Exclusive Interview: Doug Laney on Big Data and Infonomics

Doug Laney is vice president and distinguished analyst with the Gartner Chief Data Officer (CDO) research and advisory team. Doug researches and advises clients on information monetization and valuation, open and syndicated data, analytics centers of excellence, data governance and big-data-based innovation. He is the author of Gartner's enterprise information management maturity model and is a two-time recipient of Gartner's annual thought leadership award.

Gregory Piatetsky: You are well known for coming up with 3V of Big Data (Volume, Variety, and Velocity) in 2001. Gartner dropped Big Data from its Hype Curve in 2015.
Is Big Data still an important concept ?
If so, how many Vs do you see now (others have added more Vs, including Veracity, Value, ... )

Doug Laney: Big Data is still on some of the Gartner Hype Cycles, just coming out of the "trough of disillusionment" and approaching the "plateau of productivity." Note that we publish multiple Hype Cycles for various technology domains. So, yes, Big Data is more important than ever--just that it has become a more mainstream concept. Most high-value analytic solutions today involve an aspect of Big Data.

As for the three V's, velocity is becoming much more of a factor as organizations look to make more and more operational decisions or automate processes in real-time. Yes, others have suggested other V's like veracity, but these are not measures of bigness, so are not really definitional characteristics of Big Data. Nonetheless, they are important considerations for most data. In fact, some colleagues and I came up with 12 Vs of data that can be used to ensure various aspects of data are managed and leveraged appropriately.

GP: You recently came up with another breakthrough idea of Infonomics, which is the discipline of monetizing, managing, and measuring information, and wrote a book on Infonomics. Can you explain it a bit?

DL: Infonomics is the concept that information is, or should be, an actual enterprise asset. Even though antiquated accounting standards still do not recognize it as one, it clearly meets the criteria. That is, information can be owned and controlled, is exchangeable for cash, and generates probable economic value. Infonomics posits that irrespective what the accountants say, it is increasingly incumbent upon organizations to treat information like an actual balance sheet asset.

Monetizing information as an asset is about deploying it in a variety of ways to generate economic benefits. This can range from licensing it to others, to using it to improve top or bottom line results via process improvements. Generating measurable economic benefits from or attributable to available information assets.

Managing information as an asset involves applying traditional asset management principles and practices to information. This can involve adapting physical, financial, human capital, or other asset management methods. And measuring information as an asset is about gauging an information asset's quality characteristics, business relevance, impact on key performance indicators, along with applying traditional accounting valuation methods such as the cost-, market- and income approaches.

A key theme of Infonomics is that

"you can't manage what you don't measure, and you can't monetize what you don't manage."

Unfortunately, too many organizations fail to measure and manage information with the same discipline as other recognize assets, so they fail to generate sufficient value from them.

GP: Can you name some companies that are very successful in monetizing information, and some that are the laggards?

DL: Some companies like Kroger, Rite Aid and Dollar General are known to license data outright, along with most telcos. Many companies barter with information for B2B discounts or favorable terms and conditions. And companies like Vivint and MISO Energy among others have used infonomics valuation models to identify and generate millions of dollars in new economic value for their businesses. My book includes dozens of examples, and at Gartner we have a compilation approaching some 500 real-world stories. Laggards are typically those organizations in industries where transaction volumes are low, like manufacturing and government.

GP: You write that "Analytics is The Engine of Information Monetization." What are the most promising analytics methods here?

DL: We find that greater economic benefits from information are generated when it is used for diagnostic, predictive or prescriptive purposes, rather than simply producing reports, spreadsheets, or bar charts. And we're seeing more and more of these use cases include machine learning, and text and multimedia analysis. No longer can many organizations afford to hand-craft fully-optimized static analytic models. Business moves too fast and business environments change too fast, so adaptive learning models are increasingly critical.

GP: What are the main challenges for organizations in monetizing information?

DL: First, ensuring levels of data quality (accuracy, completeness, integrity, timeliness, etc.) required. DQ for data products goes beyond typical internal DQ, because internally DQ must be "fit for purpose" whereas for external purposes you may not necessarily know what those purposes are, and in aggregate they require much greater levels of DQ than any single purpose.

Second, determining type of information product (i.e. level of data prep/productization/aggregation). One customer may want raw data, while another may value it more in some derivative form or different format. This may require you, operationally, to be able to deliver information products in a variety of formats and enrichments, or suffer lost sales.

Third, following a classic product management/marketing approach. For some reason, organizations don't think of information products as products, and tend to forgo many of the typical activities in researching, designing, developing, packaging, promoting, and supporting physical products or services.

