A Customer Made Us Rewrite Our Displacement Number. He Was Right.

A few weeks ago we ran a Reddit thread off our original 22.5% displacement stat. A customer — a principal at a small federal prime — replied bluntly:

“How many of these ‘displacements’ are actually competitive losses, and how many are things like set-aside rotations where the incumbent literally couldn't re-bid, or a contract migrating from a standalone award to a task order on a vehicle where the old incumbent isn't an awardee? The raw number conflates very different outcomes.”

He was right. “Displacement” is a bucket that hides at least four very different outcomes. If you're a small shop making bid/no-bid decisions off the raw number, you're overestimating opportunity.

When an expired contract's work reappears on a new award to a different vendor, that new award falls into one of four buckets:

1. Set-aside rotation (structural).The follow-on carried a different set-aside designation — full-and-open → 8(a), SDVOSB → WOSB, etc. The incumbent wasn't allowed to bid under the new rules.
2. Vehicle migration (structural).Work migrated from a standalone contract to a task order under a MAC / GWAC / BOA where the old incumbent isn't on the vehicle. No seat, no bid.
3. Lost on vehicle (competitive).Work migrated to a vehicle AND the incumbent IS an awardee on that vehicle — they had access and still lost. A real competitive defeat.
4. Lost direct re-compete (competitive). Standalone contract → standalone contract, same set-aside, incumbent lost on merit.

Categories 1 and 2 aren't about whether an incumbent could win — they're about whether they could even bid. Rolling them into “displacement rate” makes the market look more open than it actually is.

The 20.5% number is the one that actually predicts competitive dynamics. If you're building a bid/no-bid model off displacement rates, use this one. The raw 30.2% overstates the competitive opportunity by roughly 47%.

His next question was sharper: “break it out by year and quarter — I want to see if category management push in FY23/24 or the admin change in FY25 moves the line.” It does. Meaningfully.

Three things jump out:

• FY23–Q4 and FY24–Q4 spike on vehicle migration (10.0% and 11.3% of displacements). This lines up with the category management push and the OASIS+ / Alliant 3 / Polaris on-ramps landing at end-of-FY. Work that used to go to standalone contracts was getting pushed onto vehicles.
• Set-aside rotation surged in FY23(32.4% and 32.8% of displacements in Q3/Q4). This matches the 8(a) / SDVOSB policy emphasis and the SBA's rule changes that kicked in late FY22.
• FY25-Q2 is the lowest raw displacement quarter in the entire dataset(23.8%). It's also the first full quarter under the current administration. True competitive displacement dropped to 17.6% — the lowest point in 14 quarters. Whether that's a pause during transition or a durable trend won't be clear until FY25-Q4 data locks in.

Stop writing your capture plan against the 30% number.If you're targeting contracts to displace the incumbent, your real odds are closer to 1 in 5, not 1 in 3.

Segment your pursuit list by displacement type.Of every 100 re-competes in your NAICS, ~70 won't change hands, ~10 will change hands for structural reasons you can't affect (certification / vehicle access / set-aside change), and ~20 are truly competitive. Your B&P budget should target that ~20.

Check the vehicle question first.For every big re-compete you're chasing, check whether the follow-on might land on a MAC/GWAC where the incumbent isn't an awardee. If it does, and YOU aren't on that vehicle either, the contract isn't realistically winnable by you — one of the awardees will take it as a task order.

Watch set-aside history.If you're an 8(a) looking at a full-and-open recompete, check whether the agency has been rotating similar work onto set-asides. That's your real tailwind.

Scope:27,004 matched re-compete pairs across FY2022–FY2025. “Matched” means we identified the most likely follow-on contract for every expired contract above $1M that met our matching criteria.

Matching criteria:same agency code, NAICS code, and PSC code; follow-on action date within 60 days before to 120 days after the expired contract's end date; follow-on value between 0.5× and 2× the expired value; same place of performance when specified. When multiple candidates qualified, we picked the one with the closest action date to the expired contract's end date.

Vehicle awardee detection:we derive awardee lists for every IDV / GWAC / MAC / BOA from USAspending task-order data — every contract with a parent PIID is grouped by that parent, and the union of unique vendor UEIs on those task orders is the vehicle's awardee pool. No curated vehicle names, no scraping, just the raw data. Current coverage: 41,793 parent vehicles.

Known limits: our matching is probabilistic, not ground truth — we can't know with certainty that contract A is the follow-on to contract B, though tight agency+NAICS+PSC+POP+value filters keep false-match rates low. We don't yet segment prime-to-sub flips (cases where the original incumbent continues performing as a subcontractor on the new award). Full methodology at /methodology#displacement.