6 ways CIOs can make valuation of data a business driver

Market value of info-centric businesses

Businesses that value information, like Google, appear to enjoy a higher-valued stock price; companies productizing their data, like The Kroger Co., are building new lines of business and revenue streams; organizations like San Diego's Scripps Health see infonomics as a framework to usher in mission-critical data governance reforms.

But how far along are companies in the data valuation process? "You probably have a better accounting of your tables and chairs than your own information assets," Laney said. "Now ask yourself: Which one actually generates more value for your company?"

While few companies are recording data as an asset on their balance sheets, the marketplace does put a premium on "information-centric" companies, according to Laney. As defined by Laney, these are companies that have invested in their data by hiring data scientists or a chief data officer, or by building a data science organization or a data governance function. Companies that meet these criteria, including Netflix, GlaxoSmithKline, Nokia, Apple, American Express and Ford, enjoy a market value to tangible asset or book ratio that's two to three times higher than the norm. "I'm not going to say there's a causal relationship, but it certainly is an interesting correlation," Laney said.

Information product businesses, or companies that sell data in one form or another, have an even higher market-to-book ratio, according to Laney's research. Google, Yahoo, Moody's, TripAdvisor, Harte Hanks and Dun & Bradstreet, for example, have a market-to-book ratio more than four times higher than the norm.

Data valuation is everyone's business

Doug Laney

Lest CIOs infer that trading on data is just for data brokers or internet giants, Laney's research also turned up traditional businesses that have figured out how to productize their data. The Kroger Co., for example, is generating $100 million in incremental revenue per year by selling its inventory and point-of-sale data and "making that available as a syndicated data provider," Laney said.

Stories like this one will slowly become more commonplace. Indeed, Frank Buytendijk, research vice president at Gartner, said companies should recognize the growing demand for information products as a potential source for competitive advantage. "There's true business opportunity out there to be excited about," he said.

Frank Buytendijk

Buytendijk said every CIO was sitting on data that could be productized -- in six months. He pointed to the public sector, where cities like Boston, Palo Alto, Calif., and Chicago are opening up budget or public works data to create more transparency and connectivity to constituents.

And he pointed to established companies such as General Electric. At 126 years old, the company is striving to reinvent its services business by capitalizing on sensor data to predict when industrial equipment such as a jet engines and wind turbines requires maintenance.

Even a seemingly low-tech company such as John West, a U.K. canned-seafood manufacturer, is figuring out how to capitalize on the valuation of data to enhance the customer experience, Buytendijk said. To provide visibility into its sustainability practices, the company tags the fish it catches, gathers data on where its fish are caught and then makes that data available to consumers. "With a little bit of information and master data management, you can identify certain touch points" between product and customer, Buytendijk said. The consumer may not be purchasing company data outright, but the data is helping sell the product.

Create a common language

Productizing data can lead to innovation, but infonomics can also reshape company culture, including C-suite dynamics. "It creates a common language between IT, business leaders and CFOs," Laney said. He worked with a CFO from a financial services firm to create an internal balance sheet to help data stewards and owners designate the valuation of data to the organization.

"Imagine the difference if I tell you that you're in charge of our customer database versus you're in charge of our $50 million customer information asset," Laney said. "It's a bit of an attitudinal change that can take shape."

Placing dollar signs on data to create a common language across the enterprise is an intriguing idea to Jimm Johnson, enterprise data warehouse liaison for Scripps Health.

"We're saying we have to become an information management company," Johnson said. "We're trying to turn the dial around so people understand that no matter where you are in the organization, you need information."

One of his first steps in the data valuation process was to implement a data governance program. Massive healthcare reforms, including adoption of the electronic medical record, require it, Johnson said. But "explaining what data governance is, what information management is -- any of those concepts -- to the C-suite is difficult."

Data sources are typically siloed, even duplicated in some cases, and multiple versions of the truth are common. "How do we look across the entire enterprise and get everyone on the same page with all of the data?" Johnson said. Embracing infonomics, he believes, may help by making an intangible concept -- the valuation of data -- tangible.

