Industrial base: “Portfolio” investment strategy
As soldiers met evolving threats in Iraq and Afghanistan, the Joint Improvised-Threat Defeat Organization (JIDO) needed a way to tap into a trusted group with ready solutions to counter new types of improvised explosive devices (IEDs)—and to do it quickly and often.15 By continuously scanning for the latest commercial capabilities, JIDO finds partners to help stay ahead of adversaries16—creating new technologies to counter threats from armed drones, for example17—and uses rapid acquisition authorities to get them to the frontline soon.
Building a diverse set of suppliers like JIDO’s commercial partners is an example of an “industrial base” partnership—one variant of a broader “portfolio” investment strategy, which can be used to source products whose usage is well understood. Commercially described as a “direct” investment strategy—designed to advance the strategy and operation of the current business18—it engages external partners as a way to diversify production risk and gain outside perspective to reduce “disruption risk.” A commercial example is Pfizer shifting funding from its neuroscience R&D to its venture group—gaining valuable insight into new therapies and diversifying its pipeline in a fast-moving and uncertain field.19
The US Department of Defense has long relied on a large industrial base to efficiently supply military equipment—what’s important today is to broaden the base to include nontraditional players and startups to accelerate innovation. “The companies that serve the armed forces cannot produce innovations fast enough, and the often-smaller companies that excel at innovation are not focused on defense,” says Steve Blank and Pete Newell, creators of Hacking for Defense.20
To sustain the portfolio, government can use two mechanisms: innovative acquisition approaches and venture funding arms.
Innovative acquisitions. While new vehicles alone will likely not attract new players (as the aforementioned GAO report suggests), tailoring procurement mechanisms to companies’ growth trajectory could make government a more sought-after partner. The ability to solidify a short-term revenue pipeline can allow startups to focus on product build-out, for example, rather than having to spend as much time on sales. The US Air Force tested this hypothesis in what might be the most rapid contracting process to date, awarding pilot contracts within a day. Using a process that looked more like that of startup investors, the Air Force team vetted solutions based on 15-minute pitches, and as a result, attracted a novel company profile: over half of awardees had not worked with the US government before.21
NASA has also been an early leader in using innovative procurement strategies to seed the commercial space industry—and scale and sustain companies that help the agency deliver supplies to the International Space Station or to support craft that operate in low Earth orbit. By offering SBIR grants to small companies that are rich in new ideas but short on resources, the initiative has helped to create a commercial space economy predicted to grow to more than US$1 trillion in value over the next two decades.22
Venture funding. In other cases, a solution may not yet be ready to deploy or test in a specific mission area; here, governments can use traditional venture funding to maintain access to strategically valuable technology until they can find a foothold in the market. This scenario is the classic value proposition for In-Q-Tel: The organization provides startups capital and access to frontline users in return for equity and technology. More recently, the US Department of Health and Human Services Biomedical Advanced Research and Development Authority launched the Division of Innovation, Research, and Ventures to identify promising health security solutions, and then invest in these companies as part of a PPP.23
Accelerators: “Early-stage” investment strategy
Historically, food aid as a category has been particularly vulnerable to corruption. Humanitarian agencies need to ensure that aid reaches the hands of those who need it the most—this is particularly difficult in conflict situations.24
So when WFP sought a more secure way to track aid distribution, it turned to a nontraditional source: its in-house Innovation Accelerator, which uses new technologies to improve food security. “The World Food Programme is not out to invent new technologies. We are out to adapt things that we see going on elsewhere for the benefit of the people that we serve,” says Director of Innovation Robert Opp.25 The model is based on startup accelerator Y-Combinator, which invests US$150,000 in each startup selected (alumni include Airbnb, Dropbox, and Stripe) and then spends three months working with the startup to refine its product and attract additional investors. Similarly, WFP awards up to US$100,000 to build out proofs-of-concept for promising ideas—what Opp refers to as “scale-ups”—to a point where they can qualify for outside funding.26
To improve cash transfers to refugees, the accelerator prototyped a blockchain-powered transfer system in Pakistan’s Sindh province—and is now testing an updated version that includes retinal scanning at refugee camps in Jordan.27
Accelerators offer an early-stage investment strategy, providing resources to startups with less mature products to further develop a promising capability. They speed up innovation by offering access to customer insights, sector or category expertise, and real-world environments or assets. Like WFP’s Accelerator, these short-stint programs can help to position the startup for later funding rounds by providing the resources, experience, and mentoring to help commercialize the technology.
