FAR 19 Killed Your Subcontracting Safety Net. Here's How to Rebuild Your Bid Strategy.

Executive Summary

✹ FAR 19 reforms substantially weaken mandatory small business subcontracting plans and SBA reporting requirements

✹ OASIS averages 3.4 bidders per task order, making AI-assisted bid volume the highest-return strategy for firms on that vehicle

✹ Proposal AI vendor selection is now a compliance and risk management decision

✹ AI-scored procurement is live in federal contracting, moving the winning formula from polished narratives to structured compliance

✹ The Shipley review model needs redesigning for AI-speed production timelines

What is FAR 19 reform and how does it affect proposal teams?

FAR 19 reform refers to changes in Federal Acquisition Regulation Part 19, which governs small business programs in government contracting. The 2025 reforms substantially weaken small business subcontracting plans, as the current administration has deprioritized enforcement and SBA oversight. For proposal teams, this means the teaming arrangements that previously brought small businesses into large prime bids as a compliance requirement are no longer mandated the same way.

The subcontracting plan is gone. Teaming has to earn its place.

For proposal managers who built teaming strategies around small business subcontracting requirements, the ground has moved. The obligation for large primes to include small businesses in their bids, backed by SBA reporting, is gone.

Arthur Runno, CEO of Turingon and a 20-year federal proposal veteran, puts it bluntly. If a large prime integrator doesn't have to use a small business, what are the odds they will? The enforcement mechanism, the reporting to SBA, the set percentages, that entire structure is being dismantled. For small firms that depended on mandatory subcontracting inclusion, the question is how to compete when inclusion is optional.

The answer, according to Runno, is process and teaming that adds clear value. If your firm is on OASIS or another best-in-class GWAC, you have a vehicle advantage. If you are not, the path forward is proving capability to primes who now have full discretion over their team composition.

OASIS averages 3.4 bidders per task order. The math says bid more.

Here is a number that should change how you plan your bid calendar: OASIS averages 3.4 bidders per task order per Runno's analysis of OASIS task order records. For firms on that vehicle, the odds of winning improve with every additional submission.

The constraint that historically limited bid volume was cost per bid. The labor hours, SME coordination, and management overhead required to produce a compliant response made it rational to be selective. But for this case specifically, AI changes that calculation.

Runno's approach is to use AI to generate 80 to 90 percent of the first draft. Bring in teaming partners for the remaining 10 to 20 percent, offering work share as an incentive. Attach pricing and submit. For task orders where you are qualified, the expected value of bidding now exceeds the cost of producing the response.

Additionally, Executive Order 14240 directs agencies to stop standing up their own IDIQs and route spend through Best-in-Class vehicles like OASIS+. That means task order volume on the vehicle is going to increase. The agencies that previously managed their own professional services contracts are being pushed onto a vehicle where 3.4 firms are competing per order.

For firms already on OASIS+, this is not an abstract policy change, but it is a pipeline expansion. The math on bid volume gets better as more requirements migrate onto the vehicle. Small businesses building a bid volume strategy on OASIS+ SB pools need to track one additional variable - their own size. OASIS+ requires re-representation at the master contract level, and continuous on-ramping means GSA is reviewing eligibility on a rolling basis. A firm that wins enough work to cross a NAICS size standard threshold mid-contract may find itself competing in pools it no longer qualifies for. The bid volume strategy that grows your revenue can also be the mechanism that reclassifies you out of the vehicle that enabled it. Build the tracking into your contract administration now, not after a recompete surprise.

Capture still matters for high-value pursuits, but for the long tail of GWAC task orders, AI-assisted bid production turns no-bid defaults into revenue opportunities.

The vendor evaluation checklist proposal teams need right now

If AI tools are central to this bid volume strategy, vendor selection is a high-stakes decision. Based on Runno's experience evaluating and building proposal AI, here is what to ask before shortlisting:

• Demand a free trial with your own data and a real solicitation. If the vendor will not let you test the product before buying, that is a red flag. Runno notes that at least one major vendor conducts demos entirely through pre-built PowerPoint presentations, not showing the live product until the second or third call. A proposal tool should stand on its own.

• Ask where your data is stored. For GovCon, this means asking about CMMC compliance and whether the platform runs on FedRAMP-compliant infrastructure like AWS GovCloud. Tools charging $100 per month cannot realistically operate on GovCloud. The server costs alone make it impossible. If the pricing seems too low, the data is probably not stored where it should be.

• Ask whether the vendor is bidding on government contracts itself. If so, your vendor is your competitor, and your proposal data is on their servers.

• Check user limits. Can your teaming partners and SMEs access the tool without procurement delays? Proposal timelines do not wait for license approval.

• Be wary of all-in-one integrated platforms. The best tool for proposals may not be the best tool for capture or contracting. Locking your team into a platform because another department chose it is, as Runno puts it, "like trying to write a proposal in Excel because accounting likes it."

AI is scoring your proposals. Compliance is the winning narrative.

The change from human-evaluated to AI-evaluated proposals is underway in federal procurement. Runno, who spent 20 years priding himself on strong proposal narratives, acknowledges the implication that adherence to evaluation criteria becomes the primary differentiator when a machine is reading your response.

There are known instances of AI-assisted evaluation tools being tested or piloted, but widespread AI scoring of federal proposals is not yet standard practice across civilian and DoD contracting. The FAR still requires contracting officers to make source selection decisions, and automated evaluation tools currently function more as compliance checkers than scoring engines in most documented uses. But the use of AI is an emerging trend we need to watch.

This means compliance engines, automated cross-referencing against the SOW, proposal instructions, and evaluation criteria, are no longer nice-to-have features. They are the mechanism that determines whether your response scores favorably.

For proposal managers, this moves their the daily workflow, with less time polishing prose and more time ensuring every claim is substantiated, every requirement is addressed, and the structure maps cleanly to what the evaluator (human or AI) is looking for.

The Shipley review model needs rethinking

Runno made a claim that will irritate some in the profession, but that we support at stargazy - the Shipley process is dead. He does not mean reviews are dead, but he means the timing, the participants, and the stage at which reviews happen need to be redesigned for modern AI-speed production.

When AI can produce a credible first draft in hours, scheduling a color team review on a timeline designed for weeks of manual drafting no longer makes sense. The review process needs to compress. More emphasis on red team quality with AI-assisted pre-review and less ceremony around pink and blue stages that were designed for a different production model.

This does not reduce the importance of human review, but it changes where human attention adds the most value.

FAQ

How does FAR 19 reform affect small business teaming in government contracting?

FAR 19 reform eliminates mandatory small business subcontracting plans and SBA reporting requirements for large prime contractors. Small businesses that relied on these mandates for inclusion in prime bids must now compete on capability alone.

What is the best way to increase bid volume on GWAC task orders using AI?

Use AI proposal tools to generate 80 to 90 percent of first drafts for qualified task orders. Bring in teaming partners for specialized sections. With OASIS averaging 3.4 bidders per task order, the expected value of additional submissions outweighs the reduced cost per bid.

What security questions should GovCon teams ask proposal AI vendors?

Ask about CMMC compliance, FedRAMP-compliant infrastructure (such as AWS GovCloud), data storage location, whether the vendor bids on government contracts, and whether user access can scale to include teaming partners and SMEs without procurement delays.

Is the Shipley proposal review process still relevant?

The Shipley framework's review stages need to be redesigned for AI-speed production. The principle of structured review is important, but the timing and sequence of color team reviews should compress to match the pace of AI-assisted drafting.