The Ecosystem       1

You first need to know that there’s no such thing as selling to “the government,” despite our own promiscuous use of the phrase. Most market verticals that exist within the commercial world similarly exist within the public sector, including healthcare, finance, logistics, and real estate. The main organizing principle of the federal government isn’t function, however. Ask most federal officials to describe their work, and they’ll begin with their department or agency. The only civil servants who when asked just say they work for “the government” invariably work in the intelligence community.

Just as federal workers identify themselves by the cross-coordinates of function and agency, so must you. The entire federal government is too big to treat as a single entity, and its technical environments are too diverse to presume that every agency is a potential customer. Even the giants of federal contracting break themselves down into smaller divisions (although their constant reorganizing suggests there exists no perfect solution to the problem of encompassing both agency and function within a single logical unit).

In seeking to sell to all federal agencies, you’re likely to sell to none. After all, you wouldn’t try to sell to “the private sector.” Our use of the term the government throughout is a necessary shorthand for when we talk about certain common rules and assumptions held by executive branch departments and agencies—which, despite acting very much as separate organizations, have the common bonds of federal fiscal and procurement law.

How Big a Market?

We mention this right at the beginning in order to dispel common misconceptions about size. When you see the federal market characterized as half a trillion dollar large—the amount the government spends annually on goods and services—that’s accurate only in a macroeconomic sense. That amount is not a total addressable market for any company, since it encompasses everything from vacuum cleaners to supercomputers.

Same thing goes for sales figures for General Services Administration schedule contracts—close to $40 billion annually, as of this writing. Schedule contracts also encompass the entire breadth of the American economy.

Often, a more specific figure announced annually by the Office of Management and Budget (OMB) purporting to tally projected federal information technology spending during the coming federal fiscal year gets used to represent the size of the federal IT market. In recent years, that number has hovered around $80 billion. Certainly that number is closer to practical reality, although it’s still impossible for any one vendor to be all things to all agencies. No one person or office controls the entire federal IT budget, which is fractured by the boundary lines of departments and agencies. As we want to stress, there isn’t really a unified federal market—just a collection of agencies and functions bound together by common operating and policy principles.

The OMB number is also incorrect for other reasons. As a rule of thumb, contractors get a crack at only about 60 percent of what agencies say they will spend on IT. Also, the real collective federal spend on IT is actually higher. OMB’s figure doesn’t include the intelligence community IT budget. That’s classified, but credible estimates have pegged the figure at around $10 billion. Nor does OMB include spending by quasi-governmental organizations, such as the postal service. Nor does it tally what the judicial and legislative branches spend annually on IT, although admittedly, those branches don’t spend much.

It’s even questionable how well agencies whose spending is included in the OMB analysis accurately report their IT investments, since OMB enforces no standard definition of “IT” and agency definitions can vary significantly. Some agencies include research and development IT projects, others don’t. The Defense Department doesn’t add spending on IT that’s embedded into weapons systems despite the fact that no modern weapons system is without significant hardware and software components. Nor does the Pentagon always break out the IT infrastructure supporting those programs. Civilian agencies with satellite programs act similarly.

Given current spending trends, the real IT budget could be 40 to 50 percent higher than the annual OMB figure.

Why Agencies Buy Information Technology

Federal agencies have a purpose for existing, what feds call an agency’s mission. No federal agency today could accomplish its mission without primary IT systems dedicated to data collection, creation, transmission, manipulation, analysis, or storage.

Federal agencies are also organizations that need IT to sustain the daily grind of managing resources, balancing the books, hiring people—what is often called business lines activity.

In addition, federal agencies have IT shops of their own for maintaining and defending their networks, managing data centers, developing systems, refreshing technology, and so on. See Figure 1-1 for a breakdown of categories of federal spending during fiscal year 2011.

Agencies typically initiate new projects that directly address mission execution in response to a mandate from Congress or senior administration officials. The past decade has seen a significant shift in government. The FBI, once primarily a crime-fighting agency, has transformed into part of the national security apparatus. The Homeland Security Department brought together 22 agencies previously parts of departments as varied as Justice and Agriculture.

