Skip to content
Vehicles and accessAmyn Porbanderwala, Founder14 min read

Selling software to the federal government: which contract vehicle

There is no best vehicle for selling software to the federal government. There is a portfolio, and the choice depends on six inputs. A dated comparison of every route that matters, including the 2026 changes that broke most online advice.

There is no best contract vehicle for selling software to the federal government. There is a portfolio, and which one you need depends on six things you can write down in an afternoon: deal size, market breadth, agency focus, SBIR heritage, whether you are a non-traditional contractor, and how long you can wait.

What there is, is a lot of stale advice. In 2026 the federal vehicle landscape changed more than it had in the previous five years. CIO-SP4 was cancelled. Alliant 3 was awarded. GSA made transactional data reporting mandatory on the schedule. FedRAMP replaced its impact levels with certification classes. A guide written in 2024 will send you at a vehicle that no longer exists, or into a pricing clause that has been removed.

Every vehicle fact on this page was checked in October 2026, and anything that could have changed since is dated. Where a figure rests on a secondary source rather than a government document, it is given as a range or labelled an estimate.

The short version. Selling to one agency: an agency IDIQ, or an SBIR Phase III if you have one. Selling to many civilian agencies: GSA MAS. DoD and non-traditional: OTA first. Everything else is a portfolio decision, and the table below is the reference.

Two things to understand before the table

A vehicle is a purchasing mechanism, not a sales channel. Being on GSA Advantage does not create demand. It removes friction for a buyer who has already decided. Firms that get a schedule and wait for orders wait a long time.

A vehicle does not fix a product problem. If nobody can name the agency that will buy the thing, no vehicle helps. That is why the first two steps are deciding what you sell and who buys it, which our guide to productizing a government contract covers, and why the cost of getting the vehicle wrong is a line item in what productization costs.

With that said, the vehicle decision is real, and getting it wrong costs months.

The landscape, as of October 2026

Vehicle What it is Who can hold it Carries Time to hold Cost to hold
GSA MAS The governmentwide schedule. The default route for commercial software and services Any eligible commercial seller Products and services, including SaaS and cloud 3 to 12 months 0.75% of reported sales, plus a minimum sales requirement
OASIS+ GSA's governmentwide multiple-award IDIQ contracts for non-IT professional services: 13 domains, six set-aside pools Awardees in the relevant pool Professional services, not products Rolling on-ramps Proposal cost
Alliant 3 GSA's full-and-open IT services GWAC 43 phase 1 awardees, announced February 2026 Enterprise IT services task orders Live since March 2026 Proposal cost
NASA SEWP VI NASA's governmentwide IT contract 1,490 vendors across three categories IT products, hardware, cloud and related services Awarded; ordering from 1 November 2026 Contract access fee
8(a) STARS III GSA's 8(a) IT services GWAC, $50B ceiling 8(a) firms IT services and IT-based solutions Fully awarded; ordering through July 2029 Proposal cost
Agency IDIQ One agency's own multiple-award contract That agency's awardees Whatever the agency buys 6 to 18 months Proposal cost
BPA A blanket purchase agreement under a schedule or IDIQ Schedule or IDIQ holders Repeat buys of a defined product Weeks to months None beyond the underlying vehicle
OTA Other transaction authority, DoD DoD and its components Prototypes and follow-on production Months, not years Minimal
SBIR Phase III Sole-source follow-on from an SBIR Phase I or II SBIR awardees Work that derives from the SBIR effort Weeks to months Minimal

Three of those deserve a warning before the detail. CIO-SP4 is gone. NITAAC cancelled it in late January 2026 after a review found more than 90% duplication with vehicles GSA and NASA already run, and the office is winding down: no new orders after 29 October 2026, and all task orders end by 31 December 2028. Alliant 3 is not a small-business path. Forty-three phase 1 awardees were announced in February 2026 and the contract went live in March. It is full and open, so a small business cannot hold it. SEWP VI is awarded but not yet the contract you order on. Its period of performance starts on 1 November 2026, and SEWP V was extended to 31 January 2027, so an order placed today still lands on the older contract.

