Data Union

Version 0.1

This proposal is in the public domain. It may be copied and modified -- provided that Jack Krupansky and Base Technology are credited and a link back to this original proposal is provided AND these same use and distribution terms are carried along.

Over ten years ago I came up with an idea for maintaining personal information in a way that would aid in preserving privacy, but also facilitate commerce and enable consumers to be compensated for use of their personal information. The term "data union" is a combination of "bank", "data bank", and "credit union". The essence of the data union concept is that a data union is a place where you can safely and securely store personal information, while at the same time being able to "earn interest" on that personal data. This is analogous to depositing money in a bank or credit union and knowing that you can actually earn some income and businesses and society can receive some value from your deposited money. The idea is that businesses could gain access to your personal information only to the extent that you choose to disclose that information. Of course, your "earned interest" will be lower if you disclose less, but that's a choice for each of us to make on our own. Someone may always opt to have their personal information kept 100% private, subject of course to courts and law enforcement interests. Overall, the idea is that people should feel as comfortable with their personal information in a "data union" as they would be with their money in a "credit union".

Take direct mail as a typical example. A business wants to send out an offer to consumers with certain characteristics. They may be interested in income level, your hobbies, whether you're looking for certain products, geographic location, etc. It's expensive and difficult to gather such information. And, frankly, its an invasion of your privacy. Further, if you have no interest in buying a certain product, the company not only has no need to have all you other personal information, but would probably rather not waste the money contacting you. With the data union, you choose what level of offers you're willing to tolerate. You may also price an actual price tag ("cover charge") on receiving offers. You may choose to receive no offers, or offers for only some categories, or only at certain times, or only on demand when you have a need. The goal with the data union is that the direct mail vendor can submit a query to relevant data unions and get back a one-shot list of "delivery codes" for addressing materials to be mailed. The post office would have a separate path into the data union which automatically maps each delivery code to an address for delivery. The consumer has a choice whether to receive multiple offers from the same vendor within some window of time, and the direct mail vendor has the ability to request a subsequent delivery to the same recipients of the original query. The post office would only be able to accept a delivery code once, assuring that the consumer is not bombarded with offers they are not interested in. Subsequent deliveries would require new delivery codes. The net effect of all this is that the vendor would be able to target offers more precisely, more reliably, and more cheaply, eliminating the need for a significant amount of "junk mail". The direct mail vendors would have more streamlined operations. Each delivery would have a higher value to the vendor, part of which would be given to the data union "depositor" (consumer) in return for their willingness to both receive offers and make available sufficient personal information that the offer is well targeted.

I envision that data unions are organized at several levels.

A non-profit consortium that develops standards and reference software. This group is not in any way involved with actual operations and has absolutely no contact with any consumer data. There is no requirement that only one specification be produced. The consortium would be open enough and well-funded enough that multiple avenues can be pursued, and that "the market" can determine success and evolution, rather than the opinions of a very few high-priests up on the mountain. The consortium would have a rather loose charter so that "minority" or "dissident" members could actually pursue "innovative" approaches even if "entrenched" members object.
Data union software vendors. These organizations (some for-profit, some non-profit) would utilize the standards and extend the reference software to add value and turn them into industrial grade products and services.
"Conglomerates", for lack of a better word, that are interested in making money in the data union space. Each would license the data union specs and software from a data union software vendor. They would physically build out their own network infrastructure. These might be subsidiaries of existing companies, or new ventures. The intention is that the conglomerates would be using software and services from the data union software vendors "as is", both to economize operations, and to eliminate opportunities to engage in any gaming that might be counter to the interests of consumers.
Data union branches. Ala bank branches, these are the store fronts (real or cyber) that consumers interact with for "deposits", "withdrawals", and account maintenance. These could be owned independently, but more likely would be owned by the "big" conglomerates.
Data union "bankers". These are professionals (either employees of the branches or independent "consultants") who assist consumers with the myriad of choices and options.
Data union software agents. This is software that the consumer can use that is 100% responsive to the interests of the consumer (independent of any vendor) and has the logic to accomplish tasks and pursue goals as per the instructions of the individual consumer. A data union banker may in fact assist the consumer with setting up and otherwise "programming" the agent software. The agents would run autonomously within the data union network; whether on a node within the data union, or at any network node chosen by the consumer.

One of the functions of the data union is to pre-screen offers in an effort to maximize the odds that delivered offers are likely to be accepted. This happens in one of three ways:

The consumer specifies enough detail in their profile that vendors don't bother delivering inappropriate offers.
The consumer authorizes a data union banker to do some amount of pre-screening. For example, you want to buy a car or computer or camera or whatever and you specify some general parameters (that may not easily translate into a machine-verifiable description) and the banker is presented with offers and decides whether to pass them on. In many cases a banker could discard a single offer for a wide class of consumers. A banker could also decide how to tweak the consumer's interests profile so that offers are more appropriate. Most offer matching would be done via software. The banker is simply there because the whole process is very far from a science.
Software agents perform a moderate amount of the work of a human banker.
The consumer can opt to be presented with a daily or weekly list of terse summaries of the offers and then simply select which if any of the offers are to be fully presented.

This is only a preliminary, rough sketch of the concept. The goal is to institutionalize the value of personal information.

There's a lot more to it than that, but that's the basic idea. What do you think?

Links for identity-related resources.

Updated: January 30, 2006 08:53:43 PM -0500