Big Data Summit - Executive Board Meeting
1 like526 views
During the big data executive board conference call, attendees discussed the need for a balanced focus on both data governance and analytics, pointing out that previous discussions were overly analytics-centric. Attendees emphasized that effective data management should encompass cultural accountability across the organization rather than solely relying on IT. Furthermore, the conversation highlighted staffing challenges in analytics versus data management, stressing the importance of sharing real-world case studies to enhance understanding and adoption of analytics practices.
Big Data Summit - Executive Board Meeting
- 1. www.bigdatasummit.us BIG DATA EXECUTIVE BOARD Conference Call Minutes | 02.04.2015
- 2. • Holly Starling, Autotrader • Scott Hallworth, Capital One • Michelle Norworth, Capital One • Joe DeCosmo, Enova • Dirk Garner, Macy’s • Donnie Yancey, MapQuest • Anand Raman, Impetus • Randy Lea, Teradata • Jason Cenamor, CDM Media • James Quin, CDM Media The overwhelming impression of those initially in attendance was that the previous summit had been too heavily focused on the analytics side of the data issue with insufficient focus on the governance side of things. Analytics is primarily driven by the business and if there is no business stakeholder, then there seems to be a lack of interest. Data governance on the other hand is very much an IT issue and if good governance foundations can be put in place (i.e. data dictionaries, semantic layers) that codify data then consumption becomes easier and business interest is more likely to be stimulated. This lead to the following discussion points: Governance vs. Enablement vs. Management Overall there seems to be little agreement amongst the group as to what is the best terminology around the “not analytics” portion of the data conversation. This isn’t just a semantics issue, but one of broader acceptance and understanding. Governance has a heavy regulatory feel that makes things seem unapproachable. “Enablement” achieves the same effect as “governance” but is shown to have greater business acceptability and an approach to making data more accessible rather than more regulated stimulated user community adoption of data as well as higher order work (primary usage vs. secondary usage). Extending this is the idea that “data governance” is simply an aspect of “data management” and that this broader view needs to be taken in regards to data, data ownership, data stewardship, and data usage. One aspect of data management that hasn’t really been touched on yet is that data management includes physical data management (storage levels, online/offline, etc.), which definitely puts the conversation and responsibility squarely back in IT, but for the benefit of the company rather than as a “land grab”. Culture If we accept that data management, in it’s all encompassing form, is the other side of data, then we next must consider the issue of accountability vs. responsibility. IT may well have the responsibility for data management, but it cannot also have the accountability, as a comprehensive data management aproach must be cultural, must be all encompassing. In order for everyone to be onboard, everyone must be involved which ultimately means that accountability lies at the top of the organization with the business leaders. Initiating the requisite cultural change can be exceedingly difficult however and is likely not something IT can initiate, but can co-ordinate when the external impetus presents itself. A lot of this depends on organizational maturity, but in most cases a “burning platform” is required that triggers the phase shift. Once that tipping point is achieved however, IT can and should begin the data management transition process. In the end, it is important to remember that data management is simply a means to an end and not the end itself. Big Data Executive Board Conference Call Minutes Agenda Item One: In Attendance: Regrets: • Ameet Shetty, SunTrust DATA GOVERNANCE / DATA MANAGEMENT DATA ANALYTICS Agenda Item Two: After discussing the issue of governance/management at significant length, we turned to the issue of analytics. By this point our group had rounded out with a few individuals that had more of an analytics focus within their organization. Their position was that while analytics had been discussed at the previous event, perhaps more so than governance, it was generally at a superficial level without a lot of deep dive. If we wanted to discuss analytics then, we should focus on the following:
- 3. Agenda Item Three: STAFFING Our conversations around data management and analytics consumed much of the time in the call but we did man- age to squeeze in a few points around staffing issues. On the whole the staffing conversation was split along the same lines as the rest of the conversation – data management staffing vs. analytics staffing. Data management staffing is essentially seen as a non-issue because other than top-line responsibility in the form of a CDO or equivalent, existing staff executes much of the line work. On the analytics side the situation is very different; staffing continues to be a significant challenge with far too few qualified resources around to meet the demand. Though conversations were had at the last summit on this issue, it was generally felt that continued focus in this area, with different individuals sharing their experiences, would be beneficial to all. CDM Media Contact Information: Jason Cenamor Director, Technology Summits - North America +1 312.374.0831 jason.cenamor@cdmmedia.com James Quin Senior Director, Content and C-Suite Communities +1 312.374.0809 james.quin@cdmmedia.com What is Analytics? Analytics most definitely is not tools, which seemed to be the focus of the previous event. It is important to focus on real-world case studies where data has been turned into information from which insights have been extracted that have stimulated change and/or growth. Analytics is the transformation of data to produce new data and it is important to determine how that is done responsibly. More data does not necessarily equal better analysis – better analytical processes do that. We collectively need to focus on the “how” of analytics, and almost take a myth busters type approach, exposing what doesn’t work (and showing why) versus what does. Everyone is still pretty early phase in this and so sharing experiences is the best way for everyone to scale their maturity. Subfields of Analytics Analytics is as broad a topic as governance/management. There are layers of analytical complexity and techniques and investigation into all of them is needed. Specifically focus needs to be given to Data Discovery, to Business Intelligence, to Graph Analytics, and to Machine Learning. Business Intelligence offers ease of “consumability” and the opportunity for business users to manage much of the analytics task work themselves, but breaks down at scale. Graph Analytics has tremendous potential value but broader/deeper understanding of applicability and use is needed and much the same can be said for Machine Learning. Data Discovery is seen as a more flexible, lightweight, unstructured approach with the concern being that “unstructured” eliminates necessary rigor that can make results trustworthy and viable. As an initial “gateway” step though it may make sense. Agenda Item Four: OTHER ITEMS In closing we looked at the opportunity to expand the reach beyond just IT roles. It was definitely felt that the Summit should remain open (or become even more so) rather than hyper-focusing down to just CDO/CAO roles. CIOs absolutely must continue to be part of the audience, but expanding to line of business leaders, to bring in the user vision/experience of the data issue, would be valuable. To that end it would be beneficial to expand the delegation to also include CMO attendees. The Big Data Executive Board is a panel of industry experts and leading vendor organizations that meet quarterly to help drive the agenda of CDM Media’s twice annual Big Data Summit. The Summit is attended by a carefully selected group of industry practitioners intent on creating broader and deeper understanding of data issues for the user community at large.