Start with the official Polaris industry-partner and OASIS+ contractor lists, keeping the published pool or category and observation date with each candidate. Award data does not provide a complete history of bids submitted, a firm’s win rate or future workshare. Use the recommended log below to record observed awards with date and filters, then ask candidate primes specific capacity questions.

Two sources, two different questions

The current vehicle roster is your starting point for identifying listed contract holders. GSA's Polaris page links its industry-partner spreadsheet organized by pool. The official OASIS+ page links the contractor list, which GSA labels as updated weekly. Use the categories exactly as each source labels them, and record the document version or publication date when provided, plus your observation date.

Do not substitute an apparent-awardee announcement for a current contract-holder roster. GSA's OASIS+ page separates rolling apparent-awardee announcements from its contractor list. Verify the actual contract and applicable pool/domain before relying on a candidate’s access to a specific task order.

Contract award search answers "what has this firm been awarded elsewhere?" SAM.gov describes it this way: "The contract award search function in SAM.gov allows users to search federal procurement data and to filter by keyword, agency, and legal business name." That filter set maps directly onto teaming research: search each candidate by legal business name, then narrow by the agency you sell to.

What neither source contains is the number of proposals a firm submitted. Without that denominator you cannot compute a firm’s win rate or rank vehicles by bidding difficulty. An empty scoped search does not establish that a firm is an inactive bidder. Future workshare is not established by these records. Any percentage a prime offers you is a negotiating position for that deal, not a market norm you can verify from award data.

Awardee observation log

Keep one row per candidate prime. Every observation carries its own source, filters, date range and unit, so a reader of your log can tell exactly what was counted.

CANDIDATE OBSERVATION RECORD
Candidate legal name and verified identifier: [name, UEI/CAGE if provided]
Vehicle and contract: [Polaris/OASIS+, contract number and source status]
Pool/domain/category: [label exactly as the official source states]
Roster source: [URL, publication/version date if provided, date viewed]
Award search: [filters, date range and retrieval date]
Observation: [quantity and unit, not an inferred win rate]
Unresolved evidence: [identity, current contract access or other gaps]
Question for the prime: [specific work package or capability question]

Hypothetical example: Example Co. appears in a small-business category on a roster viewed November 6, 2026. A scoped search identifies four award actions at the target department from 2023 to 2026, counted as actions rather than dollars. That does not show how many bids it submitted or whether it needs a subcontractor. Ask which upcoming requirements need your data-engineering capability and what work package is actually contemplated.

A legal-name filter alone can mix unrelated entities. Compare the UEI, CAGE code or contract identifiers the sources provide; keep an uncertain identity unresolved rather than aggregating all similar names as one prime.

From log to shortlist

  1. Tally the distribution, then set it aside. Count awardees per category as the source labels them. The tally tells you where seats are concentrated, not where task-order work will flow or which firms are stretched.
  2. Run one consistent query per candidate. Same agency filter, same date range, same unit. Inconsistent filters produce rankings that reflect your searching, not the firms.
  3. Label empty results carefully. "No actions found for this query on this date" is accurate. "Not winning" or "not bidding" is not supported by award data.
  4. Rank by fit, not by totals. A candidate whose observed work sits near your capability gap is a better first call than the firm with the largest count.
  5. Ask capacity questions the data cannot answer. For example: Which task-order requirements are you tracking where you need my capability? What scope and labor categories would a subcontractor perform? How do you document workshare in a teaming agreement? How do you handle teammates when a task order is smaller than expected?
  6. Re-date before you commit. Award data and award lists change. Rerun queries before you sign anything exclusive.

For how GWACs, IDIQs and GSA Schedules differ as buying paths, see GWAC vs IDIQ vs GSA Schedule. For what to put in writing once a prime is interested, see teaming agreements and joint ventures in federal contracting.

Next step

Explore public federal contracting opportunities on FedFinder. Review current terms on the pricing page.

Sources