Data monetization strategies add new business opportunities, IT needs

ktsdesign - Fotolia

Missions for monetizing data need lift from upfront groundwork

Organizations launching data monetization strategies should factor some key initial steps into their plans to develop revenue-generating data products and analytics services.

So, your company is looking to monetize its data? That's a logical plan: Data products, analytics services and other data-centric offerings are on the upswing, as organizations look to turn their growing stockpiles of data into both actionable information and a revenue-generating corporate asset. Clearly, there's money to be made from data, both in large enterprises and data-driven startups.

But your data isn't going to monetize itself. Various steps need to be taken to get data monetization strategies off the ground and push them forward. Let's look at some of the key to-do items that IT, data management and analytics teams will likely have to factor into formal plans for monetizing data.

Setting up a suitable -- and scalable -- data processing architecture. Data monetization initiatives often involve a lot of data, and that calls for some heavy-duty processing power -- in many cases provided by Hadoop, Spark and other big data technologies. Webtrends Inc. is a good example: It uses a 160-node Spark system to stream user activity data from websites and mobile devices into a Hadoop cluster and then runs machine learning algorithms against the info so corporate clients can personalize webpages and marketing offers on the fly. "The idea is that this data moves seamlessly through our system, and it's happening in real time," Webtrends CTO Peter Crossley said in a 2016 interview. Otherwise, the data would be less useful to the Portland, Ore., company's customers -- and less monetizable as a result.

Hiring data scientists with advanced analytics skills. If big data is involved, you're going to need some data scientists or other skilled data analysts to monetize it effectively. They're the ones who can build, test and run the analytical algorithms and predictive models that will produce insights as part of analytics services or go into data products sold to customers. Data engineers also have a possible role to play in helping data scientists pull together data sets and prepare them for analysis. But finding the right people isn't easy: A shortage of data scientists with the needed know-how continues to be the biggest roadblock companies face in big data analytics efforts, according to a survey of 370 IT and business professionals conducted in August 2016 by research and educational services provider TDWI.

Preparing your data for monetization success. Data products have to meet the diverse analytical needs of different customers, so a one-size-fits-all approach to structuring the data that goes into them could dampen user satisfaction and diminish the data's business value. For example, James Powell, CTO at The Nielsen Company, advised against designing data products "for the lowest common denominator" during a panel discussion on monetizing data at the May 2016 MIT Sloan CIO Symposium in Cambridge, Mass. Nielsen, an audience measurement and marketing research company based in New York, does "a lot of careful modeling" of data for analytics uses, Powell said. But he added that underlying data models need some flexibility for external users, which could mandate a new modeling mindset in organizations.

Embedding usable and accurate analytics capabilities. Successful monetization of data depends on it being useful to paying customers. That means providing the right data, potentially from a mix of internal and external sources. The data also needs to be clean and consistent, same as if it was going into a data warehouse or Hadoop system for internal use. And built-in analytics tools must be easy to use and produce accurate results. For business intelligence applications, embedded BI tools from various vendors are a potential option for smoothing the path. For more advanced analytics uses, analytical models have to be tested, or "trained," to ensure that they deliver valid results -- then updated on an ongoing basis to keep them current and relevant to customers.

Creating business processes to support data monetization. Efforts to monetize data depend on actually selling it so it's profitable. That could require new pricing models and sales processes alike. "We underestimated how difficult it would be to sell these products," said Ivan Matviak, an executive vice president at State Street Corp. and head of data and analytics platforms for its Global Exchange unit. Speaking as part of the MIT panel discussion, Matviak added that the Boston-based company had to educate its sales team to sell a data-as-a-service platform, a risk analytics service and other new offerings to reap the rewards of its monetization effort.

Monetizing data isn't for everyone -- not all companies have the wherewithal to become a data business, including data that lends itself to the concept. But for organizations that do fit the mold, implementing a data monetization strategy can almost literally turn data into business gold. Just be prepared for what needs to be done to unlock the treasure chest. 

Next Steps

Read expert advice on three key data monetization concepts

Analyst Howard Dresner discusses embedded BI and data monetization

Three analytics managers offer tips on building teams of data scientists

Dig Deeper on Business intelligence strategy