The day began with "Culture eats strategy for breakfast" … a quote attributed to social ecologist Peter Drucker. Our speaker went on to assert people impact business initiatives far more than business strategies or technology. Think about your recent projects. Think about the strategies, the technologies and the people. Think about what caused the success or failure of each project. I know I have seen sound strategies using solid technologies succumb to a collection of people who refuse to work together for the greater good. You need to initiate a formal alliance based on delivering instead of credit. Organizational silos exist, but cooperation across silos can accomplish more than diverting resources toward simply defending those silos. People, not technologies, decide whether to cooperate or defend.
Did you know a Ponemon Institute's study states: 1) 84% of respondents use real customer information for development or testing purposes; 2) 51% of respondents use payment information (e.g., active credit card numbers) for development or testing purposes; 3) 38% of respondents had a security breach involving real data in a development or testing environment. Why block access in production, grant test access to everyone and then copy production data to test? We should treat test data as the valuable asset it is. Of course data protection versus data usability is always a business challenge. On your next project, consider test data as a triangle of Fast, Realistic and Private … where you can only choose 2 of the 3 options.
Did you ever consider a non-disclosure agreement (NDA) does nothing to secure your data? The NDA only facilitates legal action once your data was lost.
Quality data has a shelf life. To maintain business value, the data must be regularly refreshed and replaced. If you combine data from multiple sources to derive your Single Best Record then you need to develop a survivor record strategy to keep your quality data up to date.
How do you ensure your quality data remains trusted when being presented? Consider watermarking reports and dashboards which present only trusted data. Watermarking sounds simple at first, but the watermark should be placed only on those objects which are fully vetted with proper metadata including complete data lineage. Once business users see the value of the watermark, they can focus their efforts on making sound business decisions instead of just arguing over the accuracy of different data. Focus your watermarking efforts on business critical objects and avoid data deliveries that allow users to alter the watermarked data (e.g., Excel).
Over both days, the sessions and speakers provided ample first-hand data quality experiences. Those experiences focused on bottom-line business impacts. If you consider reducing costs as the primary impact on your bottom line then you are missing the huge potential for additional revenue brought on my improved data quality.
I wonder what knowledge might be shared next year.