The Federal Recompete Playbook: Expiring Award to Dated Contact Plan

This playbook takes a lane of federal work you can deliver and turns it into a ranked list of 10 to 25 real recompetes. Each one comes with a true last possible end date, the named government and incumbent people behind it, and a dated window to start talking. It is for founders selling to government, account executives, SDR leads and capture leads who already know their codes, and it starts where most recompete guides stop: right after the end-date filter.

What a finished recompete pursuit list contains

A finished pursuit list has 10 to 25 rows. Each row is a contract you have confirmed will be re-solicited, and it carries one defensible end date, a recompete verdict, named people on the buyer side and the incumbent side with a source for each, and an outreach window with a close date. A spreadsheet of awards sorted by end date is the raw material. It is not the deliverable.

Terms used throughout, each defined once:

• Recompete: a new solicitation for work an incumbent contractor is already doing.
• Incumbent: the contractor that holds the current award.
• NAICS and PSC: the industry codes and product and service codes the government uses to classify what an award buys.
• Option period: a pre-priced extension the government can exercise without competing the work.
• IDV: an indefinite delivery vehicle, a contract under which separate orders are placed.
• Bridge: an extension or short non-competitive award that keeps work going because the follow-on is late.
• J&A: justification and approval, the document needed to award without full and open competition.

Each row carries the fields below. Copy the schema into your sheet or CRM. The source and checked-date fields are not optional. You can't re-verify a name that has no source, and you can't defend an end date that has no cap note.

award_id:
agency / contracting office:
naics / psc:
current_end_date:
potential_end_date:
ordering_end_date (IDV only):
true_last_possible_end_date:          cap applied:
agency_rfo_part17_status:             checked on:
verdict: recompete / bridge-pending / absorbed
eligibility: go / no-go               set-aside:
contracting_officer:                  source:       checked on:
contract_specialist:                  source:       checked on:
program_manager:                      source:       checked on:
cor_or_end_user:                      source:       checked on:
incumbent_program_lead:               profile checked on:
incumbent_capture_lead:               profile checked on:
window_open:
window_close:                         driver: final RFP est. / industry day / draft RFP comments
rank:

The recompete pursuit procedure, step by step

Seven steps take you from a code list to a dated contact plan: build the universe, compute the true end date, read the modifications, check eligibility, map the government people, map the incumbent team, and date the window. An analyst does steps 1 to 3, the capture lead owns step 4, SDRs do steps 5 and 6, and the capture lead dates the window in step 7.

The time budget is about one analyst day for the pull, about 30 minutes per contract for the date calculation, and 1 to 2 SDR hours per surviving row for the people map. For a final list of 25, the people work alone comes to 25 to 50 SDR hours. That is why steps 2 to 4 have to cut the list before step 5 starts.

Vendor guides on this topic usually start at step 5 or skip step 3. That is how lists fill up with contracts that are not actually ending.

Current vs potential end date: finding the true last possible end date

The true last possible end date is the latest of three dates: the Potential End Date, the Current End Date plus 6 months under FAR 52.217-8, and for IDVs the ordering end date. You then adjust it for whatever duration cap the buying agency applies. No single field in the public award data gives you that number, so you compute it per row.

What each date field proves

The federal spending data standard, documented in the USAspending Data Dictionary and the Period of Performance whitepaper, has three separate date fields. They answer three different questions:

The standard defines Potential End Date as the end date "if all potential pre-determined or pre-negotiated options were exercised." It defines the ordering end date as the date on which "no additional orders referring to it may be placed," and adds that order end dates "may extend beyond this date." So you cannot use an IDV's ordering end date as the end date of its orders. Check each order's own period of performance.

None of these fields flags a bridge. A bridge shows up only as a later modification that moves the Current End Date. Each modification is one transaction with its own action date, so the history is readable, but only if you actually read it.

Duration caps by contract type

The caps decide whether a contract can legally run past its Potential End Date without a new competition. Under the legacy FAR 17.204(e), "the total of the basic and option periods shall not exceed 5 years in the case of services," unless the agency approves otherwise. That limit does not apply to IT contracts. Other statutes, such as the Service Contract Labor Standards statute, can add further restrictions.

The 52.217-8 extension "may be exercised more than once, but the total extension of performance hereunder shall not exceed 6 months." Six months is the most the clause allows, but only if the clause is actually in the contract.

The RFO split

The Revolutionary FAR Overhaul (RFO) is a rewrite of the FAR that agencies adopt one by one through class deviations. Its Part 17 removes "the limitation in FAR 17.204(e) that the total duration of the basic contract and option periods shall not exceed five years." DOE and DoD have both issued Part 17 deviations.

Adoption is uneven. One count showed 36 agencies on the RFO version of Part 1 but only 27 on the RFO version of Part 15. That means two identical contract structures can carry different caps at two different agencies. Check the FAR Overhaul agency deviation guide on acquisition.gov, and for DoD the DFARS RFO class deviations page. The RFO is still moving through formal rulemaking, so record the date you checked on every row.

