Most SBIR Phase I Winners Never Get Phase II — Here Is Why

For every three small businesses that win an SBIR Phase I award, roughly two will never receive Phase II funding. That ratio has held remarkably steady across decades of program data, even as total SBIR appropriations have climbed past $4 billion annually. The gap between Phase I and Phase II is not primarily a funding constraint. It is a design feature — and a filter that most first-time awardees underestimate until they are on the wrong side of it.

The SBA's own performance benchmark system codifies this expectation. Companies that accumulate 21 or more Phase I awards across a five-year window must maintain a Phase I-to-Phase II transition ratio of at least 0.25 — one Phase II for every four Phase I awards. Companies with 51 or more Phase I awards face a stricter threshold of 0.50. Fall below these benchmarks, and SBA bars you from submitting new Phase I or Direct-to-Phase II applications for a full year, starting from the June 1 assessment date.

Those benchmarks set a floor, not a ceiling. The actual conversion rates, and the reasons they vary so dramatically by agency, reveal what separates companies that treat Phase I as a launching pad from those that treat it as a finish line.

The Numbers Behind the Drop-Off

Conversion rates differ by agency because each agency structures the Phase I-to-Phase II pipeline differently.

At DOD, the raw numbers look deceptively strong: an average of 938 Phase II awards per year against 1,440 Phase I awards, a ratio of roughly 65%. But that figure reflects the total award counts across the department, not the percentage of individual Phase I awardees who successfully transition. DOD's topic-based solicitation model means some Phase I contracts are explicitly designed as competitive down-selects, where only a fraction of Phase I performers are invited to propose for Phase II. The Navy, which accounts for over 50% of all DOD SBIR Phase III commercialization dollars, runs some of the most aggressive transition programs — and still sees significant attrition between phases.

At NIH, the conversion rate has been declining for years. Data from the NIH Data Book shows the rate of successful Phase I-to-Phase II conversion fell from 31% to 25% over the past decade. Phase II application success rates hovered between 30% and 31% in FY 2020 and FY 2021, while the number of regular Phase II submissions dropped from 665 to 585 over the same period. Meanwhile, Direct-to-Phase II applications surged from 588 to 851 between those years — but their success rate fell from 19% to 15%, suggesting that the alternative pathway is attracting more applicants without proportionally more funding.

At NSF, the math is straightforward. The agency makes approximately 230 to 235 Phase I awards per year and roughly 85 to 90 Phase II awards. That is a transition ratio of about 37% — higher than NIH's recent numbers, but still meaning nearly two-thirds of Phase I winners do not advance. NSF's Phase II success rate of approximately 20% (the highest among all 11 participating agencies) partly reflects the agency's invitation-based model, where Phase I awardees receive specific guidance before proposing.

The DOE averages roughly 170 Phase II awards per year. With Phase I awards numbering in the hundreds, DOE's conversion rate sits in a similar range — constrained by the agency's emphasis on technology readiness and its relatively short Phase I performance periods (often six months), which leave companies less time to generate the technical results reviewers want to see.

Why Phase I Success Does Not Predict Phase II Funding

The fundamental mistake most Phase I awardees make is assuming that Phase II is a continuation. It is not. Phase I sells the promise of an idea. Phase II sells the path to impact and revenue. The evaluation criteria shift substantially, and proposals written as if Phase II is simply "more Phase I work with a bigger budget" fail at disproportionate rates.

Three structural differences drive this gap.

The commercialization plan becomes a make-or-break section. In Phase I, the commercialization plan is often a two-page afterthought — a sketch of potential markets and a vague path to customers. In Phase II, agencies treat it as a primary evaluation criterion. The USDA requires detailed descriptions of commercial applications, intellectual property strategy, revenue assumptions, and company growth plans. NSF mandates four specific sections: Market Opportunity, Company/Team, Product/Technology and Competition, and Finance and Revenue Model. DOE's Phase II commercialization plan specifications run to dozens of pages of guidance.

