The Velocity of Data

While eating Chicken Adobo at Cendrillion every other day for the past 6 months, we talk constantly about the value of data - how can an application or a technology create value to an existing or new group of users through the creative aggregation, manipulation or structuring of data.

One of the favorite theoretical examples we use is the ability of an application to add structure to unstructured data. And, in particular, in ways that add value to end-users, publishers or advertisers, three specific groups we seem to care about at betaworks. We seem to be attracted to those examples, but in any event this can get pretty heady and theoretical and wonky.

And then sometimes the theoretical turns specific in a way that validates all the lunchtime chattering. A few months ago Summize turned their search application toward Twitter - in a way Twitter is the very definition of a mass of data -- conversational, casual, and totally unstructured. By adding search to that mass of data - and including including things such as trending analysis and other ways to search - by user, geography, sentiment, places and dates, for example, the unwashed mass of Twitter info becomes . . . structured data. Today I found out about Plurk and that Firefox RC3 was released, to name just two minor and immaterial examples from the trending topics.

So maybe this is an example of where adding structure creates new ways of even thinking about a data set. Or, much more broadly, where adding structure can take an application and make it into a platform.

Segments and Data

Sramana Mitra:

I would submit, that Yahoo hardly needs to do anything drastic, except, that it needs to reorganize its entire portfolio into Segments and Lifestyles, and align the 4Cs of each segment, so that the people interested in advertising to a particular segment can reach the community in a focused way, in Context. If they want to sell Groceries to people reading recipes, they should be able to reach them right there. Or, if they have Gadget Geeks researching a cell phone, they should be able to access that eye-ball, in Context.

Amazingly insightful. Audience and user segmentation, I believe, is the key to unlocking value from community driven publishing and media models and businesses. The ability to segment, in turn, requires a discrete level of correlated user, usage, media and context data. It needs to be collected (efficiently, which the right detail to drive its organization, which I submit is not trivial), stored, analyzed and delivered in actionable pieces, to publisher and advertiser.

While this level of analytics is only now beginning to be available, it is, as I have written before, remarkably powerful.

Advertising, Data, once more

Dave Morgan (via Andy Monfried):

We have unique consumer insights and can tell advertisers and their agencies valuable things about their audiences and their campaigns and their products and services--things that they didn't know, and need to know. This can help them reach consumers in places that are not apparent to them right now, thus opening up underdeveloped inventory. Over time, I think that we will find that online ad-based consumer insights will be more valuable than the media that we use to deliver the ad and capture the insights.

Advertising = data.