Even when the federal government isn’t undergoing a major renovation, there’s constant tinkering with agency priorities and execution. Most administrations go through at least one well-publicized event that starkly illustrates the inadequacy of federal operations, and they typically react by initiating high-priority modernization programs.

Sometimes Congress decides to mandate technological change. For example, following congressional hearings in the mid-1990s about allegations of abuse of taxpayers by the IRS, it passed the IRS Reform and Restructuring Act of 1998, requiring IRS to update its IT infrastructure.

More generally, nearly every president since World War II has declared he will “make government more efficient” and “cut out inefficiency.” This is unlikely to change since there can never be enough efficiency.

Members of Congress similarly keep up constant low-level pressure, especially since the legislative branch appropriates the money that federal agencies spend. Members of Congress get proprietary over “their” agencies and demand a constant stream of reports and briefings on all aspects of agency management, from execution of the mission to internal administration.

Officially appointed watchdogs also constantly monitor all aspects of agency performance. Each federal agency has an office of inspector general that provides a steady stream of audits nearly always loaded with recommendations, some of them regarding IT systems, others whose resolution could involve IT. The Government Accountability Office (GAO), the investigative arm of Congress, is also a report- and recommendation- generating machine, delivering about a thousand studies a year. Pressure from watchdogs to make changes isn’t insignificant, especially when Congress latches onto the reports they provide.

At the business line and IT infrastructure level, a big influencer of agencies’ plans is OMB, which uses its position as the federal budget request adjudicator to issue and enforce spending policy. OMB, if you’ve yet to otherwise come across it, is easily the most powerful federal agency you’ve never heard of. Located within the Executive Office of the President, its sole mission is to oversee the mission execution and spending of other agencies. Among OMB’s purposes is to diminish areas of redundancy, so it’s apt to do things like tell agencies to consolidate data centers. Within it is housed the position of Federal Chief Information Officer—a position that carries less authority than the title suggests, but one that nonetheless can be influential in setting the federal-wide IT agenda.

An agency’s sense of mission execution and efficiency matters greatly, too. Agencies brim with strategic plans, performance measures, and online dashboards. Feds take the performance of their organizations at every level seriously; the myth of the pencil-pushing bureaucrat doesn’t accurately describe today’s federal civil servant, who typically is well-educated, dedicated, and encouraged to come up with new and better ways of doing things.

Federal Attitudes Toward Technology

Government workers want to do their jobs well and like no more to be left behind the technology curve than anyone else. But the burden of legacy applications weighs heavily. COBOL and mainframe systems persist because ripping them all out would be expensive, not just due to the sheer size of the primary systems, but also because of the myriad secondary systems with dependencies on them.

Although there’s a trend toward centralization of IT operations within agencies, many individual components of larger departments maintain a fair degree of autonomy, resulting in a plethora of mom-and-pop shops throughout government, each with its little IT quirks. Even something as relatively minor as a medium-sized installation of Lotus Notes can be exceedingly difficult to rip out because each covered administrative unit has accumulated its own in-house programmed applications running in the Lotus environment.

In the long run, maintaining obsolete systems is more expensive than replacing them, and the government knows this as well as anybody else. But replacement requires upfront capital funds, and getting those from OMB and Congress is far more difficult than muddling through.

Cost isn’t the only thing keeping legacy systems in place. Change brings up the specter of disruption to agency operations, something that the government especially loathes. With good reason; at the upper scale of functionality, lives depend on government operations, and not just within the military. If the Social Security Administration were not able to send out disability checks, or were the Department of Veterans Affairs to lose patients’ data, some Americans would die.

As a result, the government is naturally averse to incorporating new technology into the production environment. Anything that’s bleeding edge typically won’t find a place in agencies, except around the edges or in special environments. That’s not to say the government isn’t interested in learning about the latest developments; it is very interested. But it flirts more it than commits to the latest version of anything. It likes to see examples of a technology working well in production elsewhere before incorporating it for itself.

