Emerging tech is transforming war in real-time, yet the biggest DOD incubator is stuck in the past
The DOD SBIR, America's Seed Funding program, is currently not a useful mechanism for innovation. It can be fixed.
Note: A lot has been said about the positives and negatives of the DOD SBIR, much of it at the strategic level (PPBE, valley of death). What follows is a private sector operator perspective with tactical considerations on why many emerging tech companies do not or should not consider SBIRs. OUSD(R&E) is working on reform, which will likely impact some of the following issues.
TL;DR:
Economic, geopolitical, and cultural events are reversing a decades-long separation of much of Silicon Valley and America’s National Security community. With that shift, an influx of emerging technologies will need to be matured and/or transitioned for DOD use cases. The Small Business Innovation Research (SBIR) grant — self-labeled as “America’s Seed Funding” — is a key means of early-stage funding for the DOD.
In theory, SBIRs are a fantastic on-ramp to the dual-use world. In reality, the existing SBIRs reinforce many reasons many small tech companies and VCs avoid working with the DOD.
The actual odds of receiving an award favor the few and are terrible for new companies trying to break into the DOD.
To boost odds, companies have to hire third-party advisors, which impacts margins and financial exposure.
The economics of receiving a grant does not financially justify the level of effort required to get an award.
The process is variable, byzantine, and usually a bridge to nowhere.
Without reform at a tactical and strategic level, SBIRs will unlikely drive the next generation of innovation between DC and the tech sector.
A couple of tweaks could help improve SBIRs right away. First, for the love of Hap Arnold, the DOD needs to either (a) stop selling the idea of SBIRs as the on-ramp to early tech companies or (b) dedicate the majority of resources to emerging tech companies, not a handful of organizations. Second, the DOD should strongly consider killing Phase I, focus energy and resources on direct-to-Phase IIs and increase the minimum award to $2M. Finally, the DOD should reduce point-solution solicitations in place of multi-quarter and multi-year thematic streams of funding so emerging companies can align tech and business strategies.
Background: As the world has watched the Russian invasion of Ukraine in real-time, three (among others) lessons have been laid bare to (re)learn.
A small number of powerful totalitarian regimes will continue to be some of the biggest threats to the Western world as we know it.
Emerging technologies and innovations can shatter platforms that cost orders of magnitude more to build and operate.
The US tech community has a duty and a moral role to play in supporting the DOD and countries fighting these regimes and injustices.
The lesson of morality and duty will be the underpinning of a dramatic culture shift away from the low point of relations between Silicon Valley and Washington experienced just a few years ago. In combination with economic and investment uncertainty, the DOD is poised to see a dramatic influx of companies wanting to sell to and innovate with the DOD over the next 5-10 years. A pendulum swing like this is exciting to many, but it also means the DOD needs to be ready to procure and partner with new companies at a scale that they haven’t in decades.
A primary tool for emerging tech to partner with and transition to the DOD is the SBIR. With over $1B each year that is available for DOD small businesses, it is the most important tool in the DOD innovation arsenal. Unfortunately, at its current course and speed, it is unlikely to fulfill its potential as a vehicle for emerging technology innovation. Why?
1) The award odds are terrible for most companies
Let’s look at the data. From 2016 to today (5/3/2022), there were 15,664 SBIR awards going to 3,925 companies totaling $7.7 billion dollars. That’s fantastic! However, over this six-year horizon, a deeper dive shows a snapshot of who is actually making all of this money — very few companies. How few? Of the 3,925 companies that received $7.7 billion dollars in funding:
8 companies received 10% of the total awards
1% of companies received 25% of the total
5% of companies received 50% of the total
10% of companies received 61% of the total
Voila! The DOD SBIR power curve is alive and well. Want to look at all the data? All are in a google spreadsheet here. Nerd out.
Now let’s look at the Top 25 companies. Over the last six years with the SBIR program, they’ve made an absolute killing — $1.56 billion. These companies win because have built processes and teams to specifically win SBIRs. They develop networks in the DOD and they use them. They undoubtedly serve a need for the DOD, especially in the cases of sensitive technologies. At worst, these companies are, described by SSTI, “SBIR mills.” At best, we should at least question if the DOD is actually “seeding American innovation” or supplementing small prime contractors and a select group of businesses?
Have many of the well-known, VC-backed defense-focused startups progressed through the SBIR program? Anduril has been awarded $12M, almost all from a single contract associated with Will Roper’s ABMS program in 2020-21. Others? Shield AI ($2.3M), Scale AI ($850K), Snorkel AI ($100K), Vannevar Labs ($850K). All of these SV companies have received substantially larger contracts from other organizations in the DOD, which if anything, further highlights the issue at hand. How all of these companies (including Primer AI), found success with the DOD through other deals is another post by itself? The short answer is seasoned defense veterans, strong VC and insider presence, hustle, and duct tape. The bigger question is that if emerging high-tech companies with a substantial focus on DOD aren’t getting more of the pie, is it too small, too hard to win, or too connected to incumbents? No matter the answer, there is a problem.