The selection method

Start with the six inputs. They are not equally important, and two of them usually decide the answer.

1. Deal size. Under about $100,000 in annual contract value, a governmentwide vehicle matters more, because the buyer needs a low-friction purchase. Above $1M, the buyer's own contracting shop will run a competition whatever vehicle you hold.

2. Market breadth. One agency, two or three, or governmentwide. This is the input most firms get wrong, because they assume governmentwide when they have one buyer.

3. Agency focus. Civilian, DoD, or both. Civilian buyers lean on GSA and agency vehicles. DoD has its own, and OTA is a DoD-only door.

4. SBIR heritage. If you have a Phase II, you have the only sole-source path most small firms will ever qualify for. Start there.

5. Non-traditional status. If you have not done FAR-based work with DoD in the past year, OTA is designed for you.

6. Patience. A schedule takes months. An agency IDIQ can take a year. A Phase III can take weeks. If revenue is needed inside six months, the vehicle decision is not the constraint; a services bridge is.

The sequence that follows is simple:

  • If you have an SBIR Phase II, pursue Phase III first. It is the cheapest route and it does not preclude anything else.
  • If your buyer is DoD and you are non-traditional, look at OTA before anything.
  • If you sell a product to civilian agencies and expect many buyers, GSA MAS is the default. Budget the modification or the offer.
  • If one agency is 70% of your market, an agency IDIQ or a BPA under an existing vehicle beats a governmentwide contract you will not use.
  • If you are chasing task orders above $5M in IT services, you need a GWAC or a large IDIQ, and you are competing with firms that have held one for a decade. That is a different business.

GSA MAS in detail

First, a distinction that trips up a lot of vendors: GSA MAS is not a GWAC. A GWAC is a specific statutory category of governmentwide contract for IT, awarded under a delegated authority. MAS is a schedule: an indefinite-delivery contract that any agency can order from, run by GSA's Federal Acquisition Service. Both let an agency buy without a full competition. They are governed differently, and the paperwork is different.

The Multiple Award Schedule is the closest thing to a storefront the federal government has. A buyer with a requirement can compare priced offers from schedule holders and place an order without running a full competition. For software under roughly $500,000 a year per customer, it is usually the right first vehicle.

The SINs that matter for software

A schedule is organized by Special Item Numbers, and only some of them let you sell software rather than labor. The ones a software seller usually needs:

  • 54151S, IT professional services
  • 518210C, cloud computing and cloud-related IT professional services
  • 511210, software licenses, sold as term or perpetual
  • 54151HEAL, health IT
  • 54151HACS, highly adaptive cybersecurity services

If you sell a hosted product, the cloud SIN is the one that carries it. If you sell a license, 511210 is the one that fits, including both term and perpetual licenses. Getting the SIN wrong is one of the more common reasons an offer stalls.

How you get on it

  1. An active SAM registration with a UEI. Our guide to your SAM entity record covers what has to be correct before you start, and the registration timeline covers how long it takes.
  2. A GSA offer with your pricing, your commercial sales information, past performance, and financial documentation.
  3. A negotiation. GSA will ask you to justify prices, and it evaluates reasonableness using the transaction data you report rather than the old commercial-sales benchmark.
  4. An award, then a catalog in GSA's systems.

Three to twelve months is the realistic range, and the longer end is common for a first offer. It is not a number GSA commits to, and the calendar is the part firms underestimate. The award runs for a five-year base period with three five-year options after it.

What it costs to hold

  • The Industrial Funding Fee is 0.75% of reported sales. You collect it from the customer and remit it quarterly.
  • There is a minimum sales requirement, and it is larger than most firms expect. The clause is I-FSS-639: $100,000 in the first 60 months, then $125,000 in each 60-month period after that. A schedule that produces nothing can lose its next option.

Neither number is large against a real product line. Both surprise firms that treated the schedule as a one-time award.

What changed in 2026, and why your old notes are wrong

This is the part most guides have not caught up with.

