Creating passionate crowds

I've posted before about Digg and its deep underlying philosophical difference to Google. To recap, while Digg uses explicit 'Wisdom of crowds' approaches to generating a front page, Google uses an algorithm (which takes into account implicit votes based on links) to generate its Google News front page.

These are fundamental and philosophical differences and when the issue of Digg gaming came up (and will come up again I'm sure) it was an important discussion to have.

In fact, I recently came across a wonderful article from Kathy Sierra called 'Dumbness of Crowds' where she rightly states that the term 'Wisdom of Crowds' was actually meant in sarcasm. It was supposed to highlight that crowds (read: mobs) are actually quite stupid. Real intelligence comes when measuring individual actions in aggregate (and even then in some applications and not others - e.g. designing by committee produces bland or Frankenstein results) - rather than giving individuals collective and visible control over a process.



As I've stated before, the problem with both Digg and Google is that in an era of hyperchoice and information overload these engines only show 'What's popular' instead of 'What's Personally Relevant',

Today, however, DayLife launched, to a little criticism from one of its investors (who just happens to be Michael Arrington). It has been a long time coming and, as a result, seems to have felt a little 'over anticipation' from some in the community.

Worse still (shock/horror/sarcasm) it does not use the "Wisdom of Crowds" OR a Popularity Algorithm to generate its front page. It in fact proudly and loudly declares that it uses a 3rd, age old technique - human editors.

I'm sure the DayLife founders are very passionate about their company and I wish them best of luck with their plans. Maybe in a noisy media landscape, A site that shows simple, visual and effective headlines on the front page will be a refreshing change?

Personally though I look forward to the 4th 'front page' philosophy/technique. Using a Personal Relevancy engine to generate a front page.

Blog Highlights of 2006

Hi everyone - I hope you had a great 2006. I know we certainly have. In just one year, Touchstone has gone from the back of a napkin to a funded, flying company with a number of great staff, friends, advisors and testers. We wish to thank you all very much for your efforts this year - we literally could not be doing this without you.

Here are the highlights from the blog over the last year in reverse order (most recent at the top).

Hitting the Mainstream 2
Information Overload hits the mainstream media for a second time - in a big way.

Hitting the Mainstream 1
Information Overload hits the mainstream media for the first time (or there abouts)

Democracy Now!
Web 2.0 has barley hit and people are talking about Web 3.0. We discuss how absurd that is and why YouTube is NOT Web 2.0.

Show me the money (or pain!)
Some people (read:head in the sand) think there is no information overload problem. This post explains why there is.

Filtering vs. Ranking
Some people are still talking about filtering RSS. Filtering is so 5 years ago.

Aggregation is King
Content used to be king. If that's true, then Aggregation is now master of the universe.

Desktop vs. Web-based
Web 2.0 implies that stuff is on the web. Not true. This post talks about the value of desktop applications in a web world. By the way - did you know the Browser is a desktop application? Shock/Horror.

What is Attention Data?
And no - it is not just OPML or Attention.XML.

Personalized Content
Some claim that the battle for 'People Powered News' is over. Digg and others have won. I make the argument that People Powered News MIGHT be done, but Personalized Content is just getting started.

There is no more audience
Participants have killed the audience. Media outlets that treat their audience like eyeballs are doomed to fail in a Media 2.0 world. This is a short rant about the death of the Audience.

Touchstone funded by Angel
Touchstone get's funded by an Angel Investor. What more do we need to say about that!

The Long Tail of Attention
Chris Anderson describes the three factors that have made the Long Tail a viable market. I then explain why a Tool like Touchstone empowers the 'Long Tail' (that's you and me) to take advantage.

Personal Relevancy
What is Personal Relevancy exactly? It's when your interests and personality become the basis for choosing content, rather than the whims of one editor who decides what 'the mainstream' should care about.

Tune Out the Noise
Touchstone is not about alerting - it is about NOT alerting. Think about that.

Attention, Scarcity and Demand
Markets work on Supply and Demand. Price is dictated by Scarcity. So in an era of abundance, the scarcest resource is our Attention.

Power Back in your hands
Amazon Recommendations are great... for them. They help cross-sell and up-sell their customers. But what if you could use the same technology to take control of your information across all the sites you visit?

Anti-Web 2.0
Touchstone is a desktop application. Does this make it Anti-Web 2.0?