Finally, protecting against unintended, unlicensed uses. Data is an asset easily copied, which makes it great for licensing over and over, but challenging to ensure against the customer using it in untoward ways. Other than ensuring buttoned-down contracts (We can share a listing of dozens of contractual terms to consider.), we recommend requiring licensed users to be trained on "safe and appropriate handling" of information products, enabling usage auditing, including seeds for any PII, and requiring customers to access the data on your site or via a controlled analytic environment or some monitored API, rather than providing them file dumps or the like.

GP: What is the role of Privacy in Infonomics? How do you expect GDPR (European General Data Protection Regulation) to impact companies that monetize and manage personal data?

DL: Ultimately, privacy regulations along with ethical considerations limit the variety of ways organizations can monetize their data. There are some interesting new methods and technologies to create "digital twins" of information assets that enable integrated analytics while respecting privacy controls. At an individual level, privacy is already a consideration in monetizing one's personal data. That is, we get discounts at grocery stores and the free usage of apps by sharing our personal data. I expect the methods of monetizing our personal data will expand.

GP: Companies like Facebook, Amazon, Google are so hugely valuable because they can monetize so well people's personal information, without, as some argue, adequate compensation to their users. There are various information brokers which try to help people sell their information. Should Facebook compensate their users more and what is your opinion about marketplaces for personal data?

DL: Yes, the so-called digerati have business models based on data monetization. And the extreme valuations inherent to most of these businesses is due to wide gaps between their book value and market value, again due to the fact that their primary asset, information, isn't anywhere to be found on the balance sheet. Also, information has a range of unique economic characteristics I enumerate in the book which make it lower cost and more monetizable than physical or financial assets.

As for not compensating those generating the data, I would disagree. As I mentioned earlier, we get free or highly-discounted access to services and applications in return for our data. These business models are a win-win relationship. If they weren't, we wouldn't be participating in droves. I think there's a low probability of direct personal data exchanges or marketplaces any time in the near future. Those that have been attempted to-date have been abject failures. On the other hand, marketplaces for business data such as DAWEX, Q Data and Data Republic, and industry-specific ones like Onvia and Apervita seem to be taking hold.

GP: What common misconceptions about Infonomics would you like to clear up?

DL: I often hear people say that information only has value when it's used. That's a narrow view, and one entirely inconsistent with the way that other assets are valued. For example, a can of soup sitting on a store shelf, isn't being eaten by anyone yet, but has a definitive value reflected on the store's balance sheet. Organizations need to recognize and measure this gap between their information assets' potential and realized value if they ever hope to close the gap.

GP: What inspired you to write Infonomics book?

DL: It really all started with the 9/11 terrorist attacks. In the aftermath, clients called us lamenting not only the tragic loss of life, but also the loss of their businesses' lifeblood - their data. This was in the days before cloud storage or before many companies had offsite backups. When I found out that insurers wouldn't honor claims for the value of the data they lost (because data is not considered "property"), and that data isn't considered a balance sheet asset, it set me in motion researching and advising on the topic of information value.

GP: In your predictions for Big Data Key Trends in 2018 on you included "Data governance dropping notion of data "owners" in favor of "trustees", and expanding the data steward role to include information advocacy." Can you expand on this?

DL: The concept of data ownership, and in particular the moniker itself, simply encourages people to build data silos and put up fences around their department's data--thereby limiting its value to the organization. If we're to treat information as an actual asset, we should adopt roles like trustees or fiduciaries who carry similar kinds of responsibilities and accountability, without the notion of hoarding. And I believe data stewards should be as focused on offense as they are on defense. That is, rather than spending so much of their time developing policies, they should be out there encouraging the improved and expanded use of the information assets they are stewards of.

GP: What do you like to do when away from computers and information? What recent book you read?

DL: I enjoy competitive tennis and non-competitive golf. Also, biking long distances and throwing dinner parties. And I'm a podcast junkie, with my favorites being 99 Percent Invisible, Radiolab, BackStory, and How I Built This.

Some of my favorite reads lately have included Spare Parts by Joshua Davis in which a bunch of impoverished, undocumented high school immigrants from Mexico beat MIT in a national robotics competition. There's hope for anyone who puts their mind to it, that is, if they're given the opportunity to do so, and not frightened into hiding. Also, The Year of Living Biblically by AJ Jacobs, an outrageously hilarious story of one secular journalist's attempt to live according to the strict letter of "The Law."

Other recommendations include Digital to the Core by my colleagues Mark Raskino and Graham Waller--a must read for leaders trying to digitally transform their organization, and Who Owns the Future by Jaron Lanier, a startling, informative, well-researched, and brilliantly written about our future surrounded by and integrated with technology.

Related:

• The Origins of Big Data
• Big Data: Main Developments in 2017 and Key Trends in 2018
• The 42 V’s of Big Data and Data Science