Wednesday, March 31, 2010


Eytan Adar's latest project is Zoetrope, a way of interacting with a web page's history. Rather than just seeing what's currently on a web page, Zoetrope let's you easily rewind to previous versions of that page.

Some of the applications mentioned in his video at are for viewing how a value on a web page changes over time, for example:

* Stock price
* Movie rating
* Gas prices (e.g. on gas buddy)

What is the relationship between gas prices and oil prices is a question that could be answered because it allows you to see how those values relate to each other. The video describes a really rich scenario for exploring relationships between data.

What if the correlations between changes in time-series data though could be automatically detected? What if we had the ability to mine the web for time-series data and automatically detect which events seem to be causing other events?

What's needed is a uniform data structure for representing time series data. Then we'd need to crawl the web and map the data into that format. Zoetrope adds another element to this wherein time series data is captured by monitoring the changes in a website over time.

I think it is very cool that Eytan is using the term "lens" to talk about ways of viewing data harvested off of the web. At NetBase, we use the term lens also, but for viewing data extracted through our semantic analysis.

Sunday, March 28, 2010

How to be an innovator

Not much guidance is out there for how to become an innovator. Sure there's lots on how to be an entrepreneur, an inventor, or a leader. But if you want to be an innovator, here is a how-to tip.

I heard two very interesting innovators speak last week at Eric von Hippel's Innovation Lab. One was Bre Pettis, founder of MakerBot. The other was Eric Jackson of Jackson Kayak. Von Hippel would call both of these innovators "user manufacturers". This is because they are practitioners in their field and entrepreneurs who have created companies to provide products to other practitioners in their field.

Both men talked about their passion for their field. Eric was an avid kayaker who entered and won all kinds of kayaking contests. Bre loved 3D printers and was fascinated by the idea of 3D printers that could print our 3D components needed to build other 3D printers.

At some point, both men realized they could spend even more time kayaking and building 3D printers if they could create and sell products necessary to kayak and build 3D printers, respectively. It was this critical realization that enabled both Eric and Bre to devote their lives to their passion.

So the lesson to those of us who would like to innovate is: find out what you're passionate about it. Then figure out what products and services you can develop to sell to other practitioners of your craft. But in the spirit of user-innovation, you also have to be careful not to become too much of a manufacturer and not enough of a user. Eric Jackson says that he practices kayaking every single day with his lead user buddies. He competes in all kinds of kayaking competitions. He said if he lost a day of practicing, he'd lose his edge as an inventor for his customers.

Monday, March 08, 2010

Technique – the Kozinetnographic method

I realize part of the results may be due to my experience in pulling insights or fluke of nature or or or...
I am using WordPress, a ballpoint pen, and the margin of my copy of Netnography
Anyway, I read Rob Kozinets’ opinion of trying an analytic coding on the entry followed by a hermeneutic interpretation of the post.
It does work, a nice tiger fleck insight (fairly deep) consistent analytical time (in the 20’s) good to excellent cultural immersive learning overall.
So as most things analytical are academic I am dubbing this the Kozinetnographic method....

My tiger fleck insight

Rob Kozinets sent me his Netnography book a couple of weeks ago and I've been enjoying it quite a bit. Last night I got to the Data Analysis section and thoroughly enjoyed the exercise on the posting (search for page 122 on Amazon).

At first glance I thought it would be impossible to extract any insight out of a passage of text riddled with terms foreign to me about a subject I could not possibly be more ambivalent about...espresso. But I gave it the old college try, and applying the netnographic technique I was able to find almost all of the same insights as the professor.

I'd like to contribute an insight to the analysis of this posting. I assumed going into this reading that taste was the most important quality of an espresso. I'm largely influenced by a new chocolate factory in San Francisco called Tcho. What makes Tcho special is it cuts through the chatter about percent cacao and country of origin...when it comes to chocolate the property that matters most is taste.

Figuring the same was true about coffee, I was surprised to read how Frank Rinetti described this great shot of espresso he made using the "Smith(inni)" method. He did not come out first and say the taste was excellent. He listed something about tiger fleck crema and consistent shot time first. And then when he described taste it was only on a range of good to excellent. Why write home about a coffee brewing technique unless it yields consistently excellent taste?

Well I can only imagine that in the realm of coffee, taste is not paramount. I don't know what tiger fleck crema is still but I've seen a picture of tiger fleck and it's quite beautiful. So I'm guessing that appearance of the coffee is as important, if not more important to coffee drinkers than the taste. If I remember how bitter the first sip of coffee I had was, it's frankly not surprising to me on reflection that taste isn't more important. The only "good" tasting coffee I've had was actually hot chocolate with a splash of coffee in it. If you haven't noticed, I like chocolate...a lot.

I also observed the contrast in language used to describe the crema and shot time vs. the taste. Not only was taste listed last in terms of important properties but the qualifiers were also very plain. The taste was described merely as "good" to "excellent". But the crema was described in such rich visual terms that that I had to do a google image to search to relate to it. And the shot time--still not sure what it is--was described with its own measurement system. The author said it was "in the 20's". Since no unit of measure was specified, I can only assume that the unit of measure is so well known in this community that it could be omitted.

When rich vocabulary and shared knowledge about a measurement system appear in forum content, I believe it means those properties have been given a tremendous amount of thought by that community.

Now contrast the richness of the description "tiger fleck crema" and "shot time in the 20's" with the plainness of "good to excellent taste". That's a stark contrast in my opinion and it makes me wonder if taste isn't as important as other characteristics in the realm of coffee connoisseurship.