Technically excellent SBIR proposals get declined every cycle. When the debrief arrives, the reason is rarely the science — it is a commercialization plan that read like an afterthought written the night before the deadline. The statutory purpose of the program is moving federally funded R&D into the market, and reviewers are instructed to score commercial potential alongside technical merit; at some agencies the two carry equal weight.
Engineering founders over-invest in the technical volume and treat the commercialization section as marketing filler. That is backwards. The technical volume proves you can build it. The commercialization plan proves that building it matters, and it is the section where most applicants are easiest to beat.
What the reviewers are actually scoring
Strip away the agency-specific formatting and every commercialization plan answers five questions:
- Who has this problem, how many of them are there, and what does it cost them today?
- What exists now, and why is it insufficient?
- What is your product, what does it cost to make, and what will someone pay?
- How does it reach the customer — direct, distributor, licensee, OEM?
- What happens after the award money runs out?
Question five separates funded from declined. The reviewer is asking whether a credible path runs from your feasibility result to a company that sells something. A plan ending at "then we will seek Phase II funding" has answered nothing, because Phase II is still government money. Our breakdown of how federal grant review panels score proposals covers the scoring sheet; this is about filling its commercial half credibly.
Size the market from the bottom up
The fastest way to lose a reviewer is a top-down market number pulled from an industry report. "The global sensor market is $180 billion and we need only 0.1 percent" tells a reviewer you have not thought about who buys.
Build it from the unit instead. How many US hospitals run the procedure? How many procedures each per year? What share are candidates? What is your price? Multiply. The number will be smaller and far more persuasive, because every input is something a reviewer can argue with — which means it is something you can defend.
Name the segment you sell to first, narrow enough to reach with the money you will actually have. A serviceable market of $40 million you can describe customer by customer beats a $12 billion total addressable market you cannot. The same discipline that makes a business plan for an invention useful applies here, compressed into a few pages.
Show that you have talked to customers
Reviewers can tell a founder who has interviewed thirty buyers from one who has read about the market. The tells are specific: real objections, real pricing pushback, real workflow constraints, a real name for the person who signs. The strongest evidence costs nothing:
- Letters of support from prospective customers, distributors, or clinical partners — most agencies allow them and many reviewers look for them first
- Quotes from customer discovery interviews, attributed by role rather than name
- Evidence of a pilot, LOI, or paid evaluation, if one exists
- Competitive analysis naming actual products
Letters of support are the highest-return item in the proposal relative to effort. Nobody mistakes one for a purchase commitment, but a specific letter saying "we would evaluate this in our operating room, and here is why our current device frustrates us" carries weight a generic form letter never will.
Competition: name names
Writing "there are no direct competitors" reads as either arrogance or ignorance, and reviewers treat it as both. Every problem has an incumbent solution, even if the incumbent is a manual process, a spreadsheet, or doing nothing.
Build a table. Rows are competitors, including the status quo. Columns are the dimensions your buyer cares about — cost, accuracy, throughput, regulatory status, integration effort. Fill it honestly, including cells where a competitor beats you. A plan that concedes two disadvantages and explains why the buyer trades them away is far more convincing than one that claims to win on everything. The method is spelled out in our guide to running a competitor analysis for a physical product.
The financing and revenue path
This is where you answer question five. Lay out what happens after the award in concrete terms:
| Stage | What it costs | Where the money comes from |
|---|---|---|
| Phase I feasibility | Roughly $50,000-$300,000 depending on agency | This award |
| Prototype and validation | Often $500,000-$2 million | Phase II, state match, strategic partner |
| Regulatory and certification | Highly variable by product class | Phase II, private round, revenue |
| Tooling and first production | Often $100,000-$800,000 for hardware | Private capital, purchase order financing, revenue |
Show the transition. Some agencies run commercialization readiness or matching programs that reward companies bringing in outside investment or a strategic partner, and mentioning that you understand those mechanisms signals seriousness. Structuring this without creating conflicts is covered in combining federal grants with private investment.
Financial projections should be a modest table, not a hockey stick: three to five years of unit volumes, price, gross margin, and headcount. If your margin assumption is 65 percent, say where the cost of goods number came from. Our guide on building financial projections investors trust covers the same discipline in an investor context.
Intellectual property
Address IP explicitly: what you own, what is filed, what is licensed in, and what the award will generate. Under Bayh-Dole a small business retains title to inventions made with federal funding, subject to government rights, and SBIR data rights protect your technical data. Details are in who owns the IP from a federal grant.
Practical drafting advice
- Write the commercialization plan first, then the technical volume. It forces the research questions to serve a commercial outcome.
- Match the technical milestones to commercial decision points. A Phase I task list that produces exactly the evidence a customer or a Phase II reviewer needs is a coherent plan.
- Reuse and adapt, but read the specific solicitation. Agencies differ considerably in format and emphasis — the differences by agency and by phase are outlined in SBIR Phase I versus Phase II.
- Have someone outside your field read it. If they cannot explain who buys and why after one pass, neither can a reviewer at midnight on their fourth proposal.
Projects House helps founders turn a technical concept into the cost, timeline, and manufacturing detail that a commercialization plan needs — realistic BOM costs, tooling estimates, and a development schedule that survives contact with a reviewer. If you are drafting an SBIR application and need the engineering side to hold up, reach us through the contact form.