Sometimes it’s federal procedure that makes technology adoption difficult. The U.S. Citizenship and Immigration Services, a notoriously atavistic agency when it comes to technology, has clung to paper-based processes in part because of mandatory regulations pertaining to matters such as the correct filing and handling of requests.

On its own, the government would probably never deviate from a process of incremental change, entering into major modernization efforts only when truly pressed to do so. However, the government doesn’t operate in a vacuum. As noted in the previous section, a whole slew of actors and pressures force it into a new equilibrium.

The Rise of Services

In fiscal 2010, the Defense Department spent more than half of its $367 billion budget on services contracts.¹ That same year, the Homeland Security Department spent nearly three-quarters of its $13.6 billion on contractor services.²

The rise of services contracting to a dominant position in the federal IT market has its genesis in Clinton administration–era legislation that resulted in across-the-board cuts to federal employees. It received a shot in the arm by the post-9/11 expansion of government, along with the contractor-friendly policies of the Bush administration. Although the Obama administration has said it wants to bring back in-house some of the human capital the government has outsourced via services contracting, it’s also actively promoted another type of new services contracting: the outsourcing of information technology functionality via cloud computing.

Lest you think you’ve struck a gold mine, research shows that the past decade’s sharp increase in the number of services contracts let per year has been accompanied by a decrease in average value, even when filtering out contracts worth less than $25,000; see Figure 1-2.³

But in contrast to the private sector, services companies, once engaged by the government, pretty much can count on years of steady revenue. The government typically contracts services for 12-month periods with several options to renew, and any company that falls short of incompetence gets its options renewed. Payment amounts are steady and consistent. Also unlike private sector firms, the government can’t legally address cash issues by squeezing bills outstanding without paying interest penalties.

Although products haven’t been totally eliminated from the federal marketplace, many product companies have looked for additional business by grafting services onto their offerings. Some value-added resellers have even taken to using products as loss leaders in order to obtain services contracts.

Perhaps the strangest phenomenon associated with the government’s shift away from buying things to buying services has been the emergence of traditional high-end consulting companies that found they could make more money by establishing relationships with manufacturers in order to sell “solutions” to federal clients. Their business model depends even less than that of traditional systems integrators on making things and even more on presentation and repackaging. Because these companies retain their consulting divisions, they occupy a potentially ethically dubious middle ground between a trusted advisor and a straight-out systems integrator. Some have developed reputations for playing particularly intense hardball with manufacturers.

Who in Government Buys IT

Nobody except a contracting officer can bind the government contractually. If an acquisition doesn’t pass muster with the procurement office, it won’t happen. But contracting officers aren’t in charge of identifying when or why the government must buy from the private sector. Every acquisition has a set of requirements that describe what needs to be bought, whether products or services. But who generates these requirements? For all of the hierarchy that exists in government, it’s impossible to tell precisely by looking at an organization chart.

A natural, but incorrect, assumption is that all departmental or agency IT funding is controlled by the department CIO. In fact, that’s the case in only a few major federal agencies. In a 2011 survey of 30 major agency CIOs, the GAO found just nine could say that all agency IT investments must gain their approval before they’re included in the agency budget. Large departments have CIOs at the component level, officials whom the headquarters CIO may have no say in selecting or evaluating. Some CIOs don’t even hire their own staff.

It’s difficult to accurately generalize about an organization as vast as the federal government, but the position duties of many an agency headquarters CIO are limited to policy formulation, oversight and, to a greater or lesser degree, agency IT infrastructure. See Figure 1-3 for a breakdown of how CIOs at large agencies spend their time, based on the 2011 GAO survey.

That’s not to say that headquarters CIOs aren’t involved in the IT projects that directly affect agency mission performance. Top CIOs sit on the investment review boards that major IT projects must go through to get funding. But their ability to affect the decisions of program managers can hinge on matters not captured by job title. Someone who has the ear of leadership, who is respected by peers and underlings, who has a battle scar or two from implementation of a program of his or her own, will be treated with deference. The opinions of a CIO still redolent of political fundraising dinners? Not so much.

