SBIR Phase I to Phase II: The 40% Conversion Rate and What Separates Winners From One-Time Awardees

The SBIR program funds more than 4,000 Phase I awards every year. Fewer than half of those companies will ever receive a Phase II contract. According to data compiled from SBA annual reports and agency-specific dashboards, the overall Phase I-to-Phase II conversion rate across all 11 participating agencies hovers near 40% — a number that has remained stubbornly stable even as total SBIR appropriations have climbed past $4 billion per year.

That 40% figure conceals dramatic agency-level variation. It also conceals a more important pattern: the companies that convert are not randomly distributed. A relatively small cohort of firms accounts for a disproportionate share of Phase II awards, while the majority of Phase I recipients — particularly first-time awardees — never submit a Phase II proposal at all. The gap between these two groups is not primarily about technical capability. It is about organizational behavior during the Phase I performance period and the structural choices companies make long before the Phase II deadline arrives.

The Conversion Rate Is Not What It Appears

The headline conversion number — roughly 40% across all agencies — is a blended average that obscures significant differences in how each agency structures the Phase I-to-Phase II transition.

At the Department of Defense, raw award counts suggest a conversion ratio near 65%. DOD issued approximately 1,440 Phase I awards and 938 Phase II awards in a representative recent fiscal year. But those numbers are misleading. DOD's topic-based solicitation model means many Phase II awards go to companies responding to new topics rather than advancing their own Phase I work. The Navy, which accounts for over half of all DOD SBIR Phase III commercialization dollars, runs competitive down-selects where only a subset of Phase I performers receive Phase II invitations. The actual company-level conversion rate at DOD is closer to 45-50% once you control for topic cycling.

NIH tells the opposite story. Data from the NIH Data Book shows Phase I-to-Phase II conversion declining steadily over the past decade, falling from approximately 31% to 25%. Regular Phase II submissions dropped from 665 to 585 between FY 2020 and FY 2021, even as Direct-to-Phase II applications surged from 588 to 851 over the same period. The alternative pathway attracted more applicants but funded proportionally fewer — success rates for Direct-to-Phase II fell from 19% to 15%.

NSF lands near the program-wide average. The agency makes roughly 230 Phase I awards per year and approximately 85-90 Phase II awards, yielding a transition ratio of about 37%. NSF's Phase II success rate of approximately 20% — the highest among all participating agencies — partly reflects its invitation-based model, where Phase I awardees receive structured guidance from a Program Director before proposing.

DOE sits in a similar range but with a compressed timeline that disproportionately penalizes first-time awardees. Phase I performance periods can run as short as six months, leaving companies with barely enough time to generate results before the Phase II proposal is due.

The practical takeaway: a company that wins Phase I at NIH faces fundamentally different conversion odds than one that wins at DOD or NSF — and the reasons have as much to do with agency process design as with proposal quality.

Five Factors That Predict Phase II Success

Analysis of conversion patterns across agencies and fiscal years reveals five factors that consistently separate companies that advance from those that stall after Phase I. None of them are surprising in isolation. What distinguishes repeat winners is that they execute on all five simultaneously.

1. Phase I milestones designed to produce Phase II evidence. The most consequential decision in the entire conversion process happens before the Phase I award even starts — when the company defines its Phase I milestones. Companies that convert at high rates structure their Phase I work plan so that every completed milestone generates a data point, prototype result, or customer validation that can be cited directly in the Phase II proposal. Companies that treat Phase I as an open-ended exploration arrive at the Phase II deadline with interesting findings but no compelling narrative of demonstrated feasibility.

The Air Force Research Laboratory states this explicitly: if your Phase I did not demonstrate or indicate feasibility for the proposed concept, no further funding will be awarded. That sounds obvious. In practice, a surprising number of Phase I awardees arrive at the Phase II deadline without having completed their core milestones — derailed by hiring delays, equipment procurement, subcontractor onboarding, or scope changes that consumed months of a six-to-twelve-month performance period.

