CDO Perspectives from MITCDOIQ
Last week I had the privilege of attending the MITCDOIQ symposium in Boston. This annual event features a number of distinguished data professionals from a wide variety of industries and public sectors sharing experiences of “exploiting data capital for organization performance.” With the CDO role being on the rise across the board in America, where 90% of large organizations are expected to have this role in place by 2019, the excitement and inspiration ran strong in the gathering. The session topics ranged from machine learning, advanced analytics and data monetization, to delving into the many facets of CDO role, its office and programs, such as planning for European General Data Protection Regulation (GDPR).
As I listened to the speakers and talked with attendees, I noticed a few key themes emerge: protecting information assets, creating data standards and metrics, and tackling both organizational and technical challenges that come with the role, such as overcoming silos of information within the organization. The machine learning and advanced analytics session was a staple of the first day. It focused on innovative uses of modern technology and algorithms. The new CDO of Sanofi shared a roadmap for achieving an eventual goal of being able to monetize information at “industrial,” enterprise-wide scale.
He outlined challenges including managing big data required to support genomics and infusing the passion for creating exquisite user experience for data driven interfaces. For instance, the CDO office team members are expected to regularly “cross-pollinate” with colleagues from other department to understand “our people’s user experience with data.” I found some of the strategies he outlined to be really different – and equally as intriguing. One of which was to hire someone from Louis Vuitton in a bid for their user interface for data to look sexier. Another interesting machine learning utilization tip was a suggestion to use Alexa voice recognition to communicate the data insights, which would complement the traditional visual dashboard approach to data insight communication. In addition to using Alexa, he mentioned a number of other Amazon features, most notably the story of developing an Amazon recommendations engine which nowadays accounts for 35% of the e-commerce giant’s revenue. This story echoes the bright future of our own introduction of the data “amazonification” concept within Collibra platform. Our machine learning modules currently offer data set recommendations and suggestions for linking data sets with the related business terms, which facilitates constant evolution of the data lineage in an ever-changing organizational information landscape.
Another MITCDOIQ speaker, while arguing the case for prescriptive and descriptive analytics becoming mainstream, mentioned using tags on data for prescriptive machine learning. This use case rang similar to the crowdsourced tagging of metadata that we rolled out this summer with Collibra 5.1. The metadata is arguably scarcer in volume that the data itself, which makes the proliferation of tags around it especially useful for deep learning on metadata.
Another recurring theme was human education and workforce adaptation to the digital age we live in. With potentially varied maturity levels on individual, departmental, and corporate levels, the enablement of various personas within the organization to function in a data-driven manner is critical. In our area of digitalization (one of the latest trends from Gartner) and the cultural shift within major companies and industries, it is important that we are able to collectively keep up and stay on the cusp. One of the symposium speakers was our co-founder and VP of Education & Research Pieter De Leenheer who curates our very own Collibra University. During his joint session with Dell’s Chief Data Governance Officer Barbara Latulippe, Pieter shared how the transition from scarcity to abundance affects data governance and stewardship in Big Data powered enterprises. Barbara, who won the symposium’s award for transformation from traditional data to big data, then shared the journey of empowering business teams and creating organizational information governance. She started off by stating the #1 goal for her organization: to help their users find their data faster. Barbara also shared some advice with the audience, which is worth repeating: start somewhere, move forward, and be flexible. The reality is that things will change and you need to be able to change with them. Achieving formal federated governance across the vast Dell-EMC enterprise required significant organizational, logistical, and cultural changes. The efforts to ensure their data is prepared for its business use, or “fit-for-purpose” as some say, were rewarded by attaining full cycle of sharing valuable information in a regimented fashion among business users, team members and customers alike.
The principle of valuing data based on the level of consumption enabled the CDO office at Dell to achieve agility in their data governance implementation that involved cataloging the full inventory of their information assets while implementing a minimum set of business-specific metadata attributes . Dell’s goal is to cut the data valuation cycle time by 40% – and they are well on their way to achieving this goal by implementing data governance and stewardship best practices enabled by the Collibra platform.
On the last day of the event, the presentations on advancing data-driven culture in US government and public sectors dominated the main stage. Some thought-provoking questions were asked, such as what would it take to extract the revenue from data science work and sustain its productivity on both federal and state level? The recurring theme there was system integration, data provisioning, and APIs needed to create infrastructure than can play well with the multitude of various agencies. This theme is near and dear to our hearts at Collibra, as we endeavor to support much talked about then vendor-neutral approach to potentially extend our platform’s reach to the hundreds of various data systems and tools, which is a common requirement among many of our enterprise customers. The government agencies on federal and state level have similar need for integration and API standards essential to sharing data is a safe and efficient way to create collaboration-friendly data ecosystem and replicate its best practices across our vast public system.
A number of renown MITCDOIQ symposium speakers acknowledged key role of data governance in enablement of various personas such as business users, data scientists, or analysts to be able to find the right data required for them to be successful in their respective roles. In a nutshell, our team left the event energized by this impressive group of CDOs who were, for the most part, just like ourselves, convinced that data governance is about making the data-driven world a better place.
Ellie is a passionate product manager, data warehousing aficionado, self-proclaimed fitness enthusiast, up-and-coming entrepreneur and occasional comedian. Prior to spending last ten years as multifunctional product leader, championing data management products from inception into successful adoption with high profile enterprise customers, she learned the art of building software as a consultant, programmer and system architect. Ellie holds a Marketing MBA from New York University’s Stern School of Business and bachelor’s degree in Computer Science, also from NYU.