Why Federal Contractors Need to Renegotiate Their Joint Ventures Before September 21

A company billing $531 million a year could soon be a small business. That figure is the proposed size standard for custom computer programming, an industry whose current threshold is $34 million, and it is the reason any federal contractor sitting inside a mentor-protege joint venture should reopen the agreement this week rather than after the next quarterly review.

The Small Business Administration's August 20 proposal does not rewrite the mentor-protege rules. It does something quieter and more disruptive to the economics of those deals: it makes a large number of mentors eligible to be small on their own. The one thing a protege reliably brings to the venture, its size status, is about to become available elsewhere at a lower price. Public comments close on September 21, 2026.

The scale is not marginal. SBA estimates 114,541 additional firms would qualify as small, fewer than 200 would lose the status, and 37,002 of the newly small held federal contracts in FY2025 worth more than $71 billion across roughly 105,655 awards. Those are the new competitors for a mentor's attention, and most of them do not require a mentor-protege agreement to bid alongside.

Federal size standards decide who may bid on set-aside work. Today there are roughly 1,000 of them, assigned at the six-digit NAICS level: 102 size-standard levels spread across 978 industries, plus 18 subindustry exceptions. The proposal collapses that into 338 standards at the four- and five-digit level, a cut of about 65% in line items, and eliminates every existing exception, including the ITVAR carve-out. The calculation behind each standard drops from seven factors to three: national industry size, number of geographic markets, and an adjustment for net imports.

Consolidation to a broader industry group means the threshold usually rises to accommodate the largest industry in the bucket. Hence the increases, which in professional services are not increments but reclassifications. Architectural services would move from $12.5 million to $135 million. Administrative management consulting from $24.5 million to $295 million. Engineering services from $25.5 million to $252 million. Construction leaves revenue behind entirely: NAICS 237990 converts from a $45 million receipts standard to 1,500 employees, and commercial construction to 600.

A mentor-protege joint venture is, at bottom, an exemption slip. Two firms that would otherwise be treated as affiliated get to bid as one small entity, because the larger partner has agreed to develop the smaller one. The slip has value precisely because the mentor cannot get the status any other way.

Raise the ceiling and the slip becomes redundant for a meaningful share of mentors. The lawyers at PilieroMazza put the consequence bluntly: more mentors will qualify as small on their own, which removes the need for an agreement in order to joint venture with another small firm, and makes it harder for newer firms to attract willing mentors. Thompson Hine reaches the same place from the other direction, noting that a mentor newly classified as small could form a venture with any small firm rather than its existing protege.

The compensating movement runs the other way. Spencer Fane observes that higher thresholds give small-business ventures far more room before affiliation becomes a problem, and open up small-to-small combinations that would previously have disqualified both parties. A protege at $20 million and a peer at $25 million, once forced apart by a $34 million standard, can now pair without a mentor at all. The mentor-protege structure is not being abolished. It is being priced against a substitute that did not exist before.

Consider an IT services venture in which a $180 million mentor carries the past performance and a $12 million protege carries the status. Under the proposed $531 million standard, the mentor is small without help. The protege's contribution is now capability, not eligibility, and the split should be renegotiated to reflect the work each partner actually performs. If the protege cannot show what it contributes beyond a certification, it has little to defend at the table.

Now consider a specialty trade contractor whose venture was built around a $45 million receipts standard. The measure itself changes: the industry moves to a headcount test of 1,500 employees. A labour-heavy firm well under the old revenue cap may find its position materially different, in either direction, and the venture's assumptions about which partner qualifies need rechecking against the new industry group rather than the old six-digit code. This is also a moment when a mentor quietly meeting other small firms stops being unusual behaviour and starts being a warning sign worth naming.

It is a proposal, not a rule. Nothing binds anyone today, the SBA had already received more than 65,000 comments on the size standards as of September 1, and a rule attracting that volume rarely emerges unchanged. The Federal Register document mentions the mentor-protege programme only as a certification route for newly eligible firms; it sets out no change to the affiliation exception or to joint venture rules. Anyone who tears up a working agreement on the strength of a comment-stage notice is taking a position on a document that may not survive its own docket, and no effective date has been published.

All true, and it changes the timetable rather than the analysis. Renegotiation here costs a meeting and a redline. The asymmetry is what matters: a venture that adds a recertification trigger and a repricing clause loses nothing if the rule dies, while a venture that does not loses its bargaining position on the day the final rule prints, when the mentor already knows it is small.

The pricing question deserves its own session rather than a paragraph in a redline, because the choice between a revenue share or equity determines how the venture absorbs a change in status without reopening everything again.

The proposed rule sits at docket SBA-2026-0199, RIN 3245-AI67, published in Federal Register volume 91, issue 160, as document 2026-17042. The companion methodology notice, which carries the three-factor calculation, is filed separately at docket SBA-2026-0265, and comments on both are due the same day. The Federal Register notice carries the filing instructions and names Ryan Lambert, Associate Administrator in the Office of Government Contracting and Business Development, as the agency contact.

Comments that describe a specific competitive effect tend to travel further than comments that object in general. A protege able to show, with contract numbers, that its mentor would become small under the proposed grouping and that the relationship would lapse as a result is describing something SBA's own impact estimates do not capture.

Regulatory certainty to scale is exactly what the proposal offers, and the firms it offers it to are the ones already near the ceiling. The firms at $8 million, whose entire access to federal work runs through a partner's willingness to carry them, get certainty of a different kind. They should assume their mentor has read the same alert and is running the same numbers, and open the conversation first.