Data Lakes enable a single source of truth but not necessarily a single version of truth says Doug Laney, Distinguished Analyst, Gartner
Today Doug Laney is a VP and Distinguished Analyst with Gartner's Data and Analytics Strategy Research and Advisory team. Mr. Laney covers data technology, strategy and is the author of the bestselling technology books of our times titled “Infonomics: How to Monetize, Manage, and Measure Information for Competitive Advantage” (www.gartner.com/infonomics) ’. Sangram Aglave, Contributing Editor – Technology caught up with him. Edited excerpts below:
1. Walter Benjamin in his seminal work ‘The Storyteller’ suggests that the value of information does not survive the moment in which it was new and therefore ephemeral. Was he wrong?
Doug: I get what Walter Benjamin is going after as he advocates for storytelling and rightly mentions that the narrative achieves an amplitude which information lacks. He speaks about the benefits of storytelling but also acknowledges that storytelling does not aim to convey the pure essence of the thing as does raw information. So, putting it all together we can say that storytelling is the way to amplify and distribute the utility, availability and awareness of some information, but it does not convey it in pure or necessarily accurate form. To be doctrinaire about it, quantum physics suggests that information can never be created or destroyed, but I don’t think this theory is practicable in a business context. Storytelling is a method to feed information to others and assuring the consumption and ongoing retention of information. However, increasingly the consumers of business information are not people but technology like computing systems and applications and algorithms. So, mnemonics like storytelling don’t appear so much relevant in cases where technology consumes data and information without downplaying the value that storytelling holds for humans. Certainly, storytelling will continue to have a place among humans, but is becoming devalued over time as machines become the primary consumers of. I think Information needs to be treated as any other goods or a service and in my book, I have suggested how to apply traditional economics models to Information like demand supply, diminishing marginal utility etc.,
2. Is the C level adequately biographical to tell the story of his brand, department or company through data? How does Gartner help the C level in conceiving data stories?
Doug: Many IT executives lack marketing skills. Storytelling is a premier marketing method and something which all executives could be better at. At Gartner, we curate data stories about how organizations are using analytics in innovative and high value adding ways. We share relevant stories from our repository with C level executives which help inspire business leaders to develop a vision for doing more with data and analytics.
3. By keeping data in-situ and bringing processing to data comprehensively address all information/data security challenges?
Doug: Moving computing to the data is a preferred or a default posture because of four things that can happen every time you move or copy the data. First, you end up increasing the attach surface with multiple copies. Second, you incur expense in maintain multiple copies. Third, you are exposed to data synchronization issues and the resultant risks. And fourth, you can end up making bad or inconsistent decisions when there are multiples sources of the truth. Indeed, there are legitimate reasons to move data, but only after a cost benefit and risk analysis. The default position is to not move data and to be economically circumspect when you do so.
4. Is Data Lake a necessary condition or a sufficient condition for a unified approach to Information Asset management?
Doug: If we think about data as an asset and how to inventory that asset, it rarely makes sense to have multiple ‘distribution centers’ of data. The data lake is the current fashionable architecture for co-locating multiple information sources into a single source of truth. But is not necessarily a single version of the truth, as context may differ from one business unit to the next.
5. Should CIO’s become true Information officers. Can that be a way forward bypassing the need for a CDO?
Doug: The Chief Data Officer role may not be for all companies. It is only for companies where the CIOs operate more as a chief ‘Infrastructure’ officer and are unable to give enough time to develop information as an asset. The Chief Data Officer allows organizations to hold a person ultimately responsible and accountable for realizing economic benefits generated through the wealth of information held by companies. It may seem odd to some to introduce a new executive role over information assets but consider how in the 1960s Milton Freidman and Gary Beckman came up with the idea of Human Capital even though humans cannot be a balance sheet asset. That gave rise to the modern Human Resources organization and the CHRO role. In the 60’s labor was a primary competitive differentiator, and now data is. Therefore, organizations are well-advised to create a senior executive role responsible for data assets. Some companies are even separating their IT department into separate ‘I’ and ‘T’ functions to enable managing information and technology as the distinct assets they are—and for moving the management and responsibility of Information closer to the business.
6. Sometimes organizations have information but still fail to perform. How does Gartner help such organizations?
Doug: Gartner informs its clients with information management and analytics maturity models, success stories, best practices, trends, advice on technologies, and coaching sessions. We also review our clients’ data and analytics strategies and programs to help them improve business outcomes.
