As per discussions in prior postings on Managing Data as Assets, yet another method for valuing assets is to use market prices (mark to market). This method works best in a market for a homogeneous product where no individual buyers or sellers can affect those prices by their own actions. Securities such as stocks and bonds fall into this category. If you own 1,000 shares of Informatica stock for example, you know exactly what they are worth at any time since there is a highly liquid market for them – it doesn’t matter what you paid to acquire the securities (except for tax purposes) and you don’t have to sell them to determine their value – you only need to look at what they are trading for at the moment.

The challenge with using market prices to value data assets is that there is (usually) no market for them – or if there is a market, they are usually one-of-a-kind assets and are generally illiquid. Houses for example fall into the same category. Each house is one-of-a-kind and is generally illiquid with an average selling period of 3-6 months. If you want to know what the current value of your home is without selling it (e.g. to use it as collateral for a loan) you could hire an appraiser who would compare it to similar houses that sold recently in similar areas and make adjustments up or down for differences. There may be an external market for certain kinds of data, such as customer shopping habits collected by a website, but the value in much of the data that organizations have is specifically because it is proprietary and confidential and hence there is no market for it.

There is another way however to determine market prices for data; establish an internal market. In other words, organize the IT and business groups into data providers and data consumers and establish internal financial processes to govern their interactions. This is not rocket science. There are well established management accounting and reporting processes that companies use all the time to transfer costs from one division to another – often for tax purposes in a multi-national organization or to distribute decision-making and allow each division to operate as a profit center. The basic idea is to apply these existing accounting practices to data assets.

For example, the IT organization could be organized as a collection of cost centers and Information Investment Centers (IIC for short). The cost centers operate infrastructure elements such as networks, shared computer platforms, and core operational systems. The IICs on the other hand utilize transfer prices to charge business units for the data they consume. Each IIC would “own” one or more data domains such as financial, customer, supplier, product, facilities, equipment or research data. The IIC would need to do market research (meet with the different business units to understand their information needs), develop business plans to compete for investment capital, and buy services from the IT cost centers to deliver solutions.

Business units on the other hand would be able to “shop around” with the various information providers to discover what data is currently available or what could be made available and at what cost from various data providers – then make decisions about which data would provide the greatest return for their business area.

The idea may sound radical, but this approach is similar to organizations which have departments that are self-funded.  For example, some organizations have three levels of funding for certain business units, 1) corporate tax, 2) hybrid, 3) self-funded. In the self-funded models, they operate like businesses, generating business and paying for themselves via internal transfer prices. Here’s how it could work for managing data assets:

First, establish a number of information domains that are worthy candidates to be treated as assets at the enterprise level. The number usually isn’t very large – probably in the range of five to 15. For example, typical data domains are financial, customer, workforce, supplier, product, services, facilities, equipment, market, and research data. It may make sense to divide some of the domains into smaller segments such as dividing customer data into commercial and retail, or dividing facilities data into manufacturing, distribution, retail, and customer-owned. In any event, while an organization may have hundreds of subject areas of interest from an operational perspective, the number of enterprise-level asset classes should be relatively small.

Next, establish each of the data domains as an internal Information Investment Center (IIC) and assign management accountability. The IIC managers would typically be senior individuals – direct reports to C-level executives or similar level. Their accountability and rewards, would be based on a combination of RONA (Return on Net Assets) and EVA (Economic Value Added).

It might be easier to explain how this would work with an illustrative example. But this post is already rather long so I’ll start my next one with a specific scenario. Stay tuned!