And by providing these assets, government gets hands-on experience with cutting-edge technologies for a fraction of the cost of acquisition. WFP originally built the accelerator in response to former executive director Ertharin Cousin’s concern that they were not taking advantage of emerging technologies and platforms—as Opp recalls, “She would travel around the world and see AI and blockchain and IoT, and come back to headquarters and say, ‘okay, what are we doing about this?’”28 Now, the Accelerator is experimenting with digital market places for smallholder farmers, hydroponic farming in refugee camps, and AI-based image recognition to diagnose malnutrition.29
Partnering with existing accelerators. Government can also get many of these benefits by partnering with existing accelerators. In fact, such partnering may not only allow government to use this strategy sooner but also more quickly attract startup interest; it can be tough to recruit startups initially, since they usually shy away from programs that lack an extensive commercial network.30 ENGAGE in Atlanta offers a corporate analog, where corporations fund and work with the accelerator in a “shared services” model, giving startups a range of use cases to test as they consider possible paths to profit.31
The US Department of Defense (DoD) has pioneered several similar accelerator partnerships: the Air Force Research Laboratory’s (AFRL’s) partnership with ABQid in Albuquerque, Army Futures Command’s partnership with Capital Factory in Austin, or the MD5 and AFWerx collaboration with TechStars. “Traditional technology transfer is about getting ideas out of the AFRL and into the marketplace, but this is the other way around,” says ABQid executive T. J. Cook. “We want to build AFRL awareness about new, privately built technologies that the Air Force could benefit from.”32
And it goes beyond defense: The US Department of Agriculture’s (USDA’s) National Institute of Food and Agriculture funds AgLaunch, an external accelerator for high-tech, precision agriculture startups. It offers a ready base of farms where agtech startups can test their product and get feedback from producers—helping to speed the maturity of agriculture technology.33
Sandboxes. One specific variant of accelerators is “sandboxes”—controlled environments to test new technologies or business models, often relaxing current regulations. In the fintech arena alone, Deloitte has identified more than three dozen regulatory sandboxes globally.34
Autonomous transportation has also spawned several testbeds, such as Michigan’s MCity,35 in situations where large-scale deployment would be dangerous but controlled experiments can yield useful insights for both companies and regulators. The Federal Aviation Administration, for example, granted temporary waivers to release 10 sites from drone flight restrictions such as flying at night or beyond line of sight of operators, allowing companies to test applications including medical equipment delivery, monitoring oil pipelines, and scanning the perimeter of an airport.36
Strategic partnerships: “Joint venture” investment strategy
To provide humanitarian aid in underbanked regions, the United Nations (UN) has to set up infrastructure for tracking and management—at significant cost and effort. To tackle this problem, the UN teamed up with an unlikely partner: Mastercard.
