During my October 06, 2016 webinar “Ten Ways CDOs Fail and How to Avoid the Traps” (replay available), attendees asked a number of great questions about how Gartner sees the role of the chief data officer (CDO). Pressed for time then, I promised to address them via social media, so here goes:
Q1. What role should the CDO play in the establishment of standard information architecture best practices? What is the recommended way to coordinate the application/enforcement of these standards across the entire enterprise architecture?
A1. This one might require a webinar of its own! First, the CDO absolutely should oversee the enterprise information architecture and best practices. As I mentioned during the webinar, some CDOs take on an information purview that spans the entire organization, while others may be chartered with particular business units, types of information, or only “enterprise information,” i.e., information with broad applicability. Most information architecture standards can be enforced via a system of carrots and sticks, such as limiting or allowing access to certain technologies or data sets or other resources. There will always be exceptions, so ideally your standards will address these conditions or how new exceptions are to be handled. Also standards and practices should allow for some dynamism. Change is a constant, right? Also, just as we recommend for data governance, you should establish a hierarchy of precepts (principles > guidelines > policies > standards > procedures). Don’t jump feet first into policies; that’s a recipe for discord leading to failure.
Q2. What kind of “traditional” asset do you feel is data/information in the financial statements? We always found that it depends.
A2. This may be a misunderstanding. I argued that information should be an additional type of balance sheet asset, as it meets the criteria of an asset: a) it’s owned/controlled, b) it’s exchangeable for cash, c) it generates probable future economic value. It also meets the litmus test for an intangible asset, so perhaps there should be a balance sheet line item under Intangibles for information assets.
Q3. What is the usual CDO role in the data lake? Is the CDO charged with creating it? Managing the data resources in it? Governing it?
A3. Again, the responsibilities and resources of the CDO varies quite a bit from one organization to the next. Some CDOs also wear a “CAO” (chief analytics officer) hat. But even if they don’t, yes, I believe the data lake should be part of the CDO’s purview. This would involve defining, architecting, populating/refreshing, governing and maintaining it. Any technologies and services involved may come under the CIO and the IT department, or if they are specifically for data and analytics then they may be selected, installed and maintained by the CDO and his/her department.
Q4. Do you believe the CDO should also perform the role of Chief Data Architect? If not, how should their efforts be coordinated to ensure information management alignment?
A4. I think this is addressed in A1 above. But just to clarify, the CDO probably should not be the chief data architect per-se, but should oversee data architecture as part of his or her responsibilities.
Q5. I would like to connect w Douglas off-line at a mutually convenient time…[name, company and title and email redacted] …thanks.
A5. This is in response to my request to speak with any attendees who are measuring, managing or monetizing information with the same or similar discipline as their formal balance sheet assets, to interview them for my upcoming Infonomics book.
Q6. Are there management consulting firms or other providers that have implemented some of this information asset valuation methodology? It appears rather complex and would be a large effort to drive and advance with an internal team.
A6. Yes, Gartner Consulting has done several information valuation projects based on our models. Also, Cicero Group is the only information strategy consultancy I’m aware of with an infonomics practice. Other consultancies and individuals such as John Ladley with First San Francisco Partners, James Price with the Australian information strategy firm Experience Matters, and Dr. Jim Short at UCSD have each done work in the area of information valuation. Also in Australia, Martin Spratt has developed a neat method and tool for calculating the cost of bad data. The “Data Doc” himself, Dr. Tom Redman, boasts a “a ‘toolkit’ containing dozens of techniques to measure data quality, estimate the costs of bad data,” as well, but I have not seen it yet. And of course Douglas Hubbard of Hubbard Decision Research has built a practice around his “How to Measure Anything” book and subsequent franchise. His valuation approach focuses exclusively on decision making aspects of data. For measuring the economic impact of information risk, Schedule1 in Toronto has developed a sophisticated Risk Value of Information model it deploys at financial institutions. I’m sure there are others, but that should get you started.
Q7. Feedback: I find the term Business Value of Information (BVI) a bit misleading because it sounds like it is more as about the Potential business value of information – not true/direct/actual BVI – that can only be measured by directly measuring the affected business value.
A7. Not a question, but I’ll bite. Yes, the BVI model can be used as both a leading and lagging indicator. As a leading indicator you match information assets to business processes where they could potentially be used. As a trailing indicator you consider where they are actually used. In fact, BVI(p) – BVI(a) gives you a pretty good accounting of which information assets are being underutilized and how badly. If you recall the concept of “dark data”, it’s a measure of darkness. (Se my publication on Gartner’s information value models.) Here’s an overview of them:
Q8. Can you send a link to the real-world examples of analytics?
A8. Sorry, it’s a proprietary library we keep close to vest. However, Gartner clients can request a “written response inquiry“ in which I’ll share a selection of the use cases for a particular industry, type of data, or type of business function. We have hundreds in this “art of the possible” library. If you’re a prospective Gartner client, I can share some by phone or over Webex. Contact Gartner here.
