SBIR Budget Justification: What Makes It Competitive, Not Just Compliant

An SBIR budget justification is the line-by-line explanation of how you will spend federal funds to execute the research aims. Compliance gets you past the forms. Competitiveness decides whether reviewers believe you can actually run the project. The difference shows up in three places: your indirect cost rate, your PI effort level, and whether your line items match the story in your narrative.

Most SBIR budget guides stop at "here are the categories, here is a template." This one goes further. You will see the cross-agency budget norms reviewers use as benchmarks, the red-flag thresholds that trigger scrutiny, and the consistency audit that separates credible budgets from copy-pasted ones.

Why SBIR Budget Justification Is a Reviewer Perception Test

Reviewers read your budget as a proxy for operational maturity. They are asking, underneath the math, "Does this founder know how to run a research program with federal money?"

Three signals do most of the work:

1. Indirect cost rate. Did you pick a defensible rate, or did you plug in a number to maximize total cost? A 60% indirect rate with no negotiated rate agreement is a reviewer red flag.
2. PI effort level. Is the principal investigator visibly committed in hours, or is the budget running on contract staff while the PI is listed at 10% effort?
3. Line-item specificity. Can each cost category stand on its own with a narrative justification, or does "supplies: $25,000" sit with no detail?

A compliant budget passes the form checker. A competitive budget tells reviewers you have thought about cost like an operator, not like a template-filler.

What reviewers look for in 40 to 60 words: Budget justification that matches the work described in your aims, uses agency-appropriate indirect rates, commits the PI at credible effort levels, stays within the 33% subcontract cap, and shows line items with specific, defensible numbers. Everything else is decoration.

The Cross-Agency Budget Landscape

Different agencies expect different things from the same line item. Writing an NSF budget with NIH framing, or an ARPA-H budget with AFWERX framing, tells reviewers you have not read their guidelines.

Here is the current landscape for SBIR-style awards, pulled from Cada's per-agency playbooks:

Two things to notice. First, PI effort is defined differently by agency: NIH and NSF care where the PI is employed, DoD cares what fraction of the PI's total working hours goes to this project. They are not the same rule. Second, award ceilings vary by a factor of 40 (AFWERX Phase I at $75K versus ARPA-H base period at $5M). Your budget structure has to match the scale of the program, not just the category labels.

Indirect Cost Rates: The Single Biggest Signal

Indirect cost rates (also called F&A, for facilities and administrative) cover costs you cannot attribute to a single project: rent, utilities, finance staff, IT. Reviewers pay attention because the choice reveals how much a founder understands about cost accounting.

There are three paths for a small business:

1. Negotiated rate (NICRA). You have gone through a federal cognizant agency and agreed on a rate. This is the gold standard. Most early-stage companies do not have one yet.
2. De minimis. A federally allowed default rate on modified total direct costs for entities that do not have a NICRA. Used broadly by first-time applicants. (Confirm the current de minimis percentage with your agency program officer -- federal policy on this rate has shifted in recent years.)
3. Agency-specified default. Some programs publish a default rate you can use without negotiation.

For startups with no NICRA, the AFWERX playbook is the most explicit guidance in Cada's library:

• Fringe: 15-25% is the typical startup range, covering payroll taxes, health insurance, retirement contributions
• Overhead: 0-50%. "Conservative rates (under 20%) draw less scrutiny"
• G&A: 5-20%, covering accounting, HR, executive salaries, office costs
• Profit: 5-7%. Rates above 7% invite negotiation. Grants do not allow fee; DoD contracts do.

Reviewer inference if you plug 55% indirect with no NICRA: "This founder does not understand what an indirect rate is, or is trying to maximize total cost."

Reviewer inference if you plug 15% indirect on a reasonable direct cost base: "This founder understands they have not earned a higher rate yet and is being defensible."

The second inference wins.

Worked example (fictional). A seed-stage biotech startup applies for an NIH SBIR Phase I modular budget of $250K direct costs per year. They have no NICRA. They apply a 30% indirect rate on their modified total direct costs (MTDC, which excludes equipment under standard NIH treatment). If the budget includes $20K in equipment, MTDC equals $230K and indirect equals $69K. Total proposed budget: approximately $319K per year. The narrative justification explains that this rate reflects the company's leased lab space, two administrative staff, and shared equipment. Reviewers read that as a company that has thought about its cost structure.

When to push higher: only if you have DCAA-approved rates, an established NICRA, or a documented cost accounting system that supports a higher rate. Otherwise, conservative wins.

PI Effort: The Number Reviewers Actually Check

PI effort is one of the first numbers a reviewer checks and one of the easiest to get wrong.

The eligibility floors:

• NIH SBIR and NSF SBIR: The PI must be primarily employed at the small business (at least 51% employment at the SBC during the award period). Below this, the application is ineligible.
• DoD SBIR (including AFWERX): The PI must commit more than 50% of total project work hours to the project. Note: this is hours on the project, not employment status.
• ARPA-H: Budget must specify percent effort per role; no specific PI floor documented in the playbook.