It turns out that credit card companies have core capabilities that are helpful in aid disbursement: fraud detection, electronic tracking, digital infrastructure, and even marketing. Mastercard’s collaboration with the UN helps it to monitor distribution of welfare grants using a smart card processed by a card reader that can be tethered to an Android device—replacing the manual processes used by local merchants to reconcile transactions.37 What is in it for Mastercard? An increase in brand equity, a chance to prototype new technologies, and an opportunity to expand into developing countries—potentially reaching 500 million new customers.38
When commercial assets are mature platforms—making them harder to simply procure—strategic partnerships create the public equivalent of a “joint venture,” capitalizing on areas where business and social incentives align. “When we talk about startup-driven innovation, we think of it as large organizations adopting practices that make startups successful. One of the practices is significant partnering—knowing who to partner with and how to develop these partnerships efficiently,” observes venture capitalist Evangelos Simoudis.39
And as in the case of Mastercard and the UN, these partnerships focus on solutions that benefit from consumer access or data. Domestically, several US cities have partnered with navigation app Waze to better understand traffic patterns, identify areas of congestion, and analyze modifications such as lane reductions or signal changes—all based on consumer-provided data, instead of sending out crews with traffic-counting equipment. Traffic engineers in one of these cities, Louisville, use the data to test whether the modifications work in real time.40
Similarly, producer data from the Farmers Business Network (FBN) might reduce the cost of studies for USDA researchers, and provide more recent and frequent data on US agricultural production for policymakers. Or a government agency could team up with one of the companies that specialize in genome testing: With millions of customers, these companies have a large amount of data on human genetics. With appropriate privacy controls, the data sets could be instrumental in facilitating medical research and advances—particularly for diseases with small populations or for which it has been hard to design studies.41
The Federal Bureau of Investigation (FBI), meanwhile, has been exploring the use of Amazon’s facial recognition software, whose machine learning algorithms improve based on access to a broad set of consumer training data. After a mass shooting, for example, the FBI pieces together the last weeks or months of the shooter’s movements from scores of camera footage. The FBI said that after the Las Vegas shooting in 2017, analyzing a petabyte of video footage of the days leading up to the event took three weeks, running “24/7, eight people per shift”—but it takes facial recognition software only 24 hours.42
Open architecture: “Platform” investment strategy
The health sector is a data-rich landscape, with inputs as diverse as genomic data, electronic health records, clinical trial data, and patient-generated data from wearables or mobile apps. Much of this data, however, exists in silos due to privacy concerns, competitive fears, or interoperability challenges. To break these silos, the US National Institute of Health (NIH) launched the NIH Data Commons, a cloud-based platform where investigators can store, share, access, and experiment with digital objects generated from biomedical research—thereby speeding hypothesis generation and validation. The Data Commons initiative is now in a pilot phase, using three high-value data sets as test cases for the principles, policies, processes, and architectures to be developed.43
Similarly, the US Food and Drug Administration (FDA) is beginning to integrate “real-world evidence” into its evaluations—not just data from a controlled FDA trial but also ongoing evidence about the effectiveness of treatments. For example, the next generation of the National Evaluation System for Health Technology, which regulates medical devices, will include data from clinical registries, electronic health records, and medical billing claims.44
Collectively part of a movement toward “open science,” these initiatives illustrate the fourth investment strategy governments can use to spin in new capabilities: a “platform” investment strategy. In the commercial space, Slack Fund offers one such example of a corporate venture fund’s effort, creating an US$80 million fund to seed development of services on the Slack platform, which attract and retain users (from which Slack gains additional value).45 Similarly, a public sector platform strategy makes investments in structures that increase the underlying value of the broader system—what the United Kingdom’s Nesta Innovation Foundation describes as “government as impresario,” enabling data-sharing and innovation in more nascent areas.46
“It’s hard to collaborate or innovate without common platforms, and government has a role in creating the context and ability to innovate,” says Brookings Metropolitan Policy Fellow Adie Tomer.47
Indeed, data about user satisfaction, technical performance, and interactions allows tech companies to thrive; what if cities, which have an enormous user base but fewer ways to check their pulse, had a platform for the same kinds of data?
Cities can help orchestrate the platforms that use IoT data to make smart cities a reality. Mobility-as-a-service creates this kind of integration in smart transportation systems—such as the one in Cascais, Portugal, which brings in data from telecom companies, other municipalities, and multiple modes of transit.48 The system is expanding to include a single-payment platform, to ease operability between these services.
Aggregating this data will involve partnerships with both private app providers as well as public data sources. But it will also create the context for greater private sector innovation—making it easier for new players to surface insights about civic needs, and then facilitate the integration of these solutions within the broader system.
Advances in technology have dramatically raised the level of service government can provide, but they require disciplined strategies to effectively adapt to a public sector context—making these four spin-in strategies the next generation of PPPs.