2. Commercialization planning that starts in month one. In Phase I proposals, the commercialization plan is often a two-page sketch — a brief market description and a vague path to revenue. In Phase II proposals, it becomes a primary evaluation criterion that can single-handedly determine the outcome. The USDA requires detailed descriptions of commercial applications, IP strategy, revenue assumptions, and company growth plans. NSF mandates four specific sections: Market Opportunity, Company and Team, Product and Technology and Competition, and Finance and Revenue Model. DOE's Phase II commercialization plan specifications run to dozens of pages of agency guidance.

Companies that treat commercialization planning as a Phase II proposal task — something to draft in the final weeks before submission — produce plans that read like theoretical exercises. Companies that begin customer discovery, partner outreach, and market validation during the Phase I performance period produce plans grounded in evidence. That distinction is visible to reviewers, and it is the single most common reason that technically strong Phase II proposals receive mediocre scores.

3. Letters of support that demonstrate market pull, not academic endorsement. Phase II proposals typically include three to five letters of support. NSF Phase II specifically requires at least three letters from potential customers or partners. But quantity without specificity is worthless.

The letters that influence reviewers come from organizations that would actually buy or deploy the technology. A letter from a hospital system's CTO stating intent to participate in a six-month pilot at three facilities carries more weight than a letter from a department chair affirming the importance of the research area. The strongest letters reference a specific operational problem, describe how the applicant's technology could address it, and express a concrete intent to evaluate, pilot, or purchase.

One detail that trips up applicants: NSF no longer accepts generic letters of support unless required by a specific solicitation. The agency will accept letters of collaboration documenting committed partnerships. Applicants submitting boilerplate support letters under outdated assumptions risk having them disregarded entirely.

4. Program manager engagement during Phase I, not after. Across every agency, the single most underutilized resource available to Phase I awardees is direct communication with their program manager.

At DOD, the Navy's Phase II guidance explicitly states that applicants should inform their SBIR Program Manager before submitting a Phase II proposal. This is not paperwork. Program managers have visibility into which technical areas are receiving continued investment, which topics will be re-solicited, and where transition opportunities exist. A 15-minute conversation before drafting a Phase II proposal can reveal whether the direction pursued in Phase I still aligns with agency priorities — or whether a reframing could strengthen the submission.

At NSF, the relationship is formalized. Each Phase I awardee is assigned a Program Director who provides guidance through the Phase II application process. The agency's Phase II solicitation reinforces that proposals are expected to build on Phase I results and the Program Director's assessment of those results. Companies that treat this relationship as a mandatory checkpoint rather than an optional courtesy convert at measurably higher rates.

At NIH, program officers at individual institutes can clarify scientific scope, suggest relevant study sections, and flag potential conflicts with existing funded projects. The Phase II application window — typically three standard receipt dates per year — is already compressed. Waiting until the last cycle to discover a scope mismatch wastes 12 months.

5. A Phase II budget that reflects genuine cost analysis. 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 12-18 months. DOE funds up to $1.1 million. With this increase in scale comes proportionally greater budget scrutiny. Reviewers check whether labor rates match the stated team, whether subcontractor costs align with the technical plan, and whether the overall budget is padded or underscoped. Either error signals poor planning — and both are more common than most applicants expect, because Phase II budgets require a level of operational forecasting that many early-stage companies have never performed.

Why Most One-Time Awardees Never Submit Phase II

The conversion rate framing implies that 60% of Phase I winners submit Phase II proposals and fail. The reality is different. A significant fraction of Phase I awardees never submit a Phase II proposal at all.

Three patterns explain this dropout.

The timeline trap. A typical Phase I runs six to twelve months. The Phase II proposal is due during or shortly after that period. At DOE, six-month Phase I performance periods leave companies simultaneously generating results, writing a Phase II proposal, developing a commercialization plan, securing letters of support, and managing Phase I reporting requirements. Companies that wait until results are "complete" before starting Phase II planning consistently miss the window or submit rushed proposals that do not reflect their actual capabilities.

