Craig's Lemonade

PCForum06 starts with Attention

Esther just kicked off the morning at PCForum06 with "attention" front and center although she managed to avoid using the "A" word with her introduction. Ran into Dan Farber on the way in, whom I've known from nearly 20 years in the software business, but didn't have time to chat as Esther was about to come in. Said hi and met Ed Batistsa of AttentionTrust.org. Incidentally, I just met and am sitting next to Rod Canion (Founder of Compaq) and a few seats down is Scott Cook (founder of Intuit.) Needless to say pcforum attracts the big dogs.

Seth Goldstein's pitch for ROOT Markets is making more and more sense to me. Not sure how you turn it off once you have bought a car and house but I am just beginning to understand. I met Seth at his office but we haven't really explored doing anything yet. Seems like we should be. Attensa's technology could most certainly apply to the services they are attempting to provide, perhaps we should expose to these guys the significant scale and transient meta data volume challenges we have solved. Gathering rudimentary attention data and reprioritizing information on the desktop is easy. Processing the behavior of millions of users is difficult. Seth, I think we can help.

Curiously, Seth said in his intro "Dave Winer made me promise to mention Steve Gillmor" as a pioneer around attention. So, he did. I am looking forward to GestureBank, which is Steve's new product due to be demo'd this week in the valley at SD Forum. A few have mentioned they think GB is competitive with our planned RSS centric offering based on our attentionstream technology. hmmm. I doubt it but we'll see. I've never been in competition with a journalist before. Maybe that's why Steve rarely mentions us. Well, maybe its because we haven't shipped anything yet that demonstrates our AttentionStream tech other than real-time syncing.

TrustedID (esther is an investor apparently) sounds really interesting too - 3 million people victims of identity theft last year btw and is a growing epidemic. Opinity interesting too. Personally, I am less interested in the identity component of this discussion but it is interesting.

Seth wrapped up about wanted to "suck in" his Amazon purchase history. He also mentioned he spoke to the Amazon personalization and social networking folks who not only were OK with it but thought it was a good idea. ROOT Markets created a bookmarklet to suck this info in to their vault. He suggested there should be an API to get at this info. I think this is a good idea. Looks like yet another source of interesting, never-ending stream of meta data. Dove-tailed with behavioral attention data the possibilities are seemingly endless. And it will never stop.

Tags: pcforum06, attention, ROOT Markets, identity, attensa, attentiontrust, gesturebank

Attention MetaData circa 2000

Last week Alex Barnett posted a podcast about Attention with guests Steve Gillmor (well known journalist and Chairman of AttentionTrust.org) and Joshua Porter (see his excellent web site Bokardo.) and his comments on the podcast interview. What caught my attention specifically in the podcast was also summarized by Joshua with:

"Steve’s vision is deep. He sees the whole art gallery, not just the Van Goghs. However, when Alex asks about his ideas concerning Gesture Bank, he sidesteps, fearing that he’ll give away too much. This, of course, was the reason why we asked him, of all people, to come talk to us. He doesn’t give away quite enough."

Steve is no doubt a smart guy. I've spoken with him several times, and I respect him as a journalist even though he seems to consistently contradict everyone else's viewpoint just for the hell of it (you can hear what I mean in the podcast.) At Attensa, we have committed to supporting the founding principles of Steve's AttentionTrust.org - well, we kind of are, we don't really think the rank and file knowledge worker will ever care about "managing" their attention proactively. We think people will want their "attention data" private if they were to actually think about it (they won't), and, they will just want stuff to work better for them with no effort on their part. But I'll defer debating end user's ability to care about any of this for another time.

What perked my ears was a comment on the podcast, to paraphrase, that he "didn't want to share his IP for free with startups so they could get another round of financing." Or words to that effect. I couldn't stop laughing. I wondered, is he talking about Attensa? Some other tech company? I immediately forwarded the message/podcast to Eric Hayes, Attensa co-founder and VP of R&D. Eric and I spoke with Steve in NYC at the offices of Root Markets after the BlogOn conference, and it appeared that he agreed about the significance of the opportunity with regard to attention and its specific application in the RSS area. We have always known Steve as a journalist, thought leader, and more recently Chairman of a non-profit - not a potential competitor. We'll see I guess.

Last summer at Gnomedex Technorati's Tantek Celik asked Eric Hayes to join as a co-author on the attention.xml spec. Why? Simple, after a 45 minute discussion between these two it became obvious, our technical team (led by Eric) has been working on this problem for a long time and knows, hands-on, a lot about it! At Syndicate in SF a few months ago our guys and others at Technorati teamed up and announced the move the spec over to the microformats site.

We have some 20 man years into our infrastructure that began in 2000, yes 2000, although then we referred to our work as "Transient Meta Data" so I do admit we adopted the "Attention" moniker as more clarity evolved in this area. Steve G. may have started kicking around the idea with David Sifry at Technorati three years ago but our infrastructure was processing some one million attention metadata micro transactions per minute on our development servers in 2000. That said much has changed in 6 years and there are a lot more to our AttentionStreams than just RSS.

