At any given moment, there are 300+ open SBIR topics across DoD, the Air Force, Army, Navy, DARPA, and the other participating agencies. If you're a 10-person defense company with one BD person, that number is paralyzing — and it's supposed to be. The companies that consistently win Phase I awards don't pursue 20 topics hoping one sticks. They pursue 3-5 with high conviction, deep preparation, and existing program office relationships in those specific areas.

The difference between a 40% Phase I win rate and a 5% win rate is almost entirely a filtering problem. Not a proposal writing problem. Not a technology problem. The companies losing proposals aren't writing bad proposals — they're writing competent proposals for topics they were never going to win, against incumbents or primes who had the inside track before the topic ever dropped.

Here's the filter system that works.

Diagnostic: If you're pursuing more than 8 SBIR topics simultaneously with a BD team of 1-2 people, you're almost certainly spreading too thin. A disciplined team of two people can produce high-quality Phase I proposals for a maximum of 4-5 topics per solicitation cycle. Everything beyond that is a diluted effort chasing bad-fit topics.

The 5 Signals a Topic Is a Bad Fit

These are disqualifying signals. If a topic triggers even two of these, put it in the "monitor but don't pursue" column unless you have specific intelligence that overrides them.

🚫 Reject
  • ① TRL required is above your current readiness level — you'd need to claim TRL 5 when you're at TRL 3
  • ② Program office has no history of awarding to companies your size — check SBIR.gov/awards and filter by company size
  • ③ Topic language is suspiciously specific to one vendor's existing product architecture
  • ④ Topic requires a facility clearance or special access program you don't have
  • ⑤ You have zero relationship with anyone in that program office — no prior meeting, no white paper review, no industry day exchange
✓ Pursue
  • ① Technology area is 80%+ aligned with your core IP — you could write the technical approach today without additional research
  • ② Phase I award size is in the sweet spot for your overhead rate ($150-200K for most small companies)
  • ③ Agency has a documented TACFI or STRATFI pathway for this topic area
  • ④ You have at least one prior touchpoint with the program office — a meeting, an email exchange, an RFI response
  • ⑤ The topic was preceded by an RFI or industry day where you participated — this is a strong signal the program office already knows the field

The "Wired Topic" Problem

One of the most common ways small defense companies waste BD budget is pursuing topics that were written with a specific company in mind. This is more common than the SBIR program office community likes to admit. When a program manager has been working with a company through a CRADA, a cooperative agreement, or a series of informal meetings, they sometimes write SBIR topics that, while technically open to all applicants, are structured around a capability that only one or two companies realistically have.

Signals that a topic may be "wired" for an incumbent: topic uses highly specific terminology that mirrors a known vendor's marketing language; the required deliverable is suspiciously narrow (a specific interface, a specific data format, a specific system integration); the technical requirements reference a platform or system that limits who can realistically access it; there's no prior RFI and the topic dropped without any industry engagement.

None of these signals are conclusive on their own. But two or more together should trigger a serious conversation internally before you commit BD resources to a response.

Building a Scoring System

The most rigorous small BD teams use a simple scoring rubric before committing to any topic. You don't need software to do this — a spreadsheet works. Here's the framework:

1
Map your capability vector
List your core technology areas and rate your depth in each on a 1-10 scale. Be honest. If you have a prototype in RF sensing but no deployed system, that's a 4-5, not an 8. This vector is the baseline for all topic scoring.
2
Cross-reference topic keywords against your vector
Pull the full topic description and identify the 10-15 technical keywords in the abstract and technical description sections. Count how many of those keywords map to your capability areas at a 7+ score. If fewer than 6 map cleanly, the topic is likely a poor fit regardless of the title.
3
Filter by TRL range
Every topic specifies a target TRL for Phase I deliverables, either explicitly or implicitly through the deliverable description. If the required TRL is more than 1-2 levels above your current readiness in the relevant capability area, you're likely proposing work you can't credibly commit to delivering.
4
Check agency award history on SBIR.gov/awards
Search the program office's prior Phase I awards in your technology area. Look at company size (employees, revenue), location, clearance level. If the pattern of prior awards consistently goes to companies with 100+ employees, a facility clearance, or an existing program office relationship — that's your baseline competition. Price and adjust accordingly.

Sample Scoring Output

Here's what a scored comparison looks like for three hypothetical topics from a single solicitation cycle:

Topic Scoring — Example
N243-D01 — AI for logistics optimization
Capability alignment
9/10
TRL match
8/10
Agency award history
6/10
Program office relationship
8/10
Composite Score 7.75 / 10 — PURSUE

The AI Shortcut

Building this scoring system manually is valuable as a one-time exercise — it forces you to articulate your capability vector clearly, which has benefits beyond SBIR matching. But running it against 300+ open topics every solicitation cycle is not a realistic use of anyone's time.

OpStop does this automatically. When you set up your company profile — technology areas, TRL stage, team size, clearance level, agency relationships — OpStop scores every open SBIR topic against that profile daily. High-match topics surface at the top of your feed. Wired-topic signals, TRL mismatches, and agency award history are surfaced as flags. You spend your BD time on the 5 topics worth pursuing, not on reading 300 topic descriptions.

The practical outcome for most companies is a 60-70% reduction in time spent on topic discovery and qualification — which translates directly into better proposal quality on the topics that do make the cut, because your writers are focused instead of spread thin.

The bottom line: SBIR matching is a data problem masquerading as a writing problem. Fix the filtering, and the proposals almost write themselves — because you're only writing proposals for topics where you have a genuine, differentiated answer.

Related Intelligence

Let the Algorithm Filter First

OpStop scores every open SBIR topic against your company profile daily. Stop reading 300 topic descriptions. Start writing 5 winning proposals.

Start Free Trial — $10/mo No annual contract. Cancel anytime. Setup takes 5 minutes.