Subscribe to DSC Newsletter

3 Theorems on the Economic Value of Data

Since releasing the University of San Francisco research paper on “How to Determine the Economic Value of Your Data” (EvD), I have had numerous conversations with senior executives about the business and technology ramifications of EvD. Now with the release of Doug Laney’s “Infonomics” book that builds upon Doug’s EvD work at Gartner, I expect these conversations to intensify. In fact, I just traveled to Switzerland to discuss the potential business and technology ramifications of EvD with the management team of a leading European Telecommunications company.
From these conversations, I am starting to form some “theorems” to guide organizations regarding how EvD could impact their business and technology investments. A theorem is defined as “a general proposition not self-evident but proved by a chain of reasoning; a truth established by means of accepted truths.” Well, it might be a stretch to call these “theorems” at this point, but I hope over the next several months to turn these “observations” into “theorems.”
Also, I fully expect the number of theorems to grow as the EvD concepts mature, especially as organizations look for data and analytics to the fuel their digital transformation initiatives.

Economic Value of Data Theorem #1:

It isn’t the data that’s valuable; it’s the relationships and patterns (insights) gleaned from the data that are valuable.

We highlight the difference between monetizing data versus monetizing insights when we discuss the Big Data Business Model Maturity Index (see Figure 1).

Figure 1: Insights Monetization Phase of Big Data Business Model Maturity Index

Phase 4 of the Big Data Business Model Maturity Index is the “Monetization” phase. However, organizations should not focus on the monetization of their data, especially selling their data. Selling data is a business model decision, not a business transaction. And there are significant liabilities that await an organization that moves into the business of selling data .

Instead, organizations should focus upon the monetization of the insights derived from the data. The monetization value isn’t in the data; the monetization value is in the unique customer, product, service, operational and market insights that are gleaned from the data. It is from these insights that organizations will be able to identify new services, new products, new customers, new markets, new audiences, new channels and new partnerships.

Economic Value of Data Theorem #2:

It is from the quantification of the relationships and patterns that we can make predictions about what is likely to happen.

It is the quantification of the relationships and patterns around customers, products, services, operations and markets that drive operational, management and strategic predictions. And it is the value of these predictions (in support of business use cases) that ultimately determines the economic value of your data. We want to quantify the relationships, patterns, propensities, tendencies, biases, preferences, associations and affiliations at the level of the individual customer, product, service, operational process and markets (see Figure 2).

Figure 2: Uncovering Relationships and Patterns in the Data

From these detailed insights, organizations can make predictions about their customers, products, services, operations and markets: what products and services customers are likely to buy, when customers will likely have a life stage change, what products are likely in need of servicing or retirement, what operations are likely candidates for operational optimization, what markets are likely ripe for new products or services, etc.

These predictions, though never 100% accurate, give organizations an “edge” in their operations, management and strategic decisions and use cases. For example, having better predictions about which customers are likely to attrite and the predicted lifetime value of those customers gives you an edge over the competition. It may not be much, but sometimes it is the smallest of edges that can separate the winners from the losers.

Economic Value of Data Theorem #3

Predictions drive monetization opportunities through improved (optimized) strategic and operational use cases.

It is application of predictions against business use cases (i.e., clusters of decisions) that determines the economic value of the data. For example, it is neither sufficient nor actionable to know that there is an increase in head injuries, lacerations and broken bones during and immediately after a local professional football game. That’s interesting, but not actionable.

However, if you can predict a 37% increase in head injuries, lacerations and broken bones during and immediately after the professional football game, then that is actionable! With that prediction, I can now make recommendations (prescriptive analytics) about extra nurses, doctors and supplies one might need at the hospitals nearest the stadium.

The Future of Economic Value of Data Theorems

I can see the potential for more theorems as the EVD discussions mature. I can see, for example, a theorem on “variable predictability” and its importance in attributing financial value to the appropriate data sources.

We will continue to explore, test, fail and learn as we seek to perfect the methodology and formulas that can help organizations determine economic value of their data sources. I believe that this will become a business mandate as organizations look for a management framework to help them optimize the business and technology investments that are driving digital transformations.

Views: 4066

Tags: AI, Analytics, Big, Business, Data, Database, Intelligence, Internet, IoT, Lake, More…Science, Things, artificial, digital, intelligence, to, transformation

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Bill Schmarzo on November 25, 2017 at 5:11am

Check out the supporting University of San Francisco white paper for more on the math that we used.  It was pretty rudimentary, but am working with our data science team to come up with a "Variable Attribution" formula based upon "Variable Predictability."

Also, thanks for sharing Doug's material.  It's outstanding! 

USF EvD Research Paper
Comment by Phil Phillipuk on November 24, 2017 at 10:22pm

From "The Fifth MIT Information Quality Industry Symposium, July 13-15, 2011",  see Mr. Laney's presentation, specifically, pages 11-13 of the .pdf document.  There, you'll find several general formulas for the valuation of data as an asset, which I've taken as suggested starting points.  They cover six valuation models:

  • Intrinsic Value,
  • Business Value,
  • Loss Value,
  • Performance Value,
  • Economic Value,
  • Market Value.

Infonomics has been the most practically applicable method I've come across, and used it for prioritizing and planning data projects.  Mainly by answering key questions of relevance, ubiquity, etc.you begin to paint a picture.  If you can attribute concrete figures like sales or acquisition to the use of data, the picture becomes clearer, and the exercise becomes similar to ROI analysis.

However, I've yet to apply any math to pinpoint actual value, mainly for lack of resources and organizational will to launch such a project.

Comment by Bob Wakefield on November 24, 2017 at 9:37pm

I was hoping to see some math here...

Comment by Phil Phillipuk on November 9, 2017 at 4:48pm

Done.

Sometime in the late 90's or early 00's, Doug Laney of Gartner coined the term, "Infonomics," which is all about the valuation of data and information as an asset.  

He wrote a book on the topic, which was released in Sep 2017.  Following is from his blog on why he wrote the book:

"What do I hope to achieve from writing the book? In short, it woul...

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service