This guest post is written by Dr. Vinti Agarwal, candidate for the MS in Government Analytics.
The Government Analytics Breakfast forum, held on April 8, 2015, featured Kirsten Dalboe, Program Manager at the Department of Homeland Security. In her talk titled "Strategies and successes at DHS in persuading data owners to share data for analysis via the Management Cube," Ms. Dalboe made a strong pitch for the role of relationship building in developing the integrated data platform at DHS.
Developing the data platform was essential to deal with multiple queries by Congress and the White House, especially in the areas of human capital, contracting, asset quality and security. Business intelligence also needed to be increased in the areas of investments, workforce planning, budgeting, security, contracting and real estate on the basis of portfolio, function or mission rather than just organization-specific needs.
There were, however, multiple constraints operating at the same time. Unlike in the private sector, a profit motivation to share data could be not used. Further, within DHS there were a number of independent organizations that, although working under a common umbrella, had their individual systems, styles and platforms that they had been working with for various lengths of time. The complete overhaul of existing systems was infeasible due to limited time and financial resources.
To address all the concerns that stakeholders had with regard to the sharing of data, a set of relationship building initiatives, along with an executive structure, was implemented. The entire model was built on the three pillars of governance, quality and confidence. Some of these measures included:
- Creation of a three-tiered executive team headed by the Deputy Undersecretary for Management and the Management Cube Joint Program Team with due representation from all the business units handling the operational aspects. The representatives, who met with the program team every two weeks, were expected to give a dedicated involvement of at least 25% in terms of collaborative discussions and following up with data sharing tasks within their respective organizations.
- Signing of a Memorandum of Understanding by all lines of business for the sharing of data and agreement on certain data sharing principles followed by data interface control documents to ensure continued commitment.
- Harmonization and standardization of different business terms and the creation of a logical and conceptual data model so that everyone was on the same page as to what the terms meant.
- Continued use of the existing Oracle platforms, with which certain businesses had built business intelligence tools early on, to avoid any data transfer issues.
- Avoidance of the need to recreate data and to stick with what one had. A clear communication of the need to include summary and not transactional data from different business units was made.
- Emphasis on the need to have a data-sharing relationships based on team spirit, teamwork and non- competition.
- Creation of a simple, 6-question taxonomic structure to give a basic understanding of how data needed to be structured so that all relevant questions could be answered. Further, as long as the basic data architecture was in place, and the time stamping of the data was there, no demands were made to make the data conform to a particular reporting cycle.
- Emphasis on the principle of not making any judgement calls on account of the data being submitted. To address any caveats on the quality concerns of data being submitted by the business units, a 100 character communication of their subjective assessment of the data was permitted that would be published along with the respective data used in any report.
- Provision of an opportunity for review of the data by an analyst specialized in that line of business to address any concerns by various organizational units about whether the data was being used properly.
- Frequent and continued communication with stakeholders and participants.
The success of the exercise demonstrated that it is important that stakeholders' concerns are addressed not only to avoid resistance but to make them active partners. The vital ingredients to get the task accomplished at DHS was a combination of clear and continued communication at all levels, the justification of the importance of each task, due opportunity to understand and appreciate the respective stances and efforts of all stakeholders and, finally, the right degree push efforts. That is a model worth duplicating across most organizations involved in building data sharing platforms.