve been trying to organize my thoughts on the implicit web. Previously, I gave an example of how implicit information could improve search. I realized as I was thinking about this last night that there was already a successful implicit application being used by millions of people every day; Amazon’s recommendation system.

It fits well with my idea of the implicit web; it’s using the data passively collected from users behavior to offer up recommendations, there’s no user data-entry required, it’s a behind-the-scenes servant offering up useful information. So if the implicit web’s been in use for years, why do we even need the term?

Amazon is in a very rare position; they have a massive set of trusted data to work with, and they know a lot about their users. This gives them enough information to chew on and produce something useful. Very few other sites have enough breadth (number of users) and depth (information about their user’s behavior) to do anything similar. To do something comparable without owning a site like that, services have to sit on the client side and collect information as the user browses the web.

me.dium is an example of that sort of service. At the moment, they’re offering a very simple service; show me where my friends are surfing, and related sites strangers are visiting that are similar to my current page. The second part is pretty similar to Amazon’s recommendations. The really exciting bit is that once users trust you with the data about where they’re surfing, and what their friends are, you can build some really compelling services. Here’s a few examples of what I’d do with that information:

  • Build an implicit list of favorites based on how often a user visits sites, and how long they spend there. Let them use it as a bookmark toolbar, or even publish it to their friends as their current favorites.
  • Highlight links that other people took most often to leave a page, and show pages they came from most often, giving friends a higher influence.
  • Let the user ‘stumble-upon’ pages that are popular with their friends right now.

I’m really excited about the possiblities, and I’m looking forward to some interesting conversations at Defrag!

Funhouse Photo User Count: 1,251 total, 60 active. Weird, a big jump from yesterday, almost 90 users, but the stats page claims only 22 adds in the last 24 hours. I’m not sure what’s up, but I’m not complaining!

Event Connector User Count: 36 total, 9 active. Not much change in the numbers, but there has been some progress. Tim from New Media Expo has set up a Facebook event for their 2008 show, and has around 100 guests so far. You can try out the connector here, I’m hoping that it will be as popular as it was for Defrag.

Defrag Conference

I’ve just signed up for Defrag, a conference focused on the implicit web. In their own words:

Defrag is the first conference focused solely on the internet-based tools that transform loads of information into layers of knowledge, and accelerate the “aha” moment.

People often talk about information overload, and trying to cut down the amount of data people have to deal with. That approach leads to solutions where a computer tries to do part of the user’s mental processing for them, which is a slippery slope towards talking paperclips.

I want to give people more information, but in a form they can digest. I want to present something that all our wonderful pattern-matching circuitry can sink its teeth into. We’ve had millions of years of adaption to spotting pumas in the undergrowth, we should take advantage of that.

It feels like a lot of the Defrag folks are thinking along similar lines, so I’m hoping to meet some interesting people who are working at the same coal-face, and get advice and inspiration. Plus I’ve never been to Denver, so maybe me and Liz can combine it with a vacation!

What Does Me.dium Do?

I’ve been using me.dium for a few weeks now, and it’s a really interesting service I think more people should try, so I thought I’d try and explain what it’s all about.

The first hurdle I ran into when I installed it was understanding what on earth it was! It adds a sidebar, with a radar screen at the top showing website icons, that mysteriously shift around when you move between pages, and an area at the bottom for IM-style chat. It took several days of use before it really clicked, and it’s still hard to explain, so I’ll just give some concrete examples of what it can do for you:

Uber-StumbleUpon . The radar screen shows you sites that it believes are related to the one you’re currently on, using some kind of algorithmic black magic. This led to some fun website-hopping, SU-style, as I discovered new sites that were similar to others I already visited. What really impressed me though was when it showed a technical site that had exactly the answers I needed, while I was researching a tricky coding question. It saved me days of work, since I’d been searching hard and that site never came up through Google.

IM with Context. As well as showing you sites on the radar, it also has icons representing your friends who are visiting those sites. This adds a new dimension to IM, since you can see which sites your friends are visiting, have an idea if they’re busy or bored, and generally get a bit of the same implicit social information that’s so important when you’re in the same room as someone. It’s as if you can glance over each other’s shoulders and see what you’re up to. This obviously has privacy implications, but I see it being pretty similar to the real-world. There’s times and people you’d be happy seeing what sites you’re visiting, but you want control over that. In the real-world you’d do this by keeping your screen out-of-sight if needed, and controlling who wanders around your house or office. In me.dium you can do the same by turning off sharing, and limiting your friends to those you don’t mind seeing your surfing habits.

Give Every Website a Comments Section. As well as showing related sites on the radar, the sidebar shows conversations that were written by people visiting the same or similar pages. This is really handy for annotating websites, either just shooting the breeze with strangers, or having an in-depth discussion about the content of a site. This is something I’ve wanted for a long time, there’s never been a simple, universal way to see this sort of user content on a site. Even when comments are allowed, the machinery is always fairly clunky and heavyweight, usually requiring moderation to defeat spam, and some mental effort to compose a ‘post’, which excludes the spontaneity of IM style chat. I don’t think me.dium will replace traditional comments, since the conversations are not part of the public net for posterity, but I think this really fills a gap.

Make New Friends. This is another place where me.dium is more like the real world our primate social skills were developed for. If you see there’s someone else on while you’re visiting, you’ll see an icon, with no identifying information unless they’re a friend. You might well strike up a conversation, just like you might if you bump into somebody reaching for the same book in the library. The key to this is that there’s almost no effort involved, you will naturally see people who share your interests, and might get to know them better.

As you can probably tell, I’m pretty excited about the future of the service, especially as more people begin to use it. If you have Firefox, try it for yourself! Don’t worry if you’re on IE, there should be a version for you in a few weeks.