Open Data 2007: Deep Thoughts

Today I am feeling somewhat the way I felt after the Web 2.0 Conference: overstimulated, excited, with head spinning. Seth, Devin, Gerry and Tom did an amazing job hosting what in my opinion was a seminal event, one which brought together thinkers and practitioners in the fields of media, advertising and information services. My most interesting take-away is not how different we are but how similar we are. The issues we are all dealing with are fundamentally related, and while I've written about the concept of convergence in the hedge fund/private equity/venture capital worlds, I find myself thinking those same thoughts right now. Reuters? Tacoda? Monitor110? Lijit? Morgan Stanley? Adaptive Blue? We are all thinking about and dealing with the same issues. The same tools, techniques and approaches used by consumer-oriented/social media companies are now being employed by financial services and information technology companies:

• User tags
• User-generated media
• Collaboration
• Influence measurement and tracking
• Creating rich context
• Using metadata to discover both what you're looking for and what you aren't looking for

This is by no means an exhaustive list, just one that I feel represents the convergence megatrend that will, IMHO, be a defining trend in the world of attention and standards-setting for the forseeable future. Given that I have over 30 pages of notes in my little black book, I could write about the conference for hours. I'll spare you; you'll thank me later. But I'd rather pick out some key themes/comments that really made me think, think about the implications for the future of all data-driven businesses as well as the user experience so critical to getting the most out of the massive sea of data we using - and creating - every day.

Devin Wenig - COO, Reuters

• How do we address the vertical markets we serve? How do consumer/social media tools apply to the professional vertical?
• What is the future of financial networks? Creating a platform for structured collaboration, with each individual being a portal.
• Search represents a big challenge. Is the search experience adverse because of technology or the data itself? (wow, is that a great question)
• How does metadata help us find both what were are looking for and what we didn't know we were looking for?

Seth Goldstein - AttentionTrust

• Attention is scarce, so where does it go?
• In the AOL management shake-up, Jonathan Miller, the CEO, received less buzz than Jason Calacanis after he left, yet was a much more senior person in the organization. Why? Jason's influence and his perceived authority. (hadn't thought about, but Seth is clearly right)
• One person's open data is another person's private data.
• Who has the right to use this data?

Abdur Chowdhury - Summize.com
(Former Chief Architect, AOL Search, who was involved in the release of the AOL search query database that caused such a stir. Abdur is a smart, cerebral guy, and I am so glad he made the time to speak with us.)

Three key questions:

• Why open up data? If you give up control, you hope/expect to make things better. If not, don't do it.
• What are you going to do once you open up the data? If 50 people ping you in a week to discuss the data, but you can speak to maybe 2-3 per week, how do you handle this?
• What are you going to do if something bad happens? Because companies fundamentally can't handle when stuff like this happens.

Chris Law, Aggregate Knowledge

• Social networks are bad indicators of people's preferences, i.e., I may like sushi but my good friend loves McDonald's.
• Behaviors are good indicators of people's preferences, i.e., two people who live on the 1980's section of a music website are likely to share similar tastes in music.

Sanjiv Das, Morgan Stanley

Elements of Advice:

• Data; which is transformed into
• Information; which is used for
• Analysis; from which one can gain
• Insight.

Other:

• The organization of the MS database is not driven by meta search but by organizing principles.
• Data will be a commodity, but organization of the data may not be a commodity.
• Two key outputs of MS modelware group: Operating Performance and Financing Performance (a methodology and framework for performance attribution).
• Where is the performance coming from? What happens if we stress performance by flexing key macroeconomic variables, i.e., interest rates?
• Not in the business of analyzing historical performance but in forcasting future performance.
• Another input, Non-Financial Operating Data (i.e., online discussion concerning products, etc.) is also valuable.

Everyone at the conference made major contributions to the dialogue, and this brief post doesn't begin to give the discussion the justice it deserves. That said, I've attempted to highlight some of the salient points that resonated with me as both a blogger/publisher and as an entrepreneur in this data-driven world. Thanks again to everyone for participating. You taught me a lot.

Comments

[sugiarto Setiabudi]

I need to address my strong concern regarding data integrity over internet.

People need independence body to ensure integrity of data over internet or have fiduciary duty to deter any inappropriate alteration .

IT compliance should be priority among the other thing.

[Jason Wood]

Roger,

Great to hear such excitement particularly for a New York event. Did you get much color on Reuters advancements on tagging their data inputs? While early in the process, I see that as a massively power way to generate arb for first movers.

Best,

J