Startups are revolutionizing nearly every segment of the US economy, but the federal government has remained relatively untouched. This is not owing to a lack of interest by the innovation community. BCG and Eastern Foundry conducted a survey of 109 organizations made up of startups, venture capital firms, and angel funds. (See the sidebar, “Survey Methodology.”) Most of the startups were interested in contracting with the government. But numerous obstacles deterred them, including a lengthy and overly complex contracting process, a lack of clarity on how to connect with agencies, and a sense that newcomers have little chance to win contracts over incumbents.

In this report, we closely examine these deterrents and identify bold actions that the US government can take to overcome them. By modifying its approach to contracting, cutting red tape, and communicating more effectively with startups, government agencies can access cutting-edge innovation and deliver better outcomes for citizens.

What Drives the Startup Community

Now, more than ever, startup companies drive innovation, creating new technologies and capabilities. However, because government procurement processes are slow, startups typically don’t view government agencies as potential clients. Contracting with the government is attractive to startups only if it can help them achieve their goals:

• Earning Recurring Revenue. Recurring revenue is the number one priority of startups. (See Exhibit 1.) It allows a company to grow without having to rely on investors, and it helps boost a company’s valuation. In theory, the government should be an attractive customer. As one respondent noted, its “consistent revenue is attractive, and its contracts are typically longer term than most private-sector contracts.” But because the government’s traditional procurement process is complex and long, it is costly for startups to participate.

• Raising Capital. The amount of money that startups have in the bank is closely tied to the amount of money they’ve been able to raise from investors. The amount of money in the bank, along with their burn rate (the speed at which they spend money) determines their “runway” (the amount of time they have before they need to raise more money, generate enough revenue to cover costs, or go bankrupt). Capital infusions are the only way a startup without revenue can continue to survive.
• Boosting Valuation. Embedded in the capital-raising process is the perceived value of the company. A high valuation allows the company to raise capital without giving away too much equity. But most companies would prefer to generate revenue than give away equity. Consequently, quick sales are vital.
• Finding a Market Fit. To succeed, a product must attract customers. Because typical startup solutions often possess a short “shelf life” before competing technologies or capabilities emerge, it is important for startup companies to rapidly identify customers for their products.

Startups will carefully evaluate how pursuing a government contract can help or hinder their ability to achieve these goals.

Barriers to Working with the Government

Startups have an incentive to pursue government contracts because they can be worth millions of dollars. Of the startups we surveyed, only a small percentage said that they had no interest in contracting with the US government. But numerous deterrents (some real and some perceived) stand in the way of many leaders. (See Exhibit 2.)

Lengthy Sales Cycles. The number one complaint from startups was about the time it takes to secure government contracts. In the words of one survey participant, “A company that has been in business for 6 months and has 12 months of funding remaining is not going to be looking to spend 24 months establishing a relationship with the government. The government must find a way to convince startups that they have a realistic chance of winning a contract within a relatively short period of time.”

A major government agency recently announced that it cut the time it takes to make major IT acquisitions from four years to two years. But this is still too long. Startups are racing to form a network of clients that will provide cash fast—often within 12 to 18 months. It’s not practical for startups to pursue contracts that require two years to close. By then, many startups could cease to exist.

In addition, with technology startups, innovation is their competitive advantage. A company whose product has a nine-month “shelf life” before competitors catch up will lose its edge less than halfway through a two-year sales cycle. Startups fear that their technology will be obsolete (and therefore uncompetitive) when the government finally makes a purchasing decision.

Complex Processes and Significant Effort. The complexity of the contracting process lengthens the time needed to pursue a contract. First, rigid requests for proposals (RFPs) mean that a startup must write a lengthy proposal tailored to a specific approach. Second, startups often do not have “indefinite delivery, indefinite quantity” contracts (or framework agreements) in place with the contracting agency. They also often lack a General Services Administration schedule. This effectively makes them ineligible for a high percentage of contracts.

Government agencies advertise opportunities primarily on one portal: the Federal Business Opportunities website. But startups often lack a business development department to monitor this website and sift through the thousands of opportunities. The search process requires a great amount of effort. Only larger contractors can afford to pay someone to monitor the bid boards.

Negative Government Experiences. Our survey indicated that respondents are concerned that government work is not viewed as prestigious, partially because of negative government experiences. Some startups expressed reluctance to work with the government for fear that it will tarnish their brand in the eyes of potential employees or customers.

No Clear Agency Contacts. A widespread belief among startups is that only established players with the right connections and sunk infrastructure have a chance of winning government work. (See Exhibit 3.) One survey respondent said, “From my understanding and experience, winning a government contract tends to mean a company is best at winning government contracts, not the best qualified for the job.”

