Q&A with Chief Data Officers: Steve Adler

Q1. What are the main challenges in data governance?

Evolution. The basic fundamentals of Data Governance have not changed since I invented it in 2004-5. We still organize Data Governance Councils (an idea I copied from the Federal Reserve Board of Governors as a way to bring business and data management together at the same table) benchmark organizational behavior with Maturity Models (the IBM Data Governance Council created the first Maturity Model in 2005), and think about our tasks as Six Sigma quality control operations, as ifevery human mistake that leads to a data quality, security, and operational risk exposure is actually preventable. They are not. Humans are prone to mistakes and that will never change. I regret immensely that I helped define Data Governance as a quality control discipline, which makes the role about nagging corporate hygiene.

Q2. How do you ensure data quality?

By studying the patterns of human error and discovering solutions to automate correction. Technology solutions that expect human cultures to adapt to accommodate rigorous programs of command and control may succeed temporarily but will always fail in the longer run. They will fail due to leadership and staff changes, and their successes will be Pyhrric until humans adapt to new rules and learn to escape them.

Q3. What are your suggestions for a successful master data management?

It has to become invisible. We should not be forcing people to adapt their understanding of terms to a common abstraction layer and instead should be building large relativistic thesauruses that map many different definitions to each other.

Q4. How do you create and implement a data strategy for your organization?

Data Strategy should support the brand identity and expand the business strategy. Data has a transactional value. No static value. To increase the value of data to the organization you have to increase the amount of it that can be used and reduce the latency of use. Every data strategy should seek creative ways, methods, tools, programs, and campaigns to increase data utility and reduce latency.

Q5. Data analytics and innovation management: what are the main issues?

We still use a library model of data source – stick lots of data into warehouses, marts, and lakes and expect people to know where it is and how to use it. We should instead use a Marketing Model, sharing with people based on their interests and behaviours without them even knowing or suspecting they need it, automatically linked to other data sources, with everyone learning from each others uses, analytics, and behavior.

Q6. How do you select the key Data tools and technologies?

I am constantly trying new things, working never to get stuck relying on the same solutions for too long. Restlessness and intellectual curiosity are key. Never go to the same conference twice and always seek out opportunities to try things you have never done before. Especially important to do things everyone else says you can’t.

Q7. How do you create and track of key business and data metrics?

I am extremely wary of measuring too many things.