2) To boost SBIR odds, companies pay
Setting aside the small subset of firms taking the majority of the funds, there’s still a decent of money on the table for emerging companies, so who else is winning? Companies that pay to play.
A niche market has developed for companies willing to pay to win SBIRs. Ex-DOD procurement officers and people that know the process act as third-party advisors to facilitate SBIRs — they fill out forms, make introductions, and find potential sponsors. The goal is to match-make with organizations, solicit Letters of Support (Phase I), and Memorandums of Understanding (Phase II). In return, advisors receive payments in the form of monthly expenses ($3-5K/mo) and/or a percentage of awarded deals (5-15%).
To some, this group seems like yet another bureaucratic layer, but until larger reform occurs, these organizations play a role. Their support greatly increases the likelihood of receiving an award (at least a Phase 1), they facilitate, help demystify, and frankly, make it easier to connect tech to end-users and make it easier for DOD procurement to review applications. Moreover, they’re successful.
From a contribution standpoint, anecdotal information from multiple sources says we should comfortably assume 20 –30% of companies awarded DOD SBIR contracts use these third-party advisors. If we combine the 60% of business going to the top 10% of the companies, with the 25-30% of awards for companies that use these services (unlikely they overlap), 80-90% of the SBIR pie is now spoken for. So, there’s not a lot left if you are not in one of these two categories. Would it still make sense to take a shot because of tech alignment the SBIR contract size? Doubtful.
3) The current economics likely doesn’t make sense
Most total awards for Phase I ( $50k) and Phase II ($750k) are the same across organizations. The awards for Phase I are to be used as a business development seed for Phase II, which focuses on tweaking products and technologies for DOD use.
*A simple unit economics planning tab in the spreadsheet. Feel free to download and modify.
Employee compensation, product technology readiness level (TRL), and DOD experience vary per company, but for most, they will either exceed the $800k spread out over 2.5 - 3 years or provide anemic resources to stay on budget.
Contracting a third-party advisor will undoubtedly increase the likelihood of receiving an award, but the advisors receive payment both as a monthly expense ($3-5K/mo) and a percentage of awarded deals (5-15%). Therefore, funds will be exhausted even faster when third-party advisory services, increase the likelihood of overall loss.
So what is the magic breakeven point? A guesstimate is closer to $2M for the combined Phase I and Phase II to cover the vast majority of expenses for a real team of engineers (4-5) and project managers (1-2) to take on the technology development and transition challenges. What’s the old adage? Do it right or don’t do it at all. *This number is closer to some concepts being proposed with SBIR reform. Tinker with the spreadsheet to explore on your own.
What about the bigger dollars for Phase III? Yes, pretty much all VC-backed startups, build products and customer relationships at a loss in the early years for future scaled success. Phase I and II could be a path to Phase IIIs and greener pastures, so a loss in Phase I and II would be justified. However, the number of awarded Phase III contracts is likely very, very small, so it’s not included in the decision calculus.
Economics and odds aside, any other major barriers? Yup.
4) The SBIR process itself is too variable and a bit unreliable
The DOD has highly trained tacticians and logisticians in the world, yet the award process and receiving funds is notoriously late and inconsistent (see GAO report, chart below). Late award notices and awards can create resource planning issues or could impact actual working capital levels at awarded companies. Not going to sugarcoat this — at best it makes the DOD seem like an unreliable partner, at worst it can stunt or shutter businesses trying to support the DOD.
Aside from late In a perfect SBIR scenario, going from Phase I to Phase III would take 2-3 years. In reality, that may take 4 - 5 years with delays in awarding funding, the future availability of funding, and wholesale changes in priorities. DOD personnel get reassigned and programs get absorbed all the time with the change in administrations and appointees.
What it means is that companies have very little ability to plan and resource in the near term for project execution, and mid-term sales plans are influenced by people that will rotate into a program if it still exists at all. Couple that with the low likelihood of winning a Phase III, and many emerging tech companies may execute perfectly, only to see funding dry up and be reallocated, and they (and their investors) are left holding the empty bag.
What can be done?
Decide upon the goal of DOD SBIRs and market it accordingly. If it should be mainly for a few repeatable partners, so be it, but don’t sell it differently to emerging tech companies, Silicon Valley and VCs. Alternatively, the DOD could consider a different classification or funding stream for companies that constantly submit and win SBIRs.
Strongly consider killing Phase I, focus energy and resources on direct-to-Phase IIs, and increase the minimum award to at least $2M. Given the boost in submissions and actual costs of development or transition. “Focus on picking winners.” — (stolen from Chris Brose).
Reduce point-solution solicitations in place of multi-quarter, multi-year, cross-service thematic streams of funding so emerging companies can align tech and business strategies. This starts to slip into PPBE reform, but even without going there, it is more than possible for SBIR organizations to layout areas of interest over single and multiple years to help guide company focus, resources, and future financial plans.
These recommendations are tactical in nature, to highlight changes that probably don’t require an act of Congress (I could be wrong) to enact. If enacted, SBIRs would be a more effective innovation vehicle for the DOD and emerging technology companies, and our National Security community and allies will be better prepared and better protected for the wars and crises that are happening now, and those to come.