Q9.Value would be essentially “impaired” if there are no identifiable future net cash flows – so if processes are not in place to monetize, then would potential value would be offset by that impairment?
A9. Is there an accountant in the house?! Apparently so. According to Investopedia, you’re close. Impairment is when the expected economic benefits (e.g., cash flow) of an asset has been reduced from one period to the next. Theoretically (because as we discussed, the accounting aristocracy doesn’t allow information to be capitalized), if a portfolio of data, say your customer database, degraded in quality, then yes it could be considered impaired if below its book value. You would then write-off that difference. The opposite situation is also true.
Q10. How would you assign a value to information in an organization? cost? NPV of future economic value (monetization of future net cash flows)?
A10. Our published information valuation models (see above illustration) show a variety of way to assign value. Yes, we recommend including NPV terms in the calculations, but for simplicity the models are presented without them.
Q11. Regarding Monetization…what would be the analogy for federal government agencies and in particular the Department of Defense?
A11. We have worked with government departments in various countries on this issue. The US is somewhat unique in that the Freedom of Information Act (FOIA) prohibits the licensing of information the government compiles. In other countries such restrictions are a bit looser or non-existent. But as I discussed during the webinar, we consider data monetization to include any and all ways in which information can generate measurable economic benefits. So for government departments, the possibilities are almost limitless, including all of the indirect data monetization methods I highlighted, and in the research publication, Seven Steps to Monetizing Your Information Assets.
Q12. Often privacy and consumer protection regs limit our ability to monetize data we’ve acquired. Many firms actually promise their customers that we won’t sell our data to 3rd parties. So, is this really an option in a lot of cases?
A12. No doubt. That’s the #1 inhibitor to licensing or bartering with information. I’m no lawyer, but “selling” and “licensing” are very different animals. Selling involves the transfer of ownership; licensing is the availing of limited rights. Assuming you mean licensing, your options could include selling off part of the business (including said data) to a joint venture. I know of at least one major insurance company planning to do this. Also you could process your data for them, i.e., move their processing onto your systems. Data brokers do this for ad agencies. There are other dastardly ways around this restriction, all of which could cause you more reputational harm than economic good. We call this crossing the “creepy line”. See the brilliant digital ethics research by my colleague Frank Buytendijk. Other Gartner analysts cover data privacy and security.
Q13. Absolutely agree with this presentation. As for the CDO role – knowing the position will need to speak to the business benefits for budget and buy-in, are you seeing the roles being filled by business leaders from operations or from the true data science SME vertical? I can see each source of CDO candidates having strengths and weaknesses.
A13. Thanks! Yes, CDOs with a business background tend to form better relationships with business people and focus more on the process-oriented innovation potential of information. CDOs with an IT background tend to be stronger in data management, integration and architecture. And CDOs with an analytics background of course are stronger at improving human and machine decision-making throughout the organization. Our research shows that most CDOs hail from a business function.
Q14. What’s so special about Lake Grapevine in Texas?
A14. It’s the site of Gartner’s previous and upcoming Data & Analytics Summit, 6-9 March, 2017. And also some beautiful memories.
Q15. CDO stand for chief digital officer or chief data officer.
A15. Both. And also: Collateralized Debt Obligation, Command Duty Officer, Community Development Organization(er)/Office, Central Dispatching Organization, Communications Duty Officer, Compagnia Delle Opere, Central Documentation office, Civil Defense Office, Carbon Doped silicon Oxide, Child Development Officer, Central Disbursing Officer, Corneodermatoosseous Syndrome, Customer Direct Order, Change Design/Data Order, Commercial Development Office, Chief Disciplinary Officer, Construction Development/Design Officer(er), Constrained Distortional Optimal, Classified Documents Officer, Control Document Officer, Callable Debt Obligation, Career Development Opportunities, Coaxial Digital Output, (and as I’m demonstrating here) Compulsive Disorder of Obsessiveness.
Follow me on Twitter @Doug_Laney #infonomics #GartnerEIM #GartnerBI
Category: big-data cdo infonomics
Tags: accounting big-data cdo chief-data-officer data data-capital data-management data-monetization data-privacy data-valuation data-value infonomics information information-assets information-capital information-governance information-innovation information-management information-value monetization privacy valuation
Comments or opinions expressed on this blog are those of the individual contributors only, and do not necessarily represent the views of Gartner, Inc. or its management. Readers may copy and redistribute blog postings on other blogs, or otherwise for private, non-commercial or journalistic purposes, with attribution to Gartner. This content may not be used for any other purposes in any other formats or media. The content on this blog is provided on an "as-is" basis. Gartner shall not be liable for any damages whatsoever arising out of the content or use of this blog.