Not a Gadget Engine
Touchstone compared to the current rash of Gadget/Widget engines out there.

RSS is not just about News
Imagine using RSS for something other than News. Feed readers fail for most of those other applications. Touchstone picks up the slack.


Thanks again for sticking with us. The best is yet to come!

Chris, Ash and the whole Touchstone Crew.

The PageRank of Personal Relevancy

Ilya recently posted some great data and analysis on the current state of RSS reading habits.

He goes on to explain many of the issues I alluded to in a previous post titled "Show me the money (or the pain)".

I think the question of information overload is answered. Yes there is an overload. But RSS is not the problem. In fact blogs and user generated news are not the problem either. They are just one source of information in our lives.

There are application events, presence changes from our friends, internal memos from head office, applications on our desktop and more all clamoring for our time...

So tools that try to cluster and suggest content from blogs and mainstream news sites are only (very) useful for part of the time.

Ilya goes on to make a great suggestion in his post. He recognizes that collaborative filtering has limitations, Keyword filtering is 'so 5 years ago' and that any one 'community voting' measurement will fall short.

With Touchstone we have gone to great lengths to cover all these usage scenarios. We have built a platform that accepts 'items' not 'RSS'. This means that we can source content from places other than RSS and then cache, rank and route them in a unified way.

Our 'rank' is not based on collaborative filtering or keyword filtering or community voting or previous reading behavior. It is based on some and none of these things at the same time. As such, our technology can work in a vacuum on a personal item behind the firewall, just as it can work on a news item that the whole world can see and link (read:vote) to.

Also, there is no 'handshake' period where our application tries to track your reading behavior over time. We are on the client side which means we have access to your browser history/cache, documents and email for an instant, broad and ongoing base of 'Attention Data' in order to determine your interests.

Ilya rightly compares this to PageRank. While PageRank uses incoming links as a vote to measure authority, it relies on a broader set of factors to make a decision and produce a number.

And because the result is a number rather than a binary 'yes or no' filter or an opaque recommendation, Touchstone can make intelligent presentation decisions when displaying the alert/content/information. The bigger the number the bigger the item on the page.

My friend Adam also posted about the 'Feed Overload Problem' on his blog.

Media 2.0 Roadmap

I have been thinking a lot about 'Media 2.0' lately. So I quickly wrote up a roadmap from the distant past media landscape to our near future opportunities. I'd love to hear your thoughts.

Distant Past (Local Radio Stations)

• Distribution: Costly, via radio towers and dedicated ‘wireless’ receivers
• Content: Local news and radio plays
• Advertising: Local sponsors
  

Past (National Radio Networks and TV Networks)

• Distribution: Costly (via radio and TV towers, TVs and Radios)
• Content: National shows targeted at demographic groups
• Production: Costly
• Audience: One way broadcast from the top to the masses
• Content: ‘What’s popular’ (as decided by editors) is on the air – segmented by broad demographics
• Advertising: Local and National sponsors
  

Recent Past (Internet – Web 1.0)

• Distribution: Cheaper (via modems and PCs – unstructured content in HTML)
• Production: Costly (in terms of time and skills)
• Audience: One way broadcast from the top to the masses – now also on the web
• Content: ‘What’s popular’ (as decided by the editors) is on the air – segmented by more niche demographics
• Interaction: Interest groups and communities trapped in silos
• Advertising: Local and National advertisers splitting revenue across web, tv, radio.
  

Now (Internet – Web 2.0)

• Distribution: Mostly Cheap (existing TV, Radio towers and across multiple devices using the Internet – structured content via RSS)
• Production: Cheap (just click publish on your blog)
• Audience: Two way participation within the audience (‘the bottom’) with democratic editorial control in the grassroots
• Content: ‘What’s popular’ (as decided by the audience) as measured using services like Technorati, TechMeme and Digg etc. Segmentation in the mainstream continues by more thinly sliced Demographics)
• Interaction: Interest groups unbound by silos (due to RSS)
• Advertising: Context sensitive Ads targeted at the page – served by Google, Yahoo and Microsoft
  

Coming (Media 2.0)