What Government Agencies Can Do to Attract Startups

There is much to be gained by the public sector and the innovation community connecting. The government could more rapidly adopt cutting-edge technologies and business practices while improving its efficiency and effectiveness. And startups could generate stable revenue to support long-term growth. Nonetheless, most of the startups surveyed said that they would like to see meaningful changes in the contracting process. Armed with this knowledge, we have identified several actions that the US government can take to successfully engage the innovation community.

Push opportunities through private-sector channels. The government currently expects startups to approach it. As mentioned above, federal websites such as FedBizOpps.gov are the primary way to advertise procurement opportunities. But the government should approach startups—go where they go. As one respondent noted, startups have trouble working for the government in part because of the “challenges associated with identifying appropriate contracting opportunities.”

Startups traditionally connect with potential customers at such venues as pitch competitions, hackathons, and trade shows or through accelerators. Startups have strong relationships with investors, such as venture capital firms and angel funds, and founders tend to read online news sources, such as TechCrunch, Wired, and TheFunded. If the government were to target the venues, investors, and news sources that are primarily oriented to the private sector (using advertising, for example), it could significantly increase its engagement with startups.

Focus on outcomes, not approaches. In the current acquisition model, the federal government tends to be overly prescriptive, specifying not only what it needs but also how the vendor will provide it. This model unwittingly deters companies that have novel solutions. A respondent we spoke with noted, “The biggest challenge is a lack of meaningful, constructive dialogue with an agency. The formalized process makes it difficult to figure out what an agency is really looking for, and the continued practice of wiring opportunities for specific products and solutions is highly discouraging.”

Rather, the government should focus on the results it wants to achieve. For example, if an agency wants to reduce the expense of replacing helicopter blades, it should focus on this outcome in its RFP. Companies could then respond with different approaches, such as a solution that prevents blades from breaking, a device that measures blade strength so mechanics can determine which ones need to be replaced, or a solution that lowers the cost of repairs. An overly prescriptive RFP causes vendors with other solutions to either reengineer their solutions to fit the government’s approach or refrain from bidding altogether.

The government could try to counteract the narrowness of an RFP by issuing a request for information (RFI) or hosting a government-industry day. The purpose of each would be to help agencies learn about different approaches, which should inform the RFP. But startups are unlikely to engage in either an RFI or a government industry day unless it is broadcast widely enough. Additionally, the timeline from RFI to RFP to contract award is too long for most startups.

A survey respondent suggested that the government create a portal “where startups can specify what they do so that government employees have an idea of what is possible.” Another respondent suggested that agencies fund pilot projects so that startups can get a foot in the door and show how their technology works. This approach could help address a problem reported by another respondent: “Federal procurement officers will not take any risks on new technologies that, if they fail, may reflect poorly on their performance evaluations.”

Model government procurement after private-sector procurement. A startup is building not only a product but also a team with institutional knowledge and capabilities. Although the goal is always to win new contracts, there is value in losing—so long as the startup can learn from it.

Unfortunately, survey respondents reported that the government sales process is sufficiently different from the commercial sales process to negate any learnings, capability building, and content development within sufficient time frames. If the federal government could model its RFP process after that of the commercial sector—by focusing on outcomes and being less prescriptive, for example—then the capabilities and materials developed for federal work might have a higher salvage value. This would mitigate the downside risk and likely increase the number of proposals received.

Provide startups with more feedback. Startups are eager to understand the ways that a customer wants to use their product or discover customer pain points that the product could address. Unfortunately, the government provides very little feedback to vendors and only rarely provides shareable use cases. Additionally, the government’s security, customer access, and integration requirements frequently mean that insights gained from a government customer are only relevant to other government customers. To address this, the government could establish simple procedures for government officers to give feedback on the proposed product and provide use cases.

Streamline contracting. Given the relatively small amount of money needed to interest startups, the government could deploy a wide array of fast, flexible contracting mechanisms to induce engagement. This can be done in a number of ways:

• Employing Off-Ramp Clauses. The government could make more use of off-ramp clauses, which are common in research and development contracts. Every two or three months, the government could assess a startup’s progress by examining the functionality of its prototype. The off-ramp clause would give the government the option to exit the contract if it is unsatisfied. Currently, the government often awards contracts involving emerging or complex technology to big contractors. These contractors’ large workforces enable them to commit to delivering this technology at a certain time, albeit far into the future. Using an off-ramp clause, the government could contract with a startup that could execute the contract in far less time. In the event the startup is not making adequate progress, the government could always exit the contract.
• Deploying Simplified Acquisition Procedure Contracts. Simplified acquisition procedure (SAP) contracts are available for purchases of less than $150,000 and can be completed quickly. Although this is a tiny contract in the public sector, it is well within the range that is of interest to most startups. If the procedures for an SAP contract were modeled after commercial RFP standards and opportunities were properly advertised, then SAP contracting could be an extremely effective way to engage startups. Echoing this interest, one respondent suggested that the government give “smaller contracts to startups as a way to try out new services.”
• Using Other Transaction Authorities. Other transaction authorities (OTAs) were specifically designed to be fast and flexible contracting mechanisms that allow the government to buy and test prototype technology. As such, OTAs are ideal mechanisms to use to engage with startups. Unfortunately, contracting officers are largely unfamiliar with OTAs and therefore hesitant to use them. To overcome this reluctance, contracting officers could receive training, “cheat sheets” on their use, and case studies that illustrate their value. To better reach the startup community, OTAs could be made available to only those companies whose engagement with the government—as measured by annual government sales, for example—is below a certain level for the past five years. In addition, applications for new OTAs could de-emphasize requirements for previous experience with OTAs.
• Awarding Grants. The Small Business Innovation Research (SBIR) program provides $225,000 grants (nondilutive capital) to small businesses with early-stage technology. These grants should be of interest to all but the biggest startups. However, very few startups outside academia are aware that the program exists. Furthermore, the application and evaluation processes can be opaque, complex, and slow, discouraging companies from pursuing these grants. The SBIR program could make efforts to increase the awareness of these grants in startup-friendly venues.

Increase the use of challenges. Many agencies run challenges that award $10,000 to $50,000 to companies with early-stage technology solutions that meet an agency need. The challenges usually take one to three months to complete. They have simple application processes, making it easy to apply. To increase the impact of government challenges, we recommend applying best practices from private-sector challenges:

• Ensure that all participating companies receive some value by publicizing their capabilities and giving them robust feedback.
• Clearly communicate all information about the challenge, adhere to announcement schedules, and ensure that participants are aware of major milestones.
• Mirror the applications and processes of commercial events so that startups don’t have to spend time creating new documents.
• Advertise through channels that are likely to reach startups (not only Challenge.gov and the internal Small Business Administration channels).

Treat startups as investments. Government acquisitions are typically structured as a series of discrete events: an RFP is issued, a winner is selected, a prototype is purchased, and a contract is awarded to meet a current need. Instead, the government should think of each startup as a long-term investment and work toward offering progressively larger contracts that will deepen the government’s engagement. By creating and curating a pipeline of startups, the government will likely spend less money and gain access to more innovative products over the long term.

Align requirements generation and product development. Traditional government acquisition processes draw a clear distinction between a requirements generation phase and a subsequent product development phase. Frequently, separate organizations are required to complete each phase. However, several innovative startups now combine the two phases into one, cycling between product development and requirements generation in a continuous and iterative way. In addition to the significant learning that occurs when the same company actively participates in both product design and development, a startup gains the ability to deploy usable products and capabilities to market much more rapidly than with the traditional approach.

Use large defense contractors as allies. When regulation ties the government’s hands, the government could rely on the traditional defense contracting community to provide startups with access to projects. For example, by subcontracting to startups, traditional defense contractors can help get the word out about government projects. And when a large contractor brings startup subcontractors together, these startups can profit from a project even though they are not able to cover its entire statement of work. Large defense contractors could also seed startups with cash when awards or payments are delayed. There is precedent for large defense contractors partnering with innovative players to deliver cutting-edge products to the government. Lockheed Martin, through its subsidiary Sandia Corporation, has managed Sandia National Laboratories, a federally funded research and development center, and associated startup companies that have engaged with it, for almost 20 years.

If the government can lower the barriers to entry at all stages of the procurement process, that will go a long way toward counteracting one respondent’s characterization of how government procurement currently works: “slow, designed to favor a preselected winner, and looking for a specific answer.” If the government seeks out startups, search costs can be partially shifted from the startups to the government. If startups begin to view the bidding process as a learning opportunity through which all bidders receive value, then the bidding process will move away from the current winner-take-all model, which has a large risk that is more easily borne by established incumbents. And if contracting costs can be lowered through simplification, then startups will be more willing to pursue government contracts. By applying private-sector lessons to the public sector, the government can more fully enjoy the creative destruction that startups provide.

Acknowledgments

The authors thank Sharon Marcil and Dan O’Brien for their contributions to this report. They also thank Amy Strong for her assistance in writing this report and Katherine Andrews, Gary Callahan, Kim Friedman, Abby Garland, Sean Hourihan, and Trudy Neuhaus for their contributions to the editing, design, and production.