Page 53 - FCW, August 2017

the cost is less than 5 percent of the technology currently used. That legacy technology was developed by a defense contractor in the 1950s.
Thirty years ago, I wrote a Kenne- dy School case about a tiny firm that developed and sold a technology to the Army that allowed young people who were thinking about enlisting to get a confirmed reservation for an Army training slot while talking with a recruiter, often months before gradu- ating from high school.
The previous practice had been to take enlistee training requests over the phone and write them on
a big wallboard that displayed
the number of requests for dif- ferent slots. There was no way to know which people had requested which slot, and that produced a situation in which many enlistees showed up for training only to discover that no slot was available for them.
I thought of that case study
when I noticed a post on Linked-
In by Chris Cairns, one of the co-founders of 18F. He recently left the government and was announcing the establishment of his new company.
The venue for the announcement was interesting in itself. It was a social media post rather than a more conven- tional way of doing such things. More significant, though, is what Cairns’ company is selling: He calls his ser- vice “microconsulting.”
“Not every problem requires a full- blown consulting engagement,” the announcement states. “Oftentimes a short burst of work on a specific topic is all that’s needed to help pro- pel government managers and teams forward. That’s why we’re introduc- ing digital consulting services priced within the federal micro-purchase thresholds.”
Purchases under that threshold — currently $3,500 for civilian agencies
and $5,000 for the Defense Depart- ment — do not require submission and evaluation of proposals and may be paid for using a government credit card. Such projects would correspond to about one to three days of work.
Obviously, not all consulting gigs can be so quick. But Cairns’ company seems to be trying to locate those that are. Under “agile solicitation support,” the firm offers to have “our team of technical and acquisition experts thoroughly review and recommend improvements to your solicitation package.”
and SpaceX. Government IT contrac- tors have a tendency to favor com- plex and costly enterprise software solutions rather than the cheaper, more standardized approaches favored by commercial, consumer- oriented providers. Similarly, federal contractors, facing a less competitive landscape, might be likely to provide drawn-out consulting advice from a bevy of staffers rather than find ways to provide such services at a lower level of effort.
In such a world, the service Cairns’ company plans to provide could be disruptive, just as SpaceX is. Not every service it offers fits into this category; the short training programs seem mostly to provide mar- ginal benefit, given that short training is already widely
available.
But some of the other
areas have the potential, I think, to save the govern- ment lots of money.
There is another advan- tage to staying under the micropur- chase threshold. Past performance might be taken into account without any formal rating system, not to speak of an appeals process. The approach promises to introduce the kind of robust use of past performance that happens in the commercial world, which would strongly incentivize good
performance.
If microconsulting can gain traction,
it could become a disruptive innova- tion that forces legacy contractors to change their consulting models. n
Steve Kelman is a professor of public management at Harvard Uni- versity’s Kennedy School of Govern- ment and former administrator of the Office of Federal Procurement Policy. His blog can be found at fcw.com/thelectern.
In the comfortable environment of limited competition inside
the government bubble, it is far easier for old practices to persist well beyond their use-by date and far less likely for disruptive innovations to be developed.
Under “code quality assessment,” Cairns and his colleagues write, “We’re experts at rapidly evaluating the quality of any codebase (large, small, legacy, modern), measuring technical debt and recommending improvement actions.”
The company also suggests using something it calls “photosketching,” which is a disciplined way to quick- ly develop an image of one or more requirements for a software product, which then can be used to guide devel- opment work.
My favorite is “rent a CTO,” which the company’s material characterizes as a way to “obtain quick advice on the suitability and trade-offs of using a particular architectural pattern, deliv- ery practice or technology from one of our digital experts.”
When I saw the announcement, my thoughts immediately turned to Musk
August 2017 FCW.COM 47