SBIR — an Entrepreneur’s Review

The Small Business Innovation Research (SBIR) program is a flagship innovation program by the Federal government for small businesses — it awards around $2.5B in contracts each year. As an entrepreneur who worked in the Federal sector for a decade, won several SBIRs, and has now left the sector, I thought I might be able to provide some useful insight into the program. Both for “govies” (people who work in the government) and for fellow entrepreneurs.

A quick note on me. I’m a Brit, though now have my green card (kind of a prerequisite for SBIR — you have to count as a US resident). I have never worked in the government (though I respect the people who do), and my first company just kind of ended up there based on the product and investors. I have been a subject matter expert speaker at NSA, DIA, Air Force, Navy and international defense conferences. I have created two companies that have sold into the government community — BlueSpace and Hypori — and personally sold to most of the US Federal apparatus. BlueSpace was majority owned by British nationals (I moved it here to the US after founding in the UK), so it wasn’t eligible for SBIR, though I won some Broad Area Announcements (BAAs) that are similar to SBIRs. Hypori has had multiple SBIR awards from the US Air Force and another from DHS, as well as other government and Fortune 500 contracts.

As one final caveat, I apologize for the severe lack of expletives and deplorable shortage of humor in this post compared to other Medium articles I have read. I actually went through it at the end and tried to spice it up a bit more. And maybe I need to stop thinking in bullet points so much. But I do believe in the content.

What is SBIR?

The Small Business Innovation Research (SBIR) program was started in 1982 with four original objectives:

  1. To stimulate technological innovation;
  2. To use small business to meet Federal research and development needs;
  3. To foster and encourage participation by minority and disadvantaged persons in technological innovation; and
  4. To increase private sector commercialization innovations derived from Federal research and development.

Each government agency with R&D budget over $100M has to participate in SBIR, and it is budgeted as a “tax” on their R&D budgets — 3.0% in 2016. That doesn’t sound like much, but if you are the DoD, it ends up being a lot, because things like fighter jets cost an immense amount of money to build. What’s more, SBIR budgets have been protected by the Congressional set aside through sequestration (a nasty budget freeze that hamstrung much of the DoD into solving yesterday’s problems). The tax percentage has been rising each year (increasing the SBIR budget), and competition for SBIR dollars has typically been less intense than for other funds over the past few years (from my personal conversations with govies).

Each Federal entity puts out SBIR topics — things they would like research on / problems they want solved. These are typically pretty specific, and can cover everything from IT systems to new laser weapons to carbon nanotubes to sampling drills for Mars rovers to … pretty much you name it. There’s also a related Small Business Technology Transfer Program (STTR) that is smaller and designed to foster relationships between industry and academic institutions.

SBIRs have three phases:

  1. Phase I: Startup — an award of up to $150K for 6–9 months of exploration / feasibility studies on an idea or technology. I have seen DoD organizations typically award 2 to 4 of these per topic, and they typically get 30 to 80 responses per topic (though many of these are non-starters). The proposal requirements for a Phase I are pretty low (like a 5 page “white paper” with a basic cost table), and Phase I projects are typically awarded on a Firm Fixed Price (FFP) basis.
  2. Phase II: Development — up to $1M over up to 2 years to make it work. This is typically one per topic, but sometimes you can see two awards. In the vast majority of cases, the winner of a Phase II will have already had a Phase I, though in 2014 DoD and NIH were allowed to go straight to Phase II. Also it is possible for another agency to award a Phase II with its own budget on the back of an SBIR award from another agency. Phase II’s are typically awarded on a Cost Plus Fixed Fee (CPFF) basis, though they can very occasionally be awarded on an Other Transaction (OT) basis too. More on this later.
  3. Phase III: Transition — no additional SBIR funding, but typically a very useful contract. It is a sole source contract (for both product and services) typically with a $49M ceiling that any government organization can buy from like buying out of a catalog. In government contracting land, this is a golden ticket that could make Charlie jealous.

If one or more Phase I’s are awarded, the government customer organization has budgeted for Phase II. If you win a Phase II, you are almost guaranteed to get a Phase III (unless you really screw it up). Note that there are some Phase II extensions and accelerated Phase III awards etc, but these are rare.