There is room for many players in this space ranging from those focused on solving information overload through RSS to those dedicated to improving the overall quality of  internet advertising. At Attensa, we are focusing on the former - for Business. And we're starting with RSS.

P.S. Eric tells me he has an e-mail from Alex Barnett to do a podcast interview on this very subject, so look for that soon I hope! Most of the attention discussions to date have been around ideas, not experienced implementation, and Eric has been tasked with stepping up and participating in these discussions.

Tags: Attention, AttentionTrust.org, Attention.xml, RSS, Attensa, microformats, Gnomedex

Attention is NOT Click Streams

I was recently lead to a post by Noah Brier about his spending "six hours talking to geeks about attention" with the AttentionTrust and /Root folks. He has some very astute observations. In particular was his first thought, which happens to be spot on at Attensa...

"On a personal level attention data seems to have the most potential as a way to power some kind of recommendation engine. If you could plug in all my RSS feeds, attention data, del.icio.us bookmarks, etc. into some kind of system and then compare it all to the same data from friends, I could probably get some good recommendations. Clearly this would be a hot seller for the many of us who are bogged down by 200+ RSS feeds."

I spent some time at the beautiful offices of /Root Markets visiting with CEO Seth Goldstein and some of his folks after the close of BlogOn 2005 in NYC a few months back. We are still trying to figure out how our "flavor" of attention technology blends with their vision of the opportunity. Attensa's is not lead generation, that is for sure.

Also, ATTENTION is NOT just browser clickstreams. In fact, unless you are focusing on improved advertising targeting and/or building profiles for sophisticated lead generation - it is probably the least interesting means of gathering and utilizing attention data. Sure, using clickstreams to improve ad targeting is very useful and valuable, and I wouldn't bet against Seth and his team building a successful business, but there are far more interesting applications that are user centric rather than publisher/advertiser centric. Noah identified a major opportunity (above pull quote) and it is centered on improving the user experience.

Noah goes on further and writes...

"Yeah, RSS puts all this information at your fingertips and creates an attention problem. But that's only for us geeks who are subscribed to 300+ feeds. I mean, yeah new technologies will force people to split their time more and more, but will they notice/care? I think it's really important to remember that the average person has no desire to sit around and read all these RSS feeds then blog about them. In fact, if you showed someone how I spend my attention online, they'd probably think I was an idiot who was wasting time."

RSS will be everywhere and in everything ranging from simple blog updates to sophisticated enterprise application information dissemination. Information overload is already here and growing exponentially because of RSS. We believe sophisticated utilization of "attention data" (we call them attentionstreams) can help mitigate the pain.

The net, net; Browser based clickstreams are not "Attention" but only a small piece of the puzzle.

Attention Podcast informative, and here's Attensa's $.02

I just listened to Alex Barnett’s podcast focused on Attention with Nick Bradbury (FeedDemon developer and AttentionTrust board member) and Kevin Burton (Tailrank) and Rojo co-founder) and it can be downloaded here. These are two reallly smart guys but we'd like to chime on this disucssion because we eat, sleep, and breathe attention.

A lot of thoughts came to mind listening to the podcast but I will leave the heavy lifting to one of my co-founders, Eric Hayes, a co-author of the attention.xml spec and one of several smart guys that have been working on an attention based infrastructure for years. Ideally we’d like to get online with Alex and explain what we have essentially already implemented that goes far beyond the discussion points in the podcast. While it has only been 6 months or so since we adopted the “attention” moniker to describe our technology, it is the foundation of what we have been doing for years.

The biggest challenge (and we agree attention.xml can’t handle it) is processing a massive amount of information, in near real-time, and gleaning not only useful metadata (from what we call the attention stream) but to do something really valuable with that information. We believe that is an improved experience - not better ad targeting (Google, MSN and Yahoo can focus on that I suppose)

Attensa’s business model is around just that. Your behavior will automatically improve your experience. In a collaborative work environment? Even better!

Alex, Nick, and Kevin got on the subject of standardization. As we have committed at Attensa, we will support attention.xml and have been instrumental in offering our experience in this area in reshaping and improving the spec based on our real world experience. However, we agree it is not likely to work on a massive scale with a seemingly infinite quantity of transient metadata being generated. Attensa’s strategy is predicated upon addressing the challenge of how to effectively manage very large amounts of seeminly unmanageable attention data.

So, where does that put Attensa in the debate of “does the user own their attention?” Well, our company has already publicly committed as a member of the AttentionTrust.org to its four founding principles. Where we diverge from many is that we feel the manner in which we create and utilize attention data and deliver useful benefits to individuals and businesses – that is not something we share. That is Attensa’s intellectual property. More on this as we roll out our attention based services.

Ultimately, we envision any aggregator can potentially benefit from Attensa processing attention streams (replete with strict anonymity and privacy adherence) to improve the user experience and help mitigate information overload.