Even if your product or service is related to infrastructure, the headquarters CIO might or might not be the right person to speak with. It could be that an infrastructure project has its own program manager. If the CIO shop is large enough, the CIO may have delegated authority over matters to do with your product or service to other executives within the office. The scale or scope of your product or service could fall below the threshold at which the CIO needs to be involved. One former federal CIO we spoke with said CIOs are more portfolio managers than technologists.

Program managers, the people who oversee project implementation, often are better prospecting targets than CIOs. But not always. Program managers very often come from an acquisition background and can be ignorant of technology. It’s possible they’ve delegated technology decision-making duty to subordinates or that they take their cue from above. When real decision-making is obscured rather than made clear by organizational charts, there’s little option except to ask around until you find who’s the major influencer or actual decisionmaker. Just remember that it’s not always the person with the biggest title.

The fact of delegation and division of authorities within the federal government bears some emphasis. Despite—or rather, because of—the presence of political appointees in key positions within agencies, the government takes pains to insulate buying decisions away from political or top-down pressure. As a result, going to the person sitting in the corner office might not only be a waste of time but could even damage your reputation in an agency.

“What will invariably happen if a senior executive calls on a senior government official and tries to force or sell a solution or a service is the government guy is going to say, ‘Talk to my guys down here. They take care of that.’ So you haven’t built a relationship, and you’ve actually shut that door,” a senior executive told us.

Topside Consulting, a Vienna, Virginia-based consultancy of mostly former high-level feds, tells its customers that four basic layers in government exist:

•  Political appointees

•  Members of the senior executive service (SESers) or military flag officers—the people who make sure the railroads run, but they’re not involved in the details

•  Grade 15 and 14 civil servants or O-6 and O-5 military officers—middle management in charge of supervising day-to-day operations

•  Grade 13 civil servants or O-4 military officers and below—the people who actually perform the operations and deliver the output of federal programs.

Ray Bjorklund, a well-respected thinker about all things federal acquisition, tells us this model slides up a notch outside the political appointee–heavy national capital region. In other parts of the country, SESers and flag officers have the role of leadership and public relations, GS 14/15 and O5/O6 are the chief operating officers, and GS13s and O4s step into the supervisory positions, with operations and output done by company-grade officers, GS12s, and noncommissioned officers.

In order to sell, your goal is to find the highest level that can make the decision to start pulling the levers of the procurement process, and not to go above that for a particular transaction. If you have a relationship with the people above that level, that’s fine, but don’t attempt to bigfoot the subordinate, since intervention from higher levels in procurement decisions is frowned on in the government. In the long run, it’s better for your company to gain support bottom up rather than top down; people at the top change more frequently. The median tenure of CIOs at major agencies is 25 months—only a quarter stay at their agency for at least three years, according to the 2011 GAO survey.⁵ The median tenure for politically appointed CIOs is 21 months.

“Start at the bottom. And tell them if you’re going to go up: ‘We’re seeing so-and-so next week, is there anything I need to tell him, is there any ground we need to be careful of, should we mention this at all?’ If they say don’t bring it up, don’t bring it up,” advised Bob Woods, Topside’s president.

Ultimately, you should aim to engage the government at all of its levels, but appropriately. Who you aim to talk with depends on who you are and what your intention is. A chief executive officer interested in a strategic conversation about long-term engagement with an agency should have that conversation with the CIO, chief financial officer, or agency head, but shouldn’t expect that to lead in linear fashion to a signed deal.

ENDNOTES

1. U.S. Government Accountability Office. Defense Acquisition Workforce: Better Identification, Development, and Oversight Needed for Personnel Involved in Acquiring Services. GAO-11-892. September 28, 2011.

2. David J. Berteau and others. DHS Contract Spending and the Supporting Industrial Base. Washington, D.C.: Center for Strategic and International Studies, 2011.

3. David J. Berteau and others. Structure and Dynamics of the U.S. Federal Professional Services Industrial Base, 1995–2009. Washington, D.C.: Center for Strategic and International Studies, 2010.

4. U.S. Government Accountability Office. Federal Chief Information Officers: Opportunities Exist to Improve Role In Information Technology Management. GAO-11-634. September 15, 2011.

5. Ibid.