Some other things to know:

  • Like all funded research in the government, the government mandates “Government Purpose Rights” (GPR) in the inventions / results. This can be thought of as a free site license for what they funded. This isn’t a deal breaker — more on this later.
  • The company has to be majority US owned — they include green card holders in this definition.
  • The company cannot be majority owned by investors — whether they are private equity, venture capital, public markets, whatever.
  • The company (obviously) has to be small. But small is a relative term that the government varies by sector — see this table. Broadly, if you are reading this and your company isn’t public or a unicorn, you are probably small.
  • You can have subcontractors, but the company that wins has to do most of the work.
  • Cost Plus Fixed Fee contracting requires a company to formally evaluate its cost base, which means tracking everything you spend and every hour anyone spends doing anything — so it can all be classified appropriately. The cost build up takes base salaries and then adds on overhead, G&A, etc to come to a rate per hour — it’s an accounting exercise — and that rate plus a profit of 5% to 10% is what you charge the government.

Here ends the summary, congrats for reading or skimming this far. Now on to my opinions in glorious full throttle.

Things to Like

  • Money: The SBIR program doles out a lot of it, partly why their tag line is “America’s Seed Fund”.
  • Easy Application: Compared to many government programs / contracts, it’s easy to apply, especially for the Phase I, the timeframes and processes are well known and the topics are collated so easy to browse.
  • Support: There’s tons of it. Call center helplines, accessible execs, SBIR points of contact in each government organization that does them, websites, training videos, guides, consultants, wizards, you name it, it probably exists — the program is mature.
  • Very Early: SBIR Phase I’s are designed to work right down to a sole inventor — it doesn’t matter how early you are, in fact you probably don’t even need to have an official company when you apply for a Phase I.
  • Low Bar: Depending on the SBIR office that is running the project, the expectations for output are pretty low and realistic, particularly in a Phase I. The deliverable is typically a white paper, maybe an ugly hacked up demo.
  • Transition: The SBIR program measures success in innovations transitioning into commercial use, and this is important because commercial use tends to mean longevity for the innovation.
  • Kudos: Winning and executing on an SBIR, especially for a first time entrepreneur, can help get some kudos / credibility with prospective customers and investors.
  • Further Funding: Many States have funds that focus on helping SBIR winners — Texas used to, Massachusetts has MassVentures, Indiana has Elevate Ventures, etc.
  • Successes: Corporations like Qualcomm, Symantec, iRobot and Jawbone started from SBIRs. Massachusetts claims SBIR winners in its State have generated $3.8B in sales from SBIR technologies — and that’s just Massachusetts.

Things to Dislike

CPFF Contracting: I hate this the most. Where to start?

  • Very bright people work for below market salaries in startups in exchange for stock options that they hope will end up worth a multiple of their salaries — CPFF cannot evaluate or incorporate stock options, so the math is broken in any case.
  • Due to this stock incentive, few people in startups actually work just 40 hours per week, many work 60–80 hours per week, but CPFF penalizes this extra commitment by dividing total cost (excluding stock) by hours worked to get to billable rates. The hours worked then embarrasses the contracting officer, who wonders whether you are running a sweat shop (yes, I have had this happen on a CPFF contract, though not an SBIR).
  • When you are taking investment to build something (which most innovative product companies do in tech), you are operating at a loss. The size of this loss can end up making the CPFF calculated rates ridiculously high, because there’s no way the hours worked on an SBIR can really cover the overhead including the organic product development that you raised funding to do.
  • The killer — you have to get everyone to track their time. All their time. Even people who don’t bill on the SBIR project. This is totally counter intuitive to startup culture, and the types of people attracted to startups tend to hate it — they left higher paid jobs at bigger companies to escape process bureaucracy and punch in / punch out time sheets.