A common failing is treating the commercialization plan as a theoretical exercise. Reviewers at the Phase II level want evidence of traction — letters of intent from potential customers, memoranda of understanding with strategic partners, or data from pilot deployments. As the grant consulting firm ScienceDocs notes, generic market descriptions and vague go-to-market timelines are among the most frequent reasons strong technical proposals lose during Phase II review.

Technical results must be concrete, not promising. Phase I proposals trade on feasibility arguments and preliminary data. Phase II proposals must demonstrate that the Phase I work actually produced results. If your Phase I did not demonstrate or indicate feasibility for the proposed concept, the Air Force Research Laboratory states plainly: "no further research for that particular effort will be awarded." This seems obvious, but a surprising number of Phase I awardees arrive at the Phase II proposal deadline without having completed their Phase I milestones — delayed by hiring, equipment procurement, or scope changes that consumed months of a six-to-twelve-month performance period.

The budget must match the technical plan. Phase II awards are substantially larger than Phase I. NIH Phase II awards typically reach $1 million over two years, with a hard cap of $2,095,748. NSF awards up to $1.25 million. Army Phase II contracts can reach $2 million over 12 to 18 months. DOE funds up to $1.1 million. With this increase in scale comes increased scrutiny of every budget line. Reviewers check whether labor rates match the stated team, whether subcontractor costs align with the proposed technical plan, and whether the overall budget is padded or underscoped. Either error signals poor planning.

The Letters That Actually Move Reviewers

Letters of support have become a quiet but decisive factor in Phase II evaluations, and most applicants handle them wrong.

The standard advice is to include three to five letters for Phase II proposals. NSF Phase II specifically requires at least three letters from potential customers or partners to validate market demand. But quantity without specificity is worthless. A vague endorsement from an academic collaborator praising the underlying science does nothing for a commercialization score. The letters that move reviewers come from people who would actually buy or deploy the technology — and say so in concrete terms.

The strongest letters reference a specific problem the writer's organization faces, describe how the applicant's technology could address it, and express a clear intent to evaluate, pilot, or purchase the solution. A letter from a hospital system's chief technology officer stating that they would participate in a six-month pilot of a diagnostic tool at three facilities carries more weight than a letter from a department chair at a medical school affirming the importance of the research area.

One often-overlooked detail: some agencies have changed their policies on letters. NSF no longer allows generic letters of support unless required by a specific program solicitation but will accept letters of collaboration documenting committed partnerships. Applicants submitting standard support letters under outdated assumptions risk having them disregarded entirely.

The Program Manager Conversation Most Applicants Skip

Across every agency, the single most underutilized resource available to Phase I awardees is direct communication with their program manager — and the data suggests this oversight contributes meaningfully to conversion failures.

At DOD, the Navy's Phase II guidance explicitly states that applicants should inform their SBIR/STTR Program Manager before submitting a Phase II proposal. This is not a bureaucratic formality. Program managers have visibility into which technical areas are receiving continued investment, which topics will be re-solicited, and where the agency sees transition opportunities. A fifteen-minute call before drafting a Phase II proposal can reveal whether the technology direction you pursued in Phase I still aligns with the agency's priorities — or whether a pivot in framing could strengthen the proposal.

At NIH, program officers at individual institutes can clarify whether a Phase II proposal falls within their scientific scope, suggest relevant study sections, and flag potential conflicts with existing funded projects. Grant Engine's analysis of NIH SBIR timelines notes that the Phase II application window is already compressed — typically three standard receipt dates per year — and waiting until the last cycle to discover a scope mismatch wastes an entire year.

At NSF, the relationship is even more structured. NSF assigns each Phase I awardee a Program Director who provides guidance through the Phase II application process. The agency's Phase II solicitation (NSF 24-580) reinforces that Phase II proposals are expected to build on the specific Phase I results and the program director's assessment of those results.

The companies that convert consistently treat this engagement as a standard part of their Phase I execution, not an optional last step.

