By Dennis D. McDonald

In Six Data Management Steps To Consider Before An Administration Shift William Sullivan of Cloudera lists the following steps government agency executives need to consider before the new Administration arrives:

1. Take a close look at the data you have available. Is it the right type of data that allows you to see trends and stumbling blocks to your agency’s mission? And are you using all of the data available to you, both structured and unstructured?
2. Can you easily and securely access it, share it and perform the types of analysis that enable you to put the data to work towards solutions?
3. Do you have the right tools in place, or do you need to identify and implement the best tools to leverage the data you have through complex analysis?
4. Are your systems, and in particular your legacy systems, modernized to permit them to work efficiently and cost-effectively for your agency? Or do you need to procure modernization tools to get the information you need from your mainframes?
5. Do you have the right team in place, including data scientists and other subject matter experts who can assess your systems to get the most from your data?
6. Can you answer the questions a new administration is certain to ask, validating with data the challenges and solutions you will be recommending to new agency leaders?

The above is an excellent list that any organization should be considering, but several important elements, missing from this particular list, are also important to consider:

1. How do you ask these questions?
2. How do you formulate an implementation strategy?
3. How do you manage how you decide which actions to take?
4. How do you manage implementation steps?

Making such distinctions is a major reason why it's important to distinguish between data management and data program governance, as I discussed in Thinking About “Data Program Governance.”

An effective data program governance strategy must address the data-based services you intend to provide to your users who, in all likelihood, will possess a wide range of data analytics expertise. Some may be able to access a self service portal and crunch their own numbers. Others may need much more handholding and support. Some requests will be one-time while others require recurring reports. Some may rely on existing data while others will need access to externally supplied or unstructured data.

Such considerations are normal to any type of service organization whether it be an internal network support organization, a call center, a PMO, an Agile-based development group, or a help desk. The types of questions listed by Sullivan reflect the need to treat "data" as a resource to be managed along with the standard trio of people, processes, and technologies. 

The Sullivan list provides a good foundation for a strategic assessment that will form the basis for constructing a comprehensive data program governance strategy, with one proviso: don't let the strategy formulation process take on a life of its own. Begin delivering useful analysis and data-based services as quickly as possible in parallel with assessment and strategy formulation. 

How all this is managed with everything else going on in the organization (management changes, movement to the cloud, legacy system replacement, demands for new services, etc.) will require much stakeholder support, communication, and collaboration and should not be treated as an "IT only" responsibility.

Copyright (c) 2016 by Dennis D. McDonald