Promising Practices Promising PracticesPromising Practices
A forum for government's best ideas and most innovative leaders.

How to Tame the Software Beast

ARCHIVES
Flickr user nerissa's ring

Managing software across any large organization is daunting. Unused licenses skulk unnoticed in corners, even as more are bought and added to the ranks.

Divisions and offices purchase the same applications directly from the same vendors unbeknownst to one another—and pay vastly different prices. Often, an organization doesn’t even know how many copies of any given software it owns or even where all the licenses are.

That’s why Office of Federal Procurement Policy Administrator Anne Rung and U.S. Chief Information Officer Tony Scott, along with the authors of the 2014 Federal Information Technology Acquisition Reform Act, are to be congratulated for attempting to slay the software monster not only at the organizational level, but governmentwide.

As Rung has observed, the federal government spends more than $9 billion a year on more than 50,000 transactions to buy commercial software licenses, subscriptions and basic maintenance. The average transaction is less than $200,000, so there’s a lot of room for consolidation and rationalization. Properly managing software spending promises enormous benefits and savings.

For example, a single agency, the Homeland Security Department, reported cost avoidance of $376 million from fiscal 2013 to 2015 by negotiating enterprisewide contracts with software vendors. A single DHS-wide deal with Adobe saved $125 million between March 2010 and the end of December 2012.

The draft software policy Rung and Scott issued on Dec. 21, 2015 would go a long way toward helping the 24 major federal departments tame their software spending and marshal their data to smartly manage such purchases internally and governmentwide.

The policy is part of the Office of Management and Budget’s category management initiative, which is aimed at saving money, releasing resources for mission-critical work, reducing contract duplication and rationalizing the purchase of $270 billion in common goods and services each year across agencies.

The United Kingdom began its category management journey in 2005 as a result of a 2004 public sector efficiency review. Its emphasis on the advantages of “joined-up government” is similar to Rung’s call for government to buy as one. Given the similarities in policy and intent, we wanted to share some U.K. experiences with software category management that OMB and agencies might find useful in getting faster and better results.

Much of what we learned through managing the information technology category and the software subcategory in the U.K. Government Procurement Service dovetails with OMB’s push for adoption of agile practices in information technology and procurement.

For instance, we found that taking an iterative approach to agencies’ software inventories helped win their cooperation and enabled central processing of the information and sharing of analysis. It also made it easier for software companies, systems integrators and others in the supply chain to comply with agencies’ requests for help in collecting the data.

Ultimately, we learned to stagger the delivery of inventories and break up the work leading up to them.

We broke up the management of software suppliers by size. Like most large organizations, U.K. government agencies spend the most money with just a handful of software vendors, such as Oracle, IBM, Microsoft and SAP, whose licensing models vary. So we dealt individually with each of these big vendors.

We learned to phase in software management in order of priority by other characteristics as well, rather than directing that it occur governmentwide and companywide all at once. OMB also might consider successively addressing sets of vendors based on shared characteristics. For example:

  • Those with which government spends the most.
  • Those with which agencies have the most significant relationship and governance problems.
  • The suppliers that impose the most management overhead on government through continuous or regular audits, complex licensing models, high rates of protest and litigation, and the like.
  • Vendors whose software would deliver the biggest rewards if it were effectively managed.
  • Those whose software models are most vulnerable to fraud, waste, abuse and mismanagement, or most in need of transformation, according to the Government Accountability Office.
  • Those whose products could more easily be replaced with open source software.
  • Not only did we segment software management into vendor-specific phases, but we launched those phases at specific points during the annual business cycle to achieve maximum procurement impact and gain negotiating leverage with each company or group of companies.

For example, by having software inventory data available about three months before most support contracts came up for renewal, we enabled agencies to cut their costs by canceling unneeded licenses and support agreements.

Vendors feel the most pressure to sell at the end of their fiscal years, so that’s when they are most flexible on terms and pricing. We found that having inventory data available before then can help agencies redeploy licenses identified as unused, thus driving down demand for new ones and taking advantage of vendors’ willingness to negotiate price and terms.

Managing software in prioritized, vendor-based phases also staggered the delivery of inventory data, making it easier for agencies and suppliers to provide and for us to analyze. The phased approach also provided room for iterative collaboration with and among agencies on one stage,  enabling us to fine-tune the next one. This improved our policy and delivery, while building agency engagement.

Then there is the matter of the inventory data itself. We blanched at the prospect of agencies, bureaus and vendors annually reporting software inventories in their own formats, covering different time periods and with different naming conventions. So we focused on setting minimum governmentwide software data standards.

We found it helpful to adopt some standards commonly used by the software industry, but then added fields to them to help us identify trends we could use to rationalize software utilization, manage demand and aggregate spending to drive better pricing and terms.

Requiring reporting by department, bureau, systems integrator and other organizational fields let us create powerful hierarchical reports that drilled from governmentwide software spending down into details by organization, product and vendor.

Data standards also made dealing with government easier for the software companies. Once, a systems integrator serving multiple agencies might have had to inventory and report software differently for each of them. With common data, including standard descriptions of each type of software, integrators could centralize reporting across all their agency accounts, making significant efficiency improvements that could be passed along to the government.

These are just some of the lessons we learned in battling the software dragon. We offer in them in the spirit of cooperation and support, hoping to spare U.S. agencies and OMB some pain and time in their struggle. We have seen the enormous payoffs of even partial victory, so we are invested in seeing that America’s success in IT category management exceeds our own.

David Shields is ASI Government’s Director of Procurement Transformation and Category Management and former Managing Director of the U.K. Government Procurement Service. Tony Crawley is managing director and co-founder of independent software licensing consulting firm Synyega, and former advisor to the UK Crown Commercial Service on software license reform.

Photo: Flickr user nerissa's ring

FROM OUR SPONSORS
JOIN THE DISCUSSION
Close [ x ] More from GovExec