Fast-Track and Direct-to-Phase II: When to Skip the Line

Two alternative pathways exist for companies that want to compress or bypass the sequential Phase I-to-Phase II timeline, but each carries trade-offs that are poorly understood by many applicants.

Fast-Track applications combine Phase I and Phase II proposals into a single submission, reviewed together with a single rating. The NSF Fast-Track pilot (NSF 24-582) requires applicants to have received a Project Pitch invitation from an NSF Program Director within the four months preceding submission, and to have received NSF research funding within the prior five years. The advantage is speed — a single review cycle rather than two sequential ones. The disadvantage is risk concentration: a weak Phase I technical plan sinks both phases simultaneously, and the proposal must demonstrate both feasibility and a fully developed commercialization strategy upfront, before any Phase I results exist.

NIH offers a different mechanism: Direct-to-Phase II, which allows companies that have already completed Phase I-equivalent work through other funding sources to apply directly for Phase II awards without a prior SBIR Phase I. This pathway surged in popularity — applications rose from 588 to 851 between FY 2020 and FY 2021 — but the success rate dropped correspondingly, from 19% to 15%. The pathway is not available for the STTR program or at CDC, FDA, and ACL.

For most first-time SBIR applicants, the sequential path remains the stronger bet. Fast-Track makes sense when you have existing NSF relationships and strong preliminary data. Direct-to-Phase II makes sense when you have already built a working prototype through angel funding, venture capital, or other grants and want to scale with federal support. Neither is a shortcut — both require the same level of proposal quality as their traditional counterparts.

The Timeline Trap

The most insidious obstacle to Phase II conversion is not proposal quality — it is calendar management during the Phase I performance period.

A typical Phase I award runs six to twelve months, depending on the agency. The Phase II proposal is due during or shortly after that period. At DOE, Phase I performance periods can be as short as six months, leaving companies scrambling to generate results, write a Phase II proposal, develop a commercialization plan, secure letters of support, and manage the administrative requirements of Phase I reporting — simultaneously.

The companies that convert at the highest rates front-load their Phase I execution. They begin customer discovery and commercialization planning in month one, not month six. They draft the Phase II technical plan while Phase I experiments are still running, using preliminary results to shape the narrative. They request letters of support early, giving partners weeks rather than days to respond. And they schedule their program manager conversation in the first quarter of the Phase I period, when there is still time to adjust direction based on feedback.

Companies that wait until Phase I results are "complete" before starting Phase II planning consistently miss the window or submit rushed proposals. The Phase I period is not just a research window — it is a Phase II preparation window running in parallel.

What Repeat Winners Do Differently

The SBA's benchmark data reveals an implicit truth about the SBIR program: a relatively small number of companies convert consistently, while a large number win Phase I once and never return.

Companies that maintain conversion rates above the 0.50 threshold — one Phase II for every two Phase I awards — share several patterns visible in the SBIR.gov awards database. They tend to apply to multiple agencies, spreading technical risk across DOD, NIH, NSF, and DOE rather than depending on a single agency's budget cycle. They build dedicated proposal teams rather than assigning grant writing to the PI as an afterthought. They treat Phase I deliverables as Phase II proposal inputs from day one, structuring Phase I milestones to produce exactly the evidence that Phase II reviewers will demand.

Most critically, they invest in commercialization infrastructure before Phase II requires it. By the time their Phase II proposal is due, they have already identified beta customers, filed provisional patents, and built relationships with potential acquisition or licensing partners. The commercialization plan is not a section they write for the proposal — it is a document they are already executing.

The SBIR program awarded over $4.2 billion in FY 2022 across 6,308 awards. The companies that capture a disproportionate share of Phase II and Phase III funding are not necessarily building better technology. They are building better organizations around that technology — ones that can demonstrate, in 25 pages or fewer, that the path from prototype to product is not hypothetical.

Granted tracks SBIR opportunities across all participating agencies, so you can find the Phase II solicitation that matches what your Phase I results actually proved.