Below these minimums, the application is automatically disqualified. Program officers check this during compliance screening before the science ever gets read.

Strategic effort levels above the floor:

• PI at 30-50% effort signals a founder-led technical project
• PI below 20% effort signals the science is delegated
• PI at 80-100% effort sometimes raises questions about operational bandwidth (who is running the company while the PI is full-time on the project?)

There is no universally "right" number. There is a right number given your story. A single-founder technical company probably has the PI at 40-60%. A CEO-CTO split might have the PI (the CTO) at 30-50% and the CEO at 5-10% supporting role.

The consistency trap. Cada's cross-proposal consistency audit flags as a BLOCKER: any case where the narrative PI effort differs from the budget PI effort by more than 25 percentage points. If your narrative says "the PI will dedicate 80% of her time" and your budget shows PI at 40%, reviewers will assume the budget is the truth and the narrative is marketing. That mismatch kills credibility faster than a high indirect rate.

Fictional worked example: A fictional robotics startup writes "Dr. X will lead technical execution" in the narrative (implying high effort), then budgets Dr. X at 15% of her time. One of two things is wrong: the narrative is overstating Dr. X's role, or the budget is understating it. Either way, the application loses credibility. The 25-point rule forces you to resolve the contradiction before submission.

Line-Item Red Flags Reviewers Watch For

The ARPA-H playbook contains the clearest published budget-norms table in Cada's library. These percentage ranges apply broadly across agency R&D budgets, even beyond ARPA-H, because they reflect how reviewers expect R&D money to flow.

A few of these deserve special attention.

Equipment above 30% of total: reviewers read this as "this company is buying equipment, not doing research." If you genuinely need a $100K piece of equipment on a $250K budget, you need a narrative that explains why this equipment is central to the aims, why renting or using shared facilities is not an option, and what specific experiments it enables. Otherwise, reviewers will flag it.

Labor under 30%: reviewers read this as "outside people are doing the work." If the budget is mostly subcontracts and equipment, the reviewer wonders whether the applicant is really the technical lead or just a pass-through for funding. This is especially damaging in SBIR, where the whole program is designed to fund small business technical execution.

Labor over 70%: the opposite problem. Reviewers ask, "Where are the materials, the testing, the equipment? Is this really R&D, or is it a salary grant?" Some software-only projects legitimately sit here, but the narrative has to justify the pattern.

Travel above 10%: almost always flagged. Most research projects do not require that much travel. The few that do (field research, specialized equipment access at another facility) need explicit justification. "Attending conferences" is not sufficient.

Subcontract Caps: The 33% Rule

SBA rules cap subcontracting in SBIR programs. The rule is simple on paper and easy to miss in practice.

SBIR rule: The primary applicant (your small business) must perform at least 67% of the work. Combined, subcontractors plus 1099 consultants cannot exceed 33% of total costs.

STTR rule: Different. Small business at least 40%, research institution at least 30%, no single category capped at 33%.

Both subcontractor entities and independent consultants count toward the 33% cap in SBIR. Many first-time applicants forget the consultant side and blow through the cap.

Two common failure modes, both flagged as BLOCKER-level in Cada's cross-proposal consistency audit:

1. Narrative-exceeds-budget: your narrative describes a named subcontractor leading 2 of 3 aims, but your subcontract budget is less than 33%. The reviewer asks, "If this subcontractor is leading two aims, why is the budget only 20%?" The answer is usually that the applicant underbudgeted to stay under the cap on paper, but the narrative gave away the real allocation.
2. Budget-exceeds-cap: your budget has 45% in subcontracts, and the narrative tries to say all the work is in-house. Compliance screening catches the budget math before the science is read.

The honest fix: resolve the allocation before writing either section. If the work genuinely requires 50% subcontract, SBIR is the wrong program -- consider STTR or a non-SBIR mechanism. If the work is 30% subcontract, write both sections to reflect that cleanly.

Equipment Over $5,000: Individual Justification Required

Across NIH, NSF, and DoD SBIR programs, equipment items costing more than $5,000 per unit require individual line-item justification. This is not optional. It is a format rule, and reviewers will flag missing justifications.

A strong equipment justification contains:

• What it is: make, model, specifications
• Why this specific model: why not a cheaper alternative, why not a refurbished unit
• Why you cannot use existing or shared facilities: what did you evaluate, why does it not work for this project
• How it is used in the aims: which specific experiments require it, how it maps to milestones

Common failure: "Required for experiments in Aim 2" with no specificity. Reviewers flag this as unsupported. If a piece of equipment is genuinely needed, the applicant can explain why. If the justification is thin, reviewers read that as evidence the applicant did not think hard about the cost.