• Distribution: Cheap (across multiple devices using the internet as the ‘universal pipe’ – structured content via RSS). Aggregation is the main user interface.
• Production: Cheap (just click publish on your camcorder and mobile phone)
• Audience: The audience is gone, only participants are left: Two way participation with all stakeholders and democratic editorial control of what’s on the web and what’s on the air
• Content: ‘What’s popular’ (as decided by the participants and measured by services like Technorati, TechMeme and Digg) is played on air. Segmentation by niche interest groups.
• Relevancy: With hyperchoice, ‘What’s personally relevant’ becomes far more interesting that 'What's popular' – Audiences of one.
• Advertising: Ads targeted at the individual – served by aggregators

Too much noise for 37signals

Matt from 37Signals (I'm a big fan) asks if there is any way to tame the RSS beast. These posts always make me smile. As do the responses/comments.

A lot of people almost get it, but never really provide a complete answer.

To date our answers (and the answers of others in the Attention space) have been fairly academic - over the next few months as we head into Beta - you will start to see some simple, practical solutions. Solutions that might make 37signals (and their cult of the simple) proud.

IAM responsible for this

As you could predict, Chris Anderson is a hero of mine. I almost had a chance to see him speak at Plug & Play while I was in the bay area last month but unfortunately we had a scheduling conflict. Thank heavens for blogging.

His latest post entitled 'I'm not responsible for this' , like many, strikes a chord for me personally and for Touchstone as a product.

He quotes:

The fact that you and I both watched American Idol last night probably doesn't define us, whereas our niche interests really do. We go deep and find people who share our affinities, which represent much tighter connections between us. So my suspicion is that we're going to have fewer loose connections with lots of people but tighter connections with fewer people.

I like this idea very much. It is an idea we have been discussing internally for quite some time. If Touchstone can calculate a highly granular and complete picture of your 'long-tail interests' and store it in APML, and if it can apply that model of your interests to filtering and finding content and people - have we not created both a finger-print of your identity (at least part of it) and a highly personalized world view of content and people like you (and that you like).

The whole post is put in context to Kevin Smith and his continuing popularity despite his recent box-office failures. Anderson writes:

Just like they did 10 years ago, lonely/nerdy/smart teenage boys see in Smith a humor they identify with and a personality they want to emulate. The movies are incidental: Something like Clerks II has the relationship to the Smith brand that a communion wafer has to Catholicism.

I love that :)

Filtering is so 5 years ago

In a recent Read/Write Web blog post Richard (Great guy by the way. Met him at Arrington's place for drinks and poker!) talks about Filtering RSS. He states:

[Filtering is important] Because it's the next step up from RSS aggregation, as many of us now have too much information coming at us.

I agree there is too much information coming at us, but the problem is not just the volume, it's the quality.

There is no variation in the sound. It's like using all the colors of the rainbow at once. The result is white. Or playing all sounds everywhere at the same time. The result is noise.

What we need to do is add variation back into the system. We need a way of creating our own personal signal.

I have a problem with filtering. To me, filtering is like Web 0.5 search. WebCrawler and others were doing keyword matches and returning a barley useful set of results. Filtering is rudimentary and only partially useful. Google understood that a piece was missing. A way to rank documents not only based on the number of keyword hits, but on their popularity in the community and relevancy to the subject matter.

What we need for RSS is ranking. Real ranking. Not just based on the number of keyword hits but on relevancy. Unlike search, however, relevancy is not enough. What we need is Personal Relevancy.

Personal Relevancy is not about asking 'How many keywords hits does this document have to my search query' or even 'how popular is this document based on incoming links' but rather 'Based on everything you know about me, how much do I personally care about this document right now, in what format do I want it presented to me and on what device'.

Can you see the difference? Of course you can.

Unlike filtering, relevancy is not black or white. On or off. Include or exclude. It has variation. It has rank.

Touchstone has been ranking content for 6 months now. We have decided to go one step further. Escalating Alerts. But that is a topic for another post.

Another key theme of the post was the ability to re-syndicate the 'filtered' content to RSS. Touchstone has also been doing that for ages. But again, our RSS feed is not just filtered; it is ranked. We have added an extra tag called 'Rank' so that compliant applications can make intelligent presentation decisions just like Touchstone does with its Escalating Alerts.

Filtering is neat. Ranking is powerful. And over the next few months when we start to release Touchstone into the wild I am sure the community will agree.

If you have interesting ideas about how to apply our personal relevancy technology, drop us a line.