Update:: Just found this picture on Flickr of when the diagram below was created.

Doc Searls (Cluetrain editor, Linux Journal senior editor, Harvard Fellow, revolutionary in a Firefox shirt, and all-around nice guy) is embarking on a very ambitious project at Harvard whose impact could be exceedingly far-reaching, called “Vendor Relationship Management” (in a parallel to “Customer Relationship Management”, just the other way round).

He told us more about it yesterday at the Internet Identity Workshop. While it is difficult to claim that numbers like this don’t look like they are exaggerated, I would estimate that if this kind of model becomes successful, not only billions but trillions of dollars are going to be shifted from one place to another — in short order. Steve Gillmor, who went to the session, went as far as saying that in his view, it was a given that this was going to occur, and in a reasonably short time frame. And all enabled by user-centric identity …

So how does it work? The basic idea here is that instead of you thinking what you want to buy, and then going “like a bee from flower to flower” (to quote Doc), from vendor to vendor on their terms, you create some kind of a “personal RFP” on your own terms, and let vendor offers come in.

Here is an example: let’s say you want to travel to Boston for a week, have a meeting on Tuesday in one place, and another on Thursday in another place, stay in a hotel with free WiFi and use your Gold status on United. Today, you have to figure it out all yourself, such as looking through the United website the way they want you to look for a flight on it (and only a flight), then the hotel’s website, and then map to figure out whether or not to rent a car and which rental agency has the MP3-player enabled car that seats 6 that you want.

With Vendor Relationship Management, you would construct an RFP saying what you want, put it somewhere (e.g. on your blog), and vendors would propose solutions to your problem. No self-assembly required.

The following diagram emerged on the white board during the session, which captures the flow:

At the heart of the process is the Personal RFP, which captures the information expressing what you intend to buy. It is assembled — preferably partially automated — from a variety of information sources, such as:

• written / assembled by the user
• the user’s identity information, such as residence address or frequent-flyer membership
• the user’s preferences, such as non-smoking hotel rooms
• the user’s observed actions, such as (thanks to Steve Gillmor for this example) the fact the he never uses Microsoft Office, but only Google’s on-line tools, which is a strong indicator that he prefers to fly on an airline with internet connectivity (this is a hard one, but very compelling)

This Personal RFP is then shared with a number of vendors, e.g. by submitting it to them, submitting it to some intermediary, or just publishing it to the cloud. Vendors decide whether or not to create an offer that addresses the RFP, and submit to the user. Through a process of comparison (e.g. determining the solution match with some kind of quality function), the offers are ranked and the user makes his decision.

I would love to blog “my bread machine is broken, I’m willing to entertain offers” and then have vendors offer me what matches my actual bread-baking behavior (and not what they want to sell me because it creates the highest sales commissions). Instead of opening dusty boxes in retail stores, or trying to wade on Amazon’s terms through their particular way they think I should be thinking about buying a bread machine. Of course, this is in no way limited to bread machines but applies to pretty much everything in the consumer world and much in the business world. I would think that the power of IT (or ICT) that we now have at our disposal can make this process affordable for a mass market, and not remain reserved for high-ticket items in a business-to-business context only.

It’s clear that for this to work, we need user-centric, portable identity. But given all the adoption and buzz that OpenID is getting these days, and given that Windows Vista (and thus CardSpace) is now actually being shipped, that doesn’t seem to be much of an obstacle …

<gazing to the stars>trillions … wake up Johannes, do something about it! ;-)