Other things to dislike:

  • Contracting Choices: The term “agile contracting” is kind of like an inside joke or comical oxymoron in government circles, a bit like “accreditation reciprocity”. Being a government contracting officer is a tough job because you have few choices and “fairness” oversight that forces unnatural behaviors. Creativity means risk. “Other Transactions” (OT) can be hard to justify or structure in contracting, in part because you can’t tell how much work the supplier is having to do — are they just doing a few tweaks and charging you a ton? “Time & Materials” (T&M) is used in many other contracts but not typically in SBIR — I believe the main concern here is that the supplier could be charging a high hourly rate for a junior employee, which would be bad value for money.
  • Topic Specialization: A lot of the topics are incredibly precise and specialized — for example, “Adaptive Training System for Maintaining Attention during Unmanned Aerial Systems (UAS) Operations”. Outside of photo fodder for TV show caption competitions (“electric shock via joystick”?), it’s hard to see a broad market for this. If you invented bullets that bend round corners and only kill bad guys, you would not get an award under an SBIR contract (hint: if you have one of these, go straight to SOCOM). You could suggest a topic, but that’s hard and can take a year or more for it to show up in an SBIR topic announcement.
  • Transition Failure: When topics are extremely narrow and defense specific, especially when used to prop up funding gaps in extremely well defined existing Programs of Record (POR), transition is likely to fail, and this is kind of taking the SBIR program captive for purposes that it was never intended to solve.
  • Closed Ecosystem: I have noticed that SBIR offices do sometimes pick companies they know for awards. I can understand this from a perceived risk perspective, but it can mean that small businesses that are generalist integrators with little domain knowledge are picked ahead of technology specialists with commercial experience in that specific domain. Why reinvent wheels?
  • Timing: For startups running on borrowed time, the SBIR program takes too long. It reminds me of the Hobbits trying to work with the Ents in the book and movie Lord of the Rings. Of course, this is kind of true across the board in the startup / government relationship — startups think in quarters (or months), governments think in fiscal years. But SBIRs can envisage a much longer timeframe for technology evolution than happens now in the real (especially commercial) world, so the thing you built is out of date before the SBIR Phase III award, or the company grows into a bonsai tree not a sequoia.
  • All Written: I still find it surprising that all submissions for SBIR’s are in writing, there are no presentations (though there is a presentation at the end of the Phase 1). Can you imagine a professional investor doing a deal without talking to / meeting with the entrepreneur? If transition is the goal, the personality and objectives of the entrepreneur will be key to success.
  • Government Purpose Rights: This is actually a much smaller problem than many of the uninitiated realize, but the fear is that the government will somehow end up with ownership of whatever the entrepreneur creates. In reality, the government exerts no rights over the use of IP in commercial, and as an entrepreneur if you know what you are doing, you can manage around GPR within Federal too (more on that below). Also many of the more enlightened govies I have worked with recognize that GPR can often become a poisoned chalice (see below).

Advice for Govies

  • Pricing: For most technology solutions that have true commercial transition opportunities, you don’t want GPR. I know you may feel like you do, but as one very senior exec at a three letter agency once said to me, “Every time we own GPR for something or built it ourselves, it ends up dying in a corner.” If you take GPR for a product, you give the vendor no long-term incentive to service the government’s requirements, whereas they have plenty for commercial customers who are paying on an ongoing basis to use the product. So then you might start a program office and try to pay for every change request, but who is going to field those change requests, who will prioritize them, who will fund them, and does the government want to become a product manager? You can negotiate a fair transition price up front with any vendor when you have the power of R&D funding like SBIR in your hand, but zero isn’t fair or practical.
  • Contracting: If you want true commercial companies that are trying to get commercial customers leveraging external investment, don’t force CPFF contracting. Please, as entrepreneurs we beg you, anything but that. Do OT, if you can, do T&M, do something else. CPFF can hamstring a startup’s agility and its ability to attract quality talent. At the end of the day, ask yourself do you want the government to be the company’s only customers? If you are building a new thruster for a supersonic missile, the answer is probably yes. If you are building a cyber security defense system, the answer is probably no.
  • Investors: I have never really understood the requirement to not be majority owned by investors — if it is a small startup, wouldn’t you prefer more of the bill to be footed by investors? Wouldn’t that be more validating that there’s really a commercial transition opportunity? I realize this is probably a Congressional issue, but there are likely joint ventures with buy-out clauses and other techniques for managing the issue if you really like a specific company and its technology.
  • Commercial Transition: If you have worked all your life in the government, face it, you probably know precious little about what is required for commercial success. That’s not in itself a criticism, it’s just the life path you have had in serving the nation. Most bright young Silicon Valley entrepreneurs would fail spectacularly if suddenly parachuted into your shoes. But I have watched confident young DoD officers storm off with brilliant plans of what industry is going to do with a technology without actually asking any unbiased members of industry for their own opinions. Of course the sales guy who is selling you the technology is going to agree with your vision — heck he’d dance on one leg and whistle the national anthem if he thought it would help. Instead, call a friend, ping an acquaintance, use a LinkedIn Mail — or don’t act surprised when your wonderful widget sits dusty on a shelf.
  • Topics: Don’t use (all of) them as add-ons to existing Programs of Record (POR) — use funded dollars for that. Provide some flexibility in the topics, just as BAAs tend to be more flexible (but BAAs are even harder for companies outside the government ecosystem to know about and respond to than SBIRs are). It’s hard to do “innovation research” if the topic is so specific it provides a blueprint or specification for the work.
  • Marketing: If you want commercial companies to reply to your topics, don’t confuse FedBizOpps with marketing. You are competing for entrepreneur mindshare with investors and other sources of capital. Yes, SBIR is non-dilutive, but if your target audience doesn’t know you exist, or if the hoop jumping is a turn off, how can you succeed? I know that the DoD is trying to “get the word out” in Silicon Valley at a macro-level, and I think Secretary Carter “gets it”. But Goldman Sachs and McKinsey don’t say, “heck all those MBA graduates know about us, and we post our job opportunities on our website” — they proactively engage. A commercial company outsourcing research would actively approach the academics or companies it thought would be most appropriate. I realize there may be some “fairness” challenges to doing this in the government, though I also know that many government research shops proactively build relationships with targeted academics. Having a press interview with a relevant trade magazine might help, appointing an “SBIR outreach” contractor could help to actively approach companies, working with a DoD liaison office in Silicon Valley to ask “do you know anyone at … company?”

