What is the True Value of Data?
UNLIMITED DATA | BY JAMES KULICH | QUICK READ
How much is your data worth? This is a bit of a tricky question.
On one hand, the global consulting firm Gartner, Inc., noted in a press release that the global business value derived from artificial intelligence rose to a projected $1.2 trillion in 2018—up 70% from 2017—with an expectation that this value will rise to $3.9 trillion by 2022.
On the other hand, current accounting rules do not allow the value of data to be booked as a tangible corporate asset. Clearly, we are in a time of change.
The Value of Data in Today’s ‘Prediction Machines’
How exactly do data create value? Economists Ajay Agrawal, Joshua Gans and Avi Goldfarb offer an intriguing answer to this question in their book, Prediction Machines. They draw an analogy between the state of artificial intelligence today and the internet of 1995. That was the year in which the U.S. government fully allowed the internet to carry commercial traffic, leading to an initial public stock offering by Netscape (an early web browser) that topped $3 billion.
Nobody knew exactly where this would lead, but the race was on.
Being the good economists they are, Agrawal, Gans and Goldfarb make the point that full arrival of the internet made digital interactions very cheap. Anyone could now search for information or buy and sell goods on the internet. Soon, everyone did, leading to previously unimagined stores of value that we see today throughout the world of e-commerce and beyond.
What about data? The authors make the point that artificial intelligence now makes predictions cheap. Organizations can now use data of many types to predict everything from traditional business outcomes like inventory management or demand forecasting to the actions a human driver of a car might take based on inputs from a variety of sophisticated sensors.
We don’t yet know where this may lead, but the new race to create value from data is on.
Creating Value From Data
The good news is that you can create value from your data, whether your organization is big or small. Throughout our curriculum in Elmhurst College’s master’s program in data science, we guide our students in a systematic approach to moving from their data to outcomes that matter.
The first step is to frame, in clear business terms, the goals you look to achieve. Next comes careful preparation of your data. Now, you are ready to use the power of machine learning to create models. How will your models fare in the wild? You need to test and validate your results.
Finally, you are ready to work with others in your organization to deploy your prediction machine.
Our students get good practice with this protocol. Some examples of projects they have completed are:
- Providing sales representatives with a tool for estimating the likelihood that a quote will be accepted
- Optimizing deployment of customer service personnel
- Forecasting store-level demand on regional warehouses
- Identifying key emotions from facial expressions of job candidates
- Determining customer sentiment from Twitter data
Ultimately, your projects will have impact when they marry the information gain your models produce with effective project and change management—the human side of data. We’ll see more examples of this in future posts.
Let me know your thoughts and ideas at firstname.lastname@example.org.
About the Author
Jim Kulich is a professor in the Department of Computer Science and Information Systems at Elmhurst College. Jim directs Elmhurst’s master’s program in data science and teaches courses to graduate students who come to the program from a wide range of professional backgrounds.
Learn more about Elmhurst College
Request information today!