Founder bandwidth. At most Phase I companies, the principal investigator is also the CEO, the business development lead, and the primary proposal writer. Phase I execution consumes the PI's technical bandwidth. Phase II proposal preparation consumes their strategic and administrative bandwidth. Without a dedicated proposal contributor or grants consultant — a resource most sub-10-person companies do not have — the Phase II proposal competes for attention with the Phase I deliverables it is supposed to build on.

Strategic misalignment discovered too late. Some Phase I results reveal that the original technical approach is not feasible, that the market assumption was wrong, or that the agency's priorities have shifted. These are legitimate reasons not to pursue Phase II. But companies that discover this in month ten of a twelve-month Phase I have no time to pivot. Companies that discover it in month three — through early program manager engagement and customer discovery — can either adjust their Phase I direction or make an informed decision to pursue a different opportunity.

The SBA Benchmark System: A Floor, Not a Ceiling

The SBA's performance benchmark system quantifies conversion expectations at scale. Companies that accumulate 21 or more Phase I awards across a rolling 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 the company from submitting new Phase I or Direct-to-Phase II applications for a full year, starting from the June 1 assessment date.

These thresholds were designed to flag "SBIR mills" — companies that collect Phase I awards without converting them into development programs or commercial products. But the benchmarks also reveal what the SBA considers a reasonable conversion floor. A 0.25 ratio means converting one in four Phase I awards. A 0.50 ratio means one in two. Both are below the program-wide average, which means the benchmarks catch only the worst performers.

The companies that sustain conversion rates above 0.50 — one Phase II for every two Phase I awards — share identifiable characteristics visible in the SBIR.gov awards database. They apply to multiple agencies, spreading risk across DOD, NIH, NSF, and DOE rather than depending on a single budget cycle. They build dedicated proposal infrastructure rather than assigning grant writing as a side task. They treat Phase I deliverables as Phase II proposal inputs from the first week of performance. And they invest in commercialization infrastructure — beta customers, provisional patents, licensing conversations — before Phase II requires it.

Fast-Track and Direct-to-Phase II: Alternative Pathways

Two mechanisms exist for companies that want to compress or bypass the sequential timeline.

Fast-Track applications combine Phase I and Phase II into a single submission reviewed together. The NSF Fast-Track pilot requires applicants to have received a Project Pitch invitation within the preceding four months and to have received NSF research funding within the prior five years. The advantage is speed — one review cycle instead of two. The risk is concentration: a weak Phase I plan sinks both phases, and the proposal must demonstrate feasibility and a developed commercialization strategy before any Phase I results exist.

Direct-to-Phase II allows companies that have already completed Phase I-equivalent work through other funding to apply directly for Phase II without a prior SBIR Phase I. NIH's Direct-to-Phase II pathway saw applications surge from 588 to 851 between FY 2020 and FY 2021, but success rates fell from 19% to 15% as applicant volume outpaced available funding. The pathway is not available for STTR or at CDC, FDA, and ACL.

For most first-time SBIR applicants, the sequential path remains the stronger approach. Fast-Track makes sense when existing agency relationships and strong preliminary data are already in hand. Direct-to-Phase II makes sense when a working prototype already exists through venture capital, angel investment, or other grants and the company wants to scale with federal support. Neither pathway reduces the quality bar.

What the 40% Rate Actually Measures

The Phase I-to-Phase II conversion rate is often cited as evidence that the SBIR program has a dropout problem. A more accurate reading is that it measures the gap between winning a feasibility study and building an organization capable of executing a development program.

Phase I rewards technical insight and a promising idea. Phase II rewards operational readiness — a team that can manage a seven-figure contract, a commercialization plan backed by market evidence, a technical roadmap that extends logically from demonstrated results, and letters from real customers who will test the product. These are organizational capabilities, not just technical ones. The companies that convert are not necessarily building better science. They are building better companies around their science.

The SBIR program awarded over $4.2 billion in FY 2022 across 6,308 awards. The firms that capture a disproportionate share of Phase II and Phase III funding have learned that the Phase I performance period is not just a research window. It is the Phase II preparation window — and companies that recognize this distinction convert at rates well above the 40% average.

Granted tracks SBIR solicitations across all 11 participating agencies, matching open Phase II opportunities to the technical areas where your Phase I results are strongest.