Advice for Entrepreneurs

  • Managing GPR: Don’t freak out. It’s not that bad. You probably have some existing IP or you are doing investor funded work on the side. The real trick here is to find an attorney who is knowledgeable in the FARS (Google it), and then very carefully define your existing IP and what parts you are developing under contract so providing GPR in. With appropriate negotiation, you may end up being able to “deliver” a set of diffs on a code base that was your original prototype. That’s like providing a set of instructions to get from A to B without defining where A is, or giving someone two thirds of a car laid out on the ground in pieces. This strategy also prevents an integrator from convincing a customer that they can take the code and do all the work going forwards, cutting your company out. I had an integrator prime try this on one of our non-SBIR customer contracts, not having delivered full code with GPR killed the idea immediately.
  • Dealing with CPFF: Freak out. It’s horrific. In both my companies we ended up creating wholly owned Federal subsidiaries in part to insulate the rest of the company from the onerous anguish of CPFF reporting requirements. That’s probably the least painful option if you really want to build a commercial company. Enforcing a Defense Contract Audit Agency (DCAA) approved cost accounting system on your little startup will likely make you and your team / culture die a lot more inside with each passing day. Plead for OT, beg for T&M and they can tell you whatever hourly rate they want to pay. You can also put a government friendly prime contractor in front, but then that contractor will get the lion share of the funding.
  • Topic Relevance: You can get creative. I have seen awards that bear only marginal relevance to the original topic, but were interesting to the government topic owner who typically has a great deal of discretion in selecting the awardees. Remember that you are probably not a software development outsourcing house, you are a product company, and winning a contract is only useful if it takes your ball further up the field you actually want to play on.
  • Crossing the Ocean: You’ve heard of “crossing the chasm”? Well in the government, the gap is far, far wider. Typically, different people or even different organizations fund research than the ones that actually buy technologies. You can get $M in government R&D funding without installing a single license, or having any real path to do so. Tragically, I would now say that selling new products to the government is almost impossible unless you have a really long runway. You may get a few individuals excited, but if they weren’t already planning to buy a product like yours, they (or their predecessors) probably didn’t write it into their budgets. And given how long that takes your excited individuals will likely have rotated or retired before the magical Program Objective Memorandum (POM) process runs full cycle (and you’ll be bankrupt). Why am I laboring this? Well SBIRs are great sources of funded engineering and the sole source contract can help a lot in accelerating buying. But my advice here is not to confuse SBIRs with repeatable revenue — or you risk thinking and resourcing like you won when really the Ents may have only decided that you are in fact not an Orc.

If you appreciate this review, hit me up or like below. You may help convince me that writing more articles sharing my adventures in this space would be more than purely an exercise in personal therapy :-).