A year and a half into his job with Scripps, he started seeing progress -- and momentum, he said. "We have a long way to go. But I think if you can lay out the argument in a coherent, objective way, with concrete examples, it starts making sense to people."

Six data valuation models

CIOs can help generate more value out of their information assets by applying infonomics practices to corporate data.

Working with clients from retail, financial services and technology sectors, Gartner's Laney has developed six models designed to help businesses conduct a data valuation process.

He divides the models into two categories: The first is a set of nonfinancial or noneconomic models, which do not put a price tag on data. "Some of our clients … just want to prioritize or create an aggregate of data quality characteristics to get a sense of what its relative or intrinsic value is," Laney said. The second category is a set of financial models that borrow from established accounting practices.

The six models that will help CIOs determine the valuation of data are briefly described here:

Nonfinancial methods

1. Intrinsic value of information. This model doesn't "take into account the business value at all," Laney said, but focuses instead on the data's intrinsic value. The model quantifies data quality by breaking it into characteristics such as accuracy, accessibility and completeness. Each characteristic is rated and then tallied for a final score.

Laney, who teamed up with Gartner's Ted Friedman to quantify a dozen data quality characteristics, includes scarcity in the equation. "Data that's more unique to your organization and not available to your competitors or the larger marketplace, we believe, has the potential to provide more value to you," Laney said. As with any of the six data valuation models, this one can be tailored to the company, which could, for example, "assign weighting factors" to each characteristic, he said.

2. Business value of information. This model measures data characteristics in relation to one or more business processes. Accuracy and completeness, for example, are evaluated, as is timeliness "because even if data is relevant to a business process, if it's not timely, how valuable is it really?" Laney said. The model can be tailored to fit the organization's needs and even applied to specific data types such as unstructured data or third-party data.

3. Performance value of information. This model is "much more empirical in nature" because it measures the data's impact on one or more key performance indicators (KPIs) over time, Laney said. Take the sales department, for example. "If your salespeople had access to competitor pricing data, how much quicker could they close sales?" Laney said. Businesses can run an experiment by comparing how a control group with no access to competitor pricing data performs against an experimental group. Or, if businesses have neither the time nor the ability to run an experiment, they can substitute proxy data for control group data, he said.

Financial methods

4. Cost value of information. This data valuation model measures the cost of "acquiring or replacing lost information." After the Sept. 11 terrorist attacks, as clients began calling Laney to figure out how to recoup from, in some cases, a total loss of data, they developed a method to quantify information's value based on what accountants "refer to as 'replacement costs,'" Laney said. A value is assigned to the data by measuring lost revenue and how much it would cost to acquire the data. "This is the way valuation experts value most intangible assets that don't have a discernible market value or are generating a market stream," he said.

5. Economic value of information. This model measures how an information asset contributes to the revenue of an organization. "This is our KPI model again, but instead of any given KPI, we're looking at revenue," Laney said. To illustrate his point, he returned to his sales example. An experimental group is given access to competitor pricing data and a control group isn't. "Instead of looking at time-to-sale, we're looking at revenue generated by any given salesperson" over a given period of time, Laney said. "That will give us a good sense of the value of that data." CIOs should factor in the cost it takes to acquire, administer and "bake that data into the system the salespeople are using," he said. They should also consider the data's lifespan. Competitor pricing data, for example, has a shelf life, which should be factored into its value.

6. Market value of information. This model measures revenue generated by "selling, renting or bartering" corporate data, which Laney considers one of the best ways to reap value out of an information asset. The problem is, most information assets don't have what accountants call an "open arms-length market," or what the price of the data would be on the open market, according to Laney. A way around this is to figure out what similar data from syndicated data providers or competitors is going for. After determining the data's premium price, Laney suggests figuring out what he calls a "discount value." "When we sell data, we're not really selling it," he said. "We're licensing it." The discount rate will vary based on the number of times a company sells the information and other factors. "But, again, it's not the value that's important," Laney said. "It's tracking over time."