Governance versus innovation—and why that might not be the right battle


Information governance, the orchestration of people, process and technology that enables an organization to leverage data as an enterprise asset, or innovation? It’s a question that presents itself with considerable frequency for Chief Data Officers. Lately, I’ve been considering this binary outlook as it relates to the “data integrator” role that CDOs strive to fulfill for their organizations, and I believe the question itself misses the mark.

Data integration is an important consideration for all CDOs, whether you are in your first 90 days or are a veteran of the profession. Instead of governance and innovation being a diametrically opposed “either-or” proposition, good data integration and governance facilitates and promotes innovation within the organization.

Starting with an awareness of your company’s monetization strategy and using that to create your data strategy is essential. It is also important to recognize that the level of integration and governance provided must be tailored to suit the needs of various stakeholders inside and outside of the organization. Some need the ability to explore raw data. Others wish to take analytical discovery into their own hands. Still others will benefit from packaged, value-added data. The desired level of integration and governance activities will vary considerably.

Therefore, the question becomes not, “which do I want—governance or innovation?” but rather “how much integration and governance is ideal for each group of stakeholders so that innovation may be achieved?” In the following sections, I aim to address that question more specifically for three groups of stakeholders.

Data science teams

Data science teams may balk the most at heavy handed governance. They want the opportunity to dig into raw data and build it into applications without a lot of interference. In addition, they prize solutions that are open and agile. As CDO, this necessitates an approach to integration and governance that is balanced. Seek opportunities to supply the right level of integration and governance without blocking their exploratory, agile nature by making sure that the tools they use allow for proper data management and preparation.

Consider the data repositories being used. Are they flexible? Do they provide uninterrupted data access? Are they fully managed? If not, there may be room to more fully embrace the open and agile approach the data science team desires. A number of fully managed database options are available including a NoSQL database service and a cloud hosted platform of open source database technologies.

Providing your teams with the ability to clean, reconcile, and match data in a single environment tailored for data science can also be beneficial. In this way, data preparation simply becomes part of the overall process—a step along the path to valuable insight.

Business users

Increasingly, business users are seeking to take analytics into their own hands and turn insights into innovative competitive advantage. But as we all know, insights are only as good as the data on which they are based. For this group, it is important to provide data that is analysis ready (integrated, governed and accessible), as well as tools that guide them toward insights in an intuitive manner.

The integration and governance put in place must reflect the needs of the business user. Timeliness is of the essence when making business decisions, so the data offered should be as close to a “real-time” representation as possible. Completeness and accuracy of the data is critical since insights are dependent upon its veracity. Finally, scalability and ease of data accessibility will be important considerations given the proliferation of data and the fact that business users will expect data to be readily available for immediate use.

Once that data is available, business users will need to be able to explore it and report their findings with intuitive tools that guide them through the process. Seek out solutions that allow users to employ natural language from the start to dig into the data and create visualizations of what they find. Also, make sure that the tools are available on web and mobile devices so that users can access them easily. As an added bonus, some self-service solutions offer built-in data governance practices and data quality scores for additional peace of mind.

http://www.seenews.info/wp-content/uploads/2016/08/governance_embed_1.jpgRegulatory agencies and external partners

By far, the highest level of data integration and governance is needed to create the pre-packaged and value added data that regulatory agencies and external partners wish to see. Pristine data is a necessity for these stakeholders. It is important to have sufficiently broad integration and governance capabilities working together to deliver data that is both compliant and easily monetized.

The ability to integrate and consolidate data from multiple sources in a flexible and timely way is important, as is the ability to create one trusted view of the data with a proven master data management solution. Likewise, solutions that improve data quality are critical to ensure the highest standards are met for partners and regulatory agencies. However, perhaps the most key capabilities to aid compliance and monetization are a strong data governance solution with data lineage capabilities and an information lifecycle governance solution that retains data according to company and regulatory policies, and can dispose of data that is no longer required.

With a broad set of solutions like the ones above, achieving the level of pristine data needed for these stakeholders becomes more possible. However, even the best solutions can produce a less than stellar result if they don’t work well with each other. With that in mind, look for solutions that will not only integrate data, but work together as part of a cohesive system too.

As evidenced above, when focusing on the user group instead of a governance vs. innovation battle, it becomes easier to determine how to approach the data integrator role and create value. Though it still takes considerable work and practice, this outlook does provide a more fruitful starting point. For more information on how to fulfill the data integrator role, I encourage you to visit the CDO Lookbook.

Visit the CDO Lookbook

Follow @IBMBigData

This entry was posted in Big Data. Bookmark the permalink.