The Narrative-Budget Consistency Test

Before submission, every SBIR application should pass a four-point consistency check. These are the BLOCKER-level mismatches documented in Cada's cross-proposal consistency audit -- the mistakes that sink applications at compliance screening or early review.

1. Every named collaborator in the narrative has a budget line. If you named a university PI, a consulting firm, or a specific advisor in the narrative, they need to be in the budget. Reviewers notice named individuals with no corresponding cost.
2. Every line item greater than 10% of total has narrative justification. Large line items without supporting narrative text are flagged as unsupported.
3. Budget duration and narrative timeline align within 20%. A 12-month budget with a 15-month project plan is inconsistent. Reviewers assume one or both is wrong.
4. Role titles and effort levels match between narrative and budget. The PI's effort in the narrative matches the PI's effort in the budget. The scientist's title in the narrative matches the title on the personnel table.

These four checks catch the majority of BLOCKER-level budget-narrative inconsistencies. Every Cada-written proposal runs through this audit before it goes out.

Worked Example: A Fictional $250K NIH SBIR Phase I Budget

Here is what a conservative, defensible NIH SBIR Phase I modular budget looks like for a fictional seed-stage biotech company. Numbers are illustrative. The company is entirely fictional.

Scenario: A 4-person biotech startup developing a novel drug delivery device. The PI is the founder-CTO. The example uses a single $250K direct-cost year to stay realistic under standard NIH SBIR Phase I modular limits.

Personnel ($125K, 50% of budget)

• PI (founder-CTO), 40% effort: $60K salary + $15K fringe (25%)
• Staff scientist, 100% effort: $35K salary (9 months) + $9K fringe
• Research assistant (TBD hire), 50% effort, starting month 3: $5K salary (3 months) + $1K fringe

Equipment ($20K, 8% of budget)

• Cell culture incubator (single item, $12K): justified with make, model, and aim-specific use
• Microfluidic flow controller (single item, $8K): justified similarly

Supplies ($55K, 22% of budget)

• Cell culture media and reagents: $20K (broken into categories: media, sera, supplements)
• Device fabrication materials: $25K (breakdown by component type)
• Consumables and plasticware: $10K

Travel ($5K, 2% of budget)

• One scientific conference (budget: $3K registration + travel)
• One collaborator visit ($2K)

Subcontract ($20K, 8% of budget)

• University of X core facility for specialized imaging: $20K
• Subcontract agreement attached, scope documented
• Subcontract percentage: 8% of total, well under 33% cap

Other direct costs ($25K, 10% of budget)

• Publication costs: $3K
• Animal per diem (Aim 2 validation): $18K
• IRB/IACUC fees: $4K

Direct costs subtotal: $250K

Indirect costs: applied at 30% of modified total direct costs (MTDC). Under standard NIH treatment, MTDC excludes equipment. MTDC base: $250K direct minus $20K equipment equals $230K. Indirect: 30% of $230K equals $69K. The startup uses the federal de minimis rate because it has no NICRA.

Total proposed budget: approximately $319K per year.

What this budget tells reviewers:

• The PI is committed (40% effort, founder role, explicit in narrative at 40% as well)
• Personnel is 50%, within the 40-60% labor typical range
• Equipment is 8%, well under the 30% red flag
• Subcontract is 8%, well under the 33% cap
• Travel is 2%, right at the typical minimum
• Indirect is 30%, within the 10-30% typical range, with the reason (de minimis, no NICRA) explainable in the narrative
• Every line item has a narrative justification

A reviewer reading this budget has no red flags to flag. The budget and narrative tell the same story. The founder looks like an operator who has done their homework.

What a Competitive SBIR Budget Justification Looks Like: Final Checklist

A competitive SBIR budget passes all of these before submission:

• Indirect rate is conservative (under 30%) or backed by a negotiated rate agreement
• PI effort meets agency minimum (at least 51% employment at SBC for NIH/NSF; more than 50% project hours for DoD)
• PI effort matches narrative within 25 percentage points
• Total subcontract plus consultant costs are under 33% for SBIR (or STTR split followed for STTR)
• Every equipment item over $5,000 has a four-element justification
• Labor is between 30% and 70% of total
• Equipment is under 30% of total
• Travel is under 10% of total
• Every named collaborator in the narrative has a budget line
• Every line item over 10% of total has narrative support
• Budget duration and narrative timeline match within 20%
• Role titles and effort levels are consistent between narrative and budget

If you can check all 12 boxes, your budget will not be the reason your application loses. Whether it wins depends on the science, the team, and the program fit. But the budget will not be what sinks it.

Get Help With Your SBIR Budget Justification

Cada's grant writers build agency-calibrated budgets as part of every proposal. Every SBIR budget justification we write passes the 12-point consistency check above before it goes out. If you are in the middle of a Phase I application and want an outside read on your budget before submission, we do 15-minute consultations with no pitch and no obligation.

Book a consultation at getcada.com.