Will it really be recompeted? Reading modifications and eligibility

A contract is a real recompete when its modification history shows it running out of options and extension room, nothing suggests the work is going somewhere else, and your firm is eligible for the likely follow-on vehicle. Step 3 gives every row one of three verdicts, and step 4 gives it a go or no-go.

Separating option exercises from extensions

Walk the modifications in action-date order and sort each one that moves the Current End Date:

• Option exercise: the Current End Date moves but stays within the Potential End Date. This is normal life on a contract.
• 52.217-8 extension: the Current End Date moves past the last option by a cumulative 6 months or less. The follow-on is probably a little late.
• Bridge: cumulative extension past the options goes beyond 6 months. The DoD bridge definition excludes 52.217-8 extensions until those 6 months are exceeded, which is the same line this playbook uses.

Then assign the verdict:

A bridge is not a field in the data. It is a date that moved, and you only see it if you read every modification.

Lists built only on Potential End Date leave out exactly the contracts that are delayed or under protest. GAO's review of justifications for non-competitive awards shows how common those are.

Bridges are documented as being driven by delay in awarding the follow-on. Treat bridge-pending as a timing signal, not a signal about how the incumbent performed.

Eligibility and the incumbent's position

Step 4 checks whether you can bid at all and what the public record can tell you about the incumbent:

• Set-aside status. If the follow-on looks like a set-aside you don't qualify for, the row is a no-go however attractive it looks.
• IDIQ ordering limits. DoD task order contracts have an ordering period that "shall not exceed 10 years" unless the head of the agency waives it in writing.
• Size re-certification after M&A. A small-business incumbent must re-certify "within 30 days after a merger or acquisition," and agencies should report it to FPDS within 30 days. That shows the incumbent may not qualify for a small-business set-aside follow-on. It does not show poor performance. It does not change the current contract either, because "a change in size status does not change the terms and conditions of the contract."
• CPARS. Performance evaluations "shall not be released to other than Government personnel and the contractor whose performance is being evaluated." Evaluations are used for 3 years after completion, or 6 for construction and architect-engineer work. A challenger cannot confirm poor performance from public data.
• Modification frequency and protest history. I found no published validation that either one predicts an incumbent loss. Record them, but don't rank on them.

The incumbent-weakness signals you can actually check from outside are ownership change and protests. Everything else is inference, and the row should say so.

Who is the contracting officer, program manager and COR behind a federal contract?

On the buying side, the contracting office usually shows up through notices, which name the contracting officer or contract specialist. The program manager and the COR sit in the program office and have to be found through organisation pages and professional profiles. FAR 15.201 names "the program manager, contracting officer, and other participants in the acquisition process" as the people who exchange information with industry, so a complete row names both sides.

The roles and what each contact is for:

• Contracting officer (CO): the official with authority over the contract. One-on-one meetings "substantially involved with potential contract terms and conditions should include the contracting officer."
• Contract specialist: works in the same contracting office and is often the named point of contact on notices.
• Program manager: owns the requirement and the mission need behind it.
• COR (Contracting Officer's Representative): the end-user-side person who oversees day-to-day performance, usually as a delegated duty.

How hard each role is to find varies a lot. Refolk's index of US professional profiles shows the spread:

Two things follow from the table. First, if you can't find the CO, look for a specialist in the same office. The pool is about three times larger, so someone in the right office is easier to reach.

Second, the program office side is hard to find by job title. Only 561 profiles carry a COR title against 6,969 COs, roughly 12.4 COs per COR, and 72% of the sampled CORs are Air Force. COR is a delegated duty, not a job title, so the count is an undercount and leans toward the organisations that do put it in titles. To find end users, search by program, organisation and location, not by title.

What share of notices name a point of contact, and which roles they name, is not publicly established. Track it on your own rows for a quarter and you will know for your lane.

Searching by organisation and work, rather than exact title, is the slow part of step 5, and it is the part Refolk is built for. You describe the office and the work in plain English and get named people back across LinkedIn and the open web, with profiles you can re-check before you write.

Mapping the incumbent team

Step 6 names two people at the incumbent: the program manager running the contract and the capture lead who will run the defence. They tell you who the government talks to every week and who will be writing the competing proposal. Record the date you checked each profile, because people on the incumbent's team move around, often right before a recompete. In Refolk, a search like "Program managers at a federal IT services contractor who lead a Department of Veterans Affairs contract, Washington DC area" returns people by the contract they run rather than the title they use.

When to start talking: dating the outreach window

Open the outreach window today. Close it at the earlier of two dates: the estimated final RFP date, or the last scheduled exchange event, such as an industry day or the draft RFP comment deadline. How much time is left in that window decides the row's rank.

What the rules actually allow

The legacy FAR is more permissive than most people assume. FAR 15.201 says: "Exchanges of information among all interested parties, from the earliest identification of a requirement through receipt of proposals, are encouraged." Interested parties "include potential offerors, end users, Government acquisition and supporting personnel." The techniques it lists are conferences, market research, one-on-one meetings, presolicitation notices, draft RFPs and RFIs. It also says "general information about agency mission needs and future requirements may be disclosed at any time."

