Torbz - Fotolia

Data monetization takes three routes to turn data into dollars

Data monetization is not monolithic. Instead, it takes several forms, according to MIT researcher Barbara Wixom. Each approach has its own unique traits.

In recent years, the term data monetization has gained a fair amount of currency, without actually gaining too much clarity. To cast a better light on the matter, MIT Sloan School of Management Center for Information Systems Research's Barbara Wixom decided to find out what companies meant when they said they were doing data monetization.

More than 50 interviews with more than 50 executives provided more than 50 different definitions of data monetization. But these interviews uncovered recurring patterns, according to Wixom, who is a principal research scientist at the Center.

Wixom's and her colleagues' research found three distinct approaches that mark data monetization initiatives: using data for improving operational processes and decision-making; enriching, or wrapping, existing products or services with data and analytics; and the outright selling of data generated by the organization's activities.

In a recent interview, Wixom discussed these various approaches and their implications for data quality levels, organizations' structures and more.

It seems the world has a deeper appreciation for the value of data than it had just a few years ago. What is your research telling you about this phenomenon as it plays out in organizations?

Barbara Wixom: What data monetization is about is simply generating financial returns from data. That means turning data into dollars in a top-line or bottom-line sense; getting returns from data that show up on a financial statement.

Barbara Wixom, MIT Sloan School of Management CISRBarbara Wixom

When we started looking at the different ways companies were doing this, we realized there were these three ways in which they were making money. Two of the approaches were indirect; that is, through process uplift or product uplift. The third way was a direct conversion; that is, selling.

As we explored these three approaches, we realized what varied was what was required to maximize the return. Because these approaches were distinct, they required different types of capabilities and different commitments.

Let's talk about improving processes. Sometimes, people feel processes have pretty much been harvested, and that big-bang innovation is required instead. But you've found differently, and cite a pretty well-known company as an example of process improvement and innovation.

Wixom: Sure. When it comes to improving processes, there still is a lot of money to be gained through improvements. People who have been around the data management world for a long time know companies have been improving processes with better data since the inception of decision support systems; that is not new. But the chances to create value are still great, and there is still a lot left on the table to pursue.

An interesting example is Microsoft, which created a new integrated customer system to make improvements in sales productivity. They put in place process management that made sure everyone was consistent in the way they pursued sales leads. At the same time, the company was also transforming its business model. In fact, improving processes was actually a way to achieve the completely new types of performance outcomes required for the company's shift to selling cloud services.

This wasn't just about capturing low-hanging fruit. Improving processes also meant achieving success in a new way of doing business. They used data to guide decisions in selling in ways they never had before. What was most striking was that Microsoft had to understand how to sell something completely different. People were used to selling product licenses. Now they had to understand how to sell cloud services.

Internal process improvement meant that, instead of counting up products and understanding leads associated with products, they had to understand present customer usage and future usage, which gets into customer sentiment analysis.

Another path to data monetization is wrapping information around products. You've cited FedEx among these models. Certainly, that company's online package tracking application for customers is a hall of famer.

Wixom: Yes. Back in the day, FedEx's tracking application was game-changing. But people should realize the work that was behind it. It meant that FedEx had to up its data quality level. Both the data and the platform requirements for wrapping, as FedEx found out long ago, are quite significant.

Wrapping opportunities -- where you use data and analytics to solve customers' problems -- are very different than improvement opportunities, which are internally focused. When you're wrapping, you are exposing products externally, and you really have to make sure you upped your game from a data perspective.

Wrapping is similar to selling, in that both are externally facing. The difference is that, for wrapping, the value comes from the product lift, which means the data and analytics add value to whatever the offering is. In the case of FedEx, that was the actual package delivery.

But, for selling, the data and analytics have to be able to stand on their own, and the value has to be inherent. The data and analytics have to be solving a problem for which a market will pay.

Can it be said that, when people put their data out to the public, they find things about it that they didn't know before?

Wixom: You find, when you are externally facing, that you have to understand the importance of high quality for data more than ever. If there are any problems, they will be found. There is risk to the organization, and to the relationship you have with the customer. Even with wrapping, you have some risk to your core offering.

This isn't about slapping some reports on a product. This is about delivering data analytics at an appropriate level of quality. And most organizations aren't set up to do that, nor are they likely to have that mentality around their data.

The key to selling data is appreciating that selling has its own business model and that selling is really hard. There are information businesses, and all they do is sell information. And if you talk to them, you quickly appreciate that what they do is different.

If you are going to get into that business, it's going to have to have its own reporting structure, whether it is a division or a spin-off company. You need a leadership team that understands the information business to run it, and you need fast innovation to be able to change as customers' needs change. It is a very different type of product development lifecycle.

While there are usually these three ways of data monetization, you won't see such distinctions in the digitally born organization. Think of the typical cast of characters -- Amazon, eBay, Facebook, Google and such. There they grow up with the DNA that is all about data. There you find organizations that have found ways to improve, wrap and sell their data, sometimes all at the same time. That, ultimately, is where we all want to go. But few non-digitally born companies are now set up to do that well. So, it is important for them to be very purposeful.

Next Steps

Take a look at data monetization strategy

Discover customer data via the internet of things

How to do the upfront groundwork of data monetization

Dig Deeper on Enterprise data architecture best practices