A decision-making framework for public procurement

Inmy lastpost,I talked about the seismic shift we’ve seen in recent years, away from theblack-and-white divide between open-source software (OSS) and proprietarytechnology. Today, you operate between those poles in a mixed-source world. Youhave freedom to choose best-of-breed solutions from a blend of sources and assemble the most effective mix of componentsfor the job.

As always,freedom brings responsibility as well as choice. Now that you’re free to choosewhat you want in the mixed-source marketplace, how do you make the bestprocurement decisions and invest public funds as responsibly as possible?

Start bydefining your exact needs for a project—beforeyou begin exploring solutions. When I talk to government decision-makers andproject managers, I make sure that everyone involved in a procurement project clearlyunderstands:

  • Thescope of the project
  • Itsgoals and specifications
  • Thepublic sector problem being solved
  • Thegovernment vision that needs to be advanced

This clarity laysthe foundation for evaluating your options across a range of OSS solutions, proprietarytechnologies, or both. The mosteffective tool I know for making good choices is a decision-making frameworkbased on three pillars: tech neutrality, fitness for purpose, and total cost ofownership.

1. Technology neutrality

Tech neutrality means basingprocurement decisions on impartial criteria that don’t favor any business orlicensing model, development model, approach to intellectual property, or evena specific solution. With a tech-neutralattitude, you can fairly assess the merits of all possible solutions and choose the best combination of softwareand technologies for a given situation. Your decision will be based on vital criteria:cost (more on this below), interoperability, reliability, vendor support, easeof use, and security.

The pendulumlately has swung back from a bias toward OSS as an inexpensive cure-all. Many public-sectororganizations are taking a more balanced approach. On thepolicy front, I’m seeing governments either repeal their former OSS-only stancein favor of tech-neutral policies or add tech-neutral thinking into the mix. Supportfor this trend is strong in the UN, the EU, and APEC, and in emerging as wellas developed nations. On the procurement front, many governments that havetried OSS implementations have discovered flaws, difficulties, and higher costs,and have embraced flexibility over preferential procurement policies.

A comprehensiveMicrosoft catalogue,NeutralGovernment Procurement Policies, summarizes the experiences of some 30 governments andother organizations that have recently adopted tech-neutral policies, and pointsto the substantial benefits they bring to consumers, governments, andlocal economies.

2. Fitness for purpose

Any tech solution mustgenuinely address the requirements specified in the use case analysis of a given project. Simply put: Does thesoftware do what you need it to do?

Consider the recent experience of Freiburg, Germany. From2007 to 2009, the city began migrating to OpenOffice with the goal of making itthe standard across agencies. They hoped usage would spread nationwide. But officialrecords revealed user complaints including document conversion problems,instability, and crashes, and noted that workers still relied on legacyversions of Microsoft Office to perform critical tasks. The OSS solution notonly failed to gain traction but also resulted in “reduced performance, anger,and frustration among employees and affected external parties” according to theofficial report. The city ultimately abandonedthe effort and returned to Office. The German Foreign Office, it’s worthnoting, also switched back to proprietary operating systems and office apps in2011, after a nine-year trial of OSS alternatives on its 11,000 desktops.

3. Total cost of ownership

Licensing costs typically account for about 10 percent ofsoftware expenses. If you want to find real savings, look at the other 90percent—the costsof migration, training, maintenance, security, and achieving interoperability.A clear-eyed accounting for thetotal cost of ownership over the lifetime of a product can reveal that lowupfront costs don’t necessarily save money in the long run.

Here’s an example. In 2011, Helsinki, Finland, evaluated acost-saving proposal to install OpenOffice on some 21,000 workstations in the cityadministration. The resulting studycompared all the costsassociated with migration and projected that OpenOffice would ultimatelycost 74 percent more than Microsoft Office over a seven-year period.

Similarly, the Dutch Court of Audit in the Netherlands statedin its 2011 “Openstandards and open source software in central government” report: “We concluded amongst other things that thepotential savings the government could make by making more use of open sourcesoftware were limited.”

You don’thave to take my word that these procurement guidelines really work. Just lookat the experience of the many governments that are implementing them now andenjoying the benefits.

Havea comment or opinion on this post? Let me know @Microsoft_Gov. Or e-mail us at ongovernment@microsoft.com.


Scott F. Selby
Director, Global Government Affairs, Microsoft Legal and Corporate Affairs