Avoiding the danger of custom technology

Budget pressures might be pushing some agencies inward to solve technology problems, but the risk is higher cost and less flexibility, according to Thermopylae Science & Technology president A.J. Clark.

Oftentimes the conventional wisdom inside federal agencies is that building a software solution costs much less than buying it. However, there is little to no empirical data that proves that out.

Instead, a lot of money is spent creating custom solutions that don’t scale up or crossover well to other agencies, and then cost even more money for years after to perform custom maintenance and upgrades.

For many years, intelligence, surveillance and reconnaissance  imaging used by federal agencies came only from expensive, private government-owned and operated satellites with handcrafted back-end systems that processed and analyzed those images.

With the advent of commercial satellites and cloud software – and its eventual acceptance by government agencies – millions of dollars in costs were cut. Meanwhile, time to market, efficiency and effective use of those commercial systems has increased dramatically.

It’s an ongoing debate in the defense industry whether government-created solutions or commercially created ones are better. There are advantages to both: government solutions are proprietary, which may be needed for secrecy sake, but they are more expensive in the long run. Commercial solutions, although not exclusive, have lower development and deployment costs, and allow faster time to market, quality control and flexibility.

However, after years of debate, it’s time to more widely adopt commercial solutions.

It’s not a surprise that government budget decision makers have turned to their own IT people to craft solutions. Procurement officers request funding for the handful of items they absolutely need and know will likely be approved.

But to avoid the funding request hassle, budget managers often turn to their in-house staff. They’ll ask engineers and IT employees to handle other software needs. The situation is complicated by the fact that the internal staffers, like many smart and curious people, become interested and intrigued by the project and then don’t want to adopt a commercial solution because they enjoy solving the problem themselves.

There are reasons why this happened. Free flow funding is not a federal legislative strength. Between 1976 and 2013, there were 18 government shutdowns over budget disagreements. And while continuing resolutions are common – they’ve happened all but four of the last 40 years – it makes for tenuous and tentative budgeting styles by procurement managers.

Agencies that need funding have to make requests for specific projects or needs, but then legislative budget hemming and hawing can stop, stall or prolong the request. And what happens if an agency request makes it through the usually more than year-long budget process only to be turned down? Time to start all over.

What can be done?

The addiction to building a solution internally is entrenched and budgets are still tight, so there is no easy fix. However, there are some ways we can all work toward better, more efficient and less expensive outcomes.


Create crossover standards for government and commercial solutions. With specific standards, commercial software builders can make sure that government agencies would be able to “plug in” all or parts of their solutions.

Right now, there are a few efforts underway, such as the Defense Intelligence Information Enterprise or DI2E, which is working to establish an enterprise approach of standards and specifications among the intelligence community, Department of Defense and international partners. As those efforts show success and provide a model for other federal agencies to adopt, there could be some great lessons learned.


Evaluate commercial technologies upfront. Procurement officers should review commercial solutions to look for potential off-the-shelf resources as a matter of course, but when it comes to specific projects, they should assess those solutions in the beginning stages of the problem.

When commercial software is an afterthought, it’s less likely to be used. The intense aversion to paying license costs, which comes as a reaction to major overruns from Fortune 100 software firms is always top of mind but does not accurately reflect the likely outcome of every commercial technology.


Use experts to do cost analysis measurements. Today, analysis inside the government is often done by engineers who have specific skill sets, not necessarily broad budgeting or software aptitude. To them, the upfront spending on a boxed solution will always look more expensive than getting five engineers already on the clock to figure it out. Time, ongoing costs and maintenance are almost never figured in.

All too often it seems that procurement decision makers assume that the commercial solution won’t work until it’s proven beyond a shadow of a doubt that it will.  I believe the opposite should be occuring; the procurement decision makers should prove that commercial solutions will not work before they try to roll their own.


Tie government funding to accurate analysis. As part of DNI or DoD acquisition office rules and regulations for contracts, there should be controls or audits to make sure proper analysis is done. This applies to ballooning world of shadow IT most specifically.

Government employees or contractors that happen to have some time available to apply to building some replica of a commercial solution are oftentimes the cracks that the seeds of new starts grow out of.  Their cost is already budgeted for and before the Government knows it, a new tools is birthed that requires care and feeding for years to come.

Another important factor that is helping to change the landscape is also happening organically. Young people moving into government positions are forcefully driving the demand for commercial solutions. They’re accustomed to, and expect, the latest technology that government can’t reproduce in-house or quickly. Hundreds of billions of dollars are spent on software every year and while the Federal Government was a leader in this area 25 years ago, they do not have a competitive advantage there any longer.

We can learn from public sector CIOs who have seen software costs falling over the past few years as they leverage the cloud and bring-your-own-device policies. The government can’t capture all of those same cost savings, but agencies can take lessons from the commercial sector. Technology planning, for example, shouldn’t be about playing catch up, but rather figuring out what will be needed in the next year or two, identifying the costs and budgeting for that future.

It’s time to re-evaluate the widespread use of government-created software solutions. The recession and budgetary crises of the past few years have mandated that cost effectiveness as the new operating mantra for defense agencies and their partners. That means adopting the latest technologies already available in the commercial sector. New trials and experiments, not defaulting to build it yourself, is the modern business model our industry needs today.