San Francisco -- This story will update throughout the day.
8:30 PDT: Confrerence convenes, opening keynote from conference chair Aaron Zornes
Compliance, single customer iew (Know your Customer/supplier, etc.) are "universal drivers." Customers have stopped calling for multiple domain MDM and now assume it is baked in. Large enterprises have moved past CDI and PIM to address reference master data. (Large vendors are capitulating to package/bundle RDM).
"Party" MDM providers under pressure to manage "thing" (supplier location, etc. and vice versa). Pim vendors are challenged to be good at party data.
Also, "big data" seems incresingly a part of the discussion/landscape of master data.
Zornes says application providers will increasingly package MDM "innately" in application software across a variety of categories. Low cost departmental MDM will also begin to arrive from select large vendors.
Master data governance market is nascent but emergine, stand by as IBM, SAP, Oracle, Informatica deliver tools/applications, choices will have to be made.
9:15: Zornes continues "Social MDM" is a challenging overheated term that is hard to discuss. But, prototype systems exist that cull D&B info, Facebook etc. Near term, cloud based services will be more prevalent than commercial products.
On IT selling MDM internally: "Promote MDM as an essential business strategy with deliverables used repeatedly." "Position MDM as enabler of key business activities"
Bottom Line: Ivest for long term sustainability.
9:30: Rob Harris, Customer Data Strategist at Dell talking about data management transformation.
Identified core capabilities
Identify the customer data; Relate the customer data; Govern the customer data -- Build customer master data management into the fabric of business, not just a project.
Created a customer reference architecture based on capabilities, added to roadmap and ensured channels of communication with stakeholders.
Customer data governance was stood up in advance of any customer data MDM programs. "Isn't this just more work" a typical reply, so targeted business benefits were identified to sell the program. Direct benefits of data governance must be continuously and clearly conveyed. The concept of centralized data governance is still maturing. "It's going to be a long road," so have a targeted approach.
10:00: Alex Bentley, principal product mgr. at IBM:
Look at data governance across three lanes:
- Policy administration: how are policies created, approved, distributed
- Monitoring: How do we know policies are adopted, enforced
- Enforcement: Model agile remediation, collaborative remediation of policy violations.
On social media, big data and MDM: "When we think of big data and MDM, the two will have to work together. Take the information you have internally, seed the processes that look at social media streams and use that to extract the information. It is better because you already know what you are looking for."
11:45 Keynote: Kalyan Vishwanathan, global practice head, Tata Consulting Services on "Enterprise Information Fusion."
"The customer is much more than a buyer, he is an innovaor and influencer with many nonlnear relationships." There is a new network of relationships beyond attributes like name, address and other things we are used to. This has new meaning in the age of social media and it impacts all elements of the enterpeise from customer acquisition to retention to brand.
A new challenge is to understand 50 billion web pages. Response is fragmented, some dealing with social, others mobile, different parts of the enterprise interacting with partner portals. Need is to move to unified vision, "from big data to big insights" to quote IBM.
Big data stands to recast business intelligence, MDM and data warehousing. New capabilitries in social intelligence, process of generating knowledge from noise, techologies need a variety of capabilities: to listen, to filter, to fuse info sources structured and not.
This will create a quantum shift toward analytics across devices and unified customer engagement, new kid of connected marketing, one to one marketing over the Web.
Need to store, merge, analyze structured and unstructured data, connect at levels of predicted reliability. Tata calls Enterprise Information Fusion. Includes machine learning, not just listen & learn, but reason. Executing in Web scale is possible in an infrastructure cloud connected to enterprise data including master data systems. "Master data is absolutely essential to playing this new game of big data analytics."
Rajan Chandras, Delivery Lead, Data Strategy and Management at Horizon Blue Cross Blue Shield of New Jersey
The vision at Horizon is for centralized master data.
To avoid situations like Rajan has heard "We've been doing fine without MDM all these years, why do we need it now?" Rajan suggests to look at existing business problems and determine how MDM can help. The more visible the technology is to the business users, the easier the business case. The lower in the technology stack, the more difficult.
One of the things Rajan believes Horizon has done well is the collaboration betwen MDM and data governance. "There's so much more to data governance than just MDM, though MDM is a critical piece."
Also, Rajan highly recommends doing a pilot. It required time and money, but it was "one of the smart things we did."
Other success factors: staff for success. "If you don't have the right people, you're hosed." If you put good, passionate people in a room, "there will be a lot of noise" but the result will be good. Also, never leave the soap box, and provide clear vision for the future.
3:30 Laura Duvall Department of Homeland Security on data governance. On jumpstarting data governance across multiple unconnected agencies. Her message is about organization and simplicity.
Small collaborative group met weekly, developed mission statement. Reduced strategic plan for data governance from 250 pages to 30.
Strategic goals: governanence policies, collaborative working environment, insert data decision points.
Products included business lexicon, enterprise conceptual data model.
Outcomes included better trust, colllaboration.
Takeaways: Focus on top mission priorities, ensure opportunities are not wasted, educate, inspire by sharing pitalls and triumphs.
4:30 Best Practices in Data Governance Panel
Eugenia Rutenberg, Actelion Pharmaceuticals; Kira Chuchom, Data Mgmt org at Microsoft; Matthrew P. Hoying, Data Governance Manager at Stanford University.
From concept to action:
Eugenia: "It's always driven by need ... like sales side getting bad data. Tech side knew solution was tightening processes and policies, called it data governance later" but it was always tied to a real need.
Kira: There is a tendency to lead with tools and not governance but that is not right.
Matthew: Act locally, learn the language, do favors, avoid enemies.
Eugenia: Agree before implementing
Matthew: Scorecard and agree on metrics
4:30 pm. Dan Myers, Enterprise Data Management Manager, Farmers Insurance
Dan's case study presentation focused on the pre-MDM phase, and the crowd revealed that the great majority in attendance has not yet implemented MDM.
Although typical MDM hurdles may be similar in many companies, you need to prirotize and decide what needs to be addressed. Dan calls this a focus funnel.
At Farmers, the focus funnel, and the mechanisms to overcome the obstacles, were as follows:
1) Understand source data through metadata management.
2) Find consensus on business terminology through development of a tiered dictionary and stewardship.
3) Appropriate sourcing through systems portfolio documentation.
4) Data completeness and accuracy through ABC (audit balance and control) and data quality monitoring/profiling.
"Data quality is really a usability issue." Dan's goal is to expose data quality issues now, before MDM.