The legacy text names no milestone after which only the contracting officer may be contacted. Where that restriction exists, it comes from each solicitation's own instructions. Whether the RFO version of Part 15 changes 15.201 is not established here. Check the text that applies at the buying agency, using the same deviation guide as in step 2.

Practitioners treat the final RFP as the practical cutoff even though the FAR allows exchanges until proposals are received. I use the earlier, more conservative date. It costs you a few weeks of theoretical access, and it keeps you clear of whatever instructions the solicitation turns out to contain.

The widely repeated "12 to 18 months before the RFP" capture window has no published method behind it. The 12 to 30 month universe in step 1 exists to give you room, not to hit a magic number.

What the incumbent advantage actually is

Timing matters because incumbents win often, though nobody has measured how often with a published method. The published figures disagree:

Lisa Pafe of Lohfeld Consulting has said incumbents lose half of recompetes. The honest summary is a range from about half to three quarters, depending on who was asked and when. Vendor blogs also claim challengers who wait for the RFP win about 18% of the time. That figure is unverified. Use it as a reason to start early, not as a forecast.

Ranking the rows

Drop every no-go and every absorbed row first. Rank the rest on two things: how much window is left and how complete the people map is.

The first message to a program manager or end user should be about the mission need, which FAR 15.201 says can be discussed at any time. Leave contract terms out of it. Anything about terms belongs in a meeting that includes the contracting officer.

Subject: [Program or office name] - market research input on [capability area]

Hi [First name],

I lead [your company], which delivers [one-line capability] for [similar agency or mission]. I understand [program or office] is likely to look at [capability area] again ahead of the follow-on to [contract name or number].

Before any formal notice goes out, I'd value 20 minutes to understand the mission need as you see it: what's working, what has to change, and what the end users are asking for. I'm happy to send a short capability summary in advance, or respond to an RFI or sources sought notice if that's the better route.

If someone else in the office is the right person, a pointer would be very welcome.

[Name]
[Title, company, phone]

How recompete pursuit goes wrong

Recompete lists fail in predictable ways. Most of them come from trusting one field, one rule or one signal that does not prove what it looks like it proves. These are the failure modes that produce false positives, each with the check that catches it.

1. Treating Potential End Date as the real end. Up to 6 months of 52.217-8 extension, plus any later bridge, pushes the work out. False positive: the "expiring" contract runs for another year. Check: compare against the latest modification's Current End Date.
2. Applying the 5-year cap after the RFO. Agencies on the RFO Part 17 deviation have no FAR default cap, so a contract you expected to hit its limit may legally keep running. Check: the agency deviation list on acquisition.gov, dated on the row.
3. Using an IDV's ordering end date as the end of its orders. Orders placed before the ordering end date can run past it. Check: each order's own period of performance end date.
4. Missing DoD awards. DoD and Army Corps of Engineers actions reach the public data about 90 days late, according to a secondary source citing USAspending. A pull that only covers recent months under-counts DoD awards and misses recent extension modifications. Check: re-pull after 90 days.
5. Assuming a bridge means the incumbent is weak. In the GAO sample, 5 of 18 bridges came from protests, and bridges in general are driven by follow-on delay, not poor performance. Check: the GAO docket, to see who protested.
6. Assuming a size change means an open recompete. Re-certification does not change the terms of the current contract, and the incumbent can still compete full and open. Check: whether the follow-on is actually likely to be a set-aside before you count it as an opening.
7. Contacting a stale person. Contracting officer and program manager titles change, and an email to someone who has left the office costs you the window. Check: re-verify every name against a current profile before outreach.
8. Citing an "80% incumbent win rate" as fact. No published method supports it, and a forecast built on it will misjudge every row. Check: use the dated survey ranges in the win-rate table instead.

Failure mode 7 is the one that hits even disciplined teams, because names go stale between the people map and the send. Re-running the same plain-English search in Refolk the day before outreach is a cheap way to catch a CO or program manager who has moved on.

Keeping the pursuit list current

A pursuit list starts to decay the day you finish it. Dates move with every modification, names move with every reassignment, and agency rules move with every deviation. Treat it as a standing document with a re-check rhythm, not a one-off export.

Keep it current on four triggers:

• 90 days after each pull. Re-pull your codes for DoD, recompute the true end date on any row whose modifications changed, and add DoD rows that have since appeared.
• When a notice posts. A presolicitation notice, RFI, draft RFP or industry day announcement changes the close date of the window. Re-date it the same day.
• When the agency changes its deviation. Re-check the FAR Overhaul deviation guide and the DFARS RFO deviations page, and update the cap note and check date on affected rows.
• Before every send. Re-verify the named person's current title and office.

Run this checklist before you call a row ready for outreach:

When a row passes, it moves from list to plan. When a row's window closes without contact, record why: no names, late start, or a date you got wrong. Those reasons tell you which step of the procedure to tighten next quarter.