Bronte Media

The Wall St Journal recently had an intriguing article on Prosper, the P-2-P lending marketplace. The article is excellent in its own right, but what I found most interesting is that a site, savagenumber.com, has taken the time to continuously monitor the marketplace of loans on Prosper. How well they perform, the interest and default rates for lenders, and so on. I absolutely love the fact that someone has done this, and is another trend towards greater transparency and at ever increasing granular levels. There are other great businesses built upon eBay’s market data too.

Using crawlable data is obviously a subject close to my heart but the article got me thinking about my second best idea. Chris Yeh recently ran an event whereby he solicited a group of second-best-ideas and the audience of the event duly voted on it. I wasn’t able to make the event but here’s my second-best-idea anyway:

The crux of the idea is basically a search industry analytics provider targeted at search marketers and perhaps also hedge funds/investors looking at stocks like Google and Yahoo.

The idea is to stich together a number of different data sources. The main one would be campaign data from large search marketers and search marketing agencies. They would participate in the inner circle and receive very granular analysis of keyword costs and trends, and over the longer haul as conversion events become more standardized, value creation.

No one would ever be able to break out the activity by advertise source, only compare against the industry average (as defined by a portfolio of keywords).

The second piece is user search behavior. The idea being that keyword history could somehow be reconciled through a service like Alexa and the APIs and Tools of Google and Yahoo. The adware guys would also be able to become an inner circle data partner for their keyword history (i.e. keyword per user, which would measure how often the average user searched for mortgage for instance).

The financial results of Google and to a lesser extent Yahoo are used as a continual quality assurance mechanism.

In the end you would be able to see at a 30,000 foot level that search is a $10 billion dollar industry but also to click down to the level of the keyword and examine for instance that digital camera was being searched 10% more and the cost of digital camera keywords rising by 5%, for instance.

It would also clean out the noisy metrics that are out there that are based on what people bid for keywords, which is very different to what they actually end up paying for the clicks.

Marketers could benchmark their spending and ultimately the service would aid them in discovering new profitable keyword activity, as the efficiency of the current campaigns comes under the attack of competition. The smarter people are already trying to do it.

Hitwise could possibly launch the service but their data seems too aggregated at a high level and just the first piece of the equation (i.e. raw searcher data without the cost and efficiency metrics built in).

The challenge is that the first 10-15 inner circle data relationships with advertisers/agencies are the key. Getting the first few to contribute without the others having contributed. But that’s chicken and egg and business.

The trend that is helping is that so many sophisticated developers and agencies are creating bidding platforms that abide by xml standards and so repurposing (i.e. stripping any sign of individual client or identifiable information) that data is relatively easy. I.e. the challenge is in the relationship not the technical possibility of making it happen (still a very tough challenge).

And so there you have it, my second best idea Thoughts?