Is My Company’s Data Worth Anything?

I’d like to present a framework for data value – not valuation, just the thesis that the data has value whatsoever. In order for data to have value, it must meet 2 criteria.

First, in order to hold value, data must be exclusive in its ownership. My knowing the capital of Madrid is of no commercial value (except, perhaps, on a game show or a bar bet in Valencia) because that data is non-exclusive. Anyone can look it up.

Second, in order to hold value, data must also enable the owner to accomplish something that could not otherwise be accomplished without such data. An example of data failing to meet this criterion might be tracking the direction in which the sun rises for a year so that on day 366, I know to face east that morning to watch the sunrise. My data is not valuable because even a 4^th grader knows that the sun rises in the east. No data is required to obtain that knowledge and act upon it.

1. Size of the data set

Generally speaking, the greater the quantity of data the set contains, the more value it holds. This seems a fairly commonsense argument. If I’m a marketing company, then, all else equal, I’d rather have a data set with 10,000 customer records than 100. But there are exceptions. Imagine that you have only one stolen password, but it’s the password to Fort Knox that would enable you to have them ship their gold reserves right to your door (succeeding where Goldfinger failed.). That one piece of data is potentially worth billions of dollars.

2. Age of the data set

Again, generally speaking, the more recent the data was validated (note, I didn’t say collected), the more valuable the data set will be. Some data can perish quickly, like relationship status or car insurance carrier. Other data is much more sticky, like social security and cellphone numbers. In some cases, older data may be better. For example, a data set may be more robust before a series of data collection/privacy laws were enacted.

3. Depth of the data set

At the most basic level, data is quantified by records and fields. Records is the number of distinct data subjects (such as customers), and fields are the information that the records contain (such as income, physical address, email address). The more fields a data set has, the greater its depth. The record potentially contains more information. But, the more fields a record contains, the more likely that many of the fields will contain garbage information. Quantity over quality is absolutely the name of the game in data.

We are embarking on the third age of data monetization–the first being analog data collection (remember when Radio Shack asked for your zip code, every darned time?), the second being digital collection (as far as social media is concerned, we are the product) and the third age is when data starts to become a fungible asset. Data marketplaces are arising, and some companies are even making loans using data sets as collateral. The most successful businesses will have to speak the language of data value fluently.

Now, let’s assume that a data set meets these two criteria. What then? How do we differentiate the value of one data set from another? Here are some broad themes that can be used to differentiate the value of a data set.