Refresh 31 made transactional data reporting mandatory for every SIN. The solicitation removed all non-TDR and CSP-1 references. That means the old advice about the Price Reduction Clause, the Basis of Award, and the Commercial Sales Practices disclosure describes a regime the current solicitation no longer contains. In its place you report transaction data: order date, ship date, zip code, customer agency code, order type, and for the cloud SIN, the cloud service type. The data feeds GSA's price reasonableness work.

The practical consequence is that the trap moved. It used to be a pricing clause that could trigger a refund. It is now a reporting obligation that runs monthly, within 30 days, and a firm that under-reports has a compliance problem rather than a pricing one. The current price-adjustment clause, GSAR 552.238-81, covers Government-requested and voluntary reductions only. The legacy Price Reduction Clause and the Most Favored Customer benchmark are gone from the solicitation.

One more change worth knowing if you sell a subscription: GSA permits upfront payment for SaaS under Acquisition Letter MV-2024-01, issued in March 2024. It has conditions, including contemporaneous access, fixed-price terms, and no extra upfront fee. A contracting officer who does not know the letter exists will still tell you the government cannot prepay.

Refresh 32, released in June 2026, added three things worth knowing: a new clause at FAR 52.222-90 addressing DEI discrimination by federal contractors, end-of-support requirements for products in every large category, and a new Grants Management SIN. The end-of-support language matters for software: you are responsible for the lifecycle, and GSA can remove items it considers a security risk.

Refresh 33 followed in October 2026, four months later and four days before this page was checked. It adds two supply-chain risk clauses, GSAR 552.540-70 and 552.540-71, and expands the FASt Lane offer process to every category. The point of listing both refreshes is not completeness. It is that the schedule you are on changes two to three times a year, and a compliance process that does not read the refresh notices is not a compliance process.

SBIR Phase III in detail

If you have an SBIR Phase I or II, Phase III is the best vehicle most small firms will ever qualify for, and almost nobody uses it properly.

What it is. A Phase III award is a follow-on that derives from, extends, or completes the SBIR effort. It can be awarded by any federal agency, not only the one that made the Phase II award. It can be sole-source, without competition. There is no dollar ceiling.

Why it matters. It is the only sole-source route where the justification is the work you already did. Other sole-source paths exist: 8(a) rests on the firm's status, and OTA follow-on production rests on a prototype awarded under an OTA. Phase III is the one that carries a prototype built under a different award instrument into a new contract. For a small firm with a real prototype and no vehicle, it is often faster than a schedule and cheaper than a GWAC.

The data rights are the other reason to start here. You keep title, and under the DFARS clause the government's license is narrow for twenty years. The civilian FAR clause still states four years, so the contract decides. The SBIR Phase III guide walks the clause detail. The contrast that matters for a vehicle decision is with non-SBIR DoD work, where government purpose rights typically run five years from the contract that required the development before converting to unlimited.

The limit. A Phase III lets you sell more of what you built to the agency that funded it. It does not give you a governmentwide storefront. Firms that assume Phase III solves distribution discover the same problem a year later. Our SBIR Phase III guide covers the commercialization sequence, and the Replicate stage covers the vehicle portfolio that has to exist alongside it.

OTA in detail

Other transaction authority lets DoD buy prototypes outside the FAR. The authority sits at 10 U.S.C. 4021 for research projects and 10 U.S.C. 4022 for prototype projects, and 4022 includes a follow-on production path for a successful prototype.

OTA is not a loophole. It is a deliberate instrument for work the FAR does not fit well, and it is used heavily by the Defense Innovation Unit and the service labs. There are three ways to qualify under 4022(d): you are a non-traditional contractor, every participant is a small business or non-traditional, or the participants cover at least one third of the cost. Two things make it attractive to a small software firm: the timeline can be months rather than the year a FAR competition takes, and non-traditional contractors are the point of the program rather than an exception.

Two limits. It is a DoD authority, with a few civilian agencies holding their own separate authority under different statutes, so it does not solve a civilian market. And a prototype award is not a product business; the follow-on production path is the thing to negotiate for, and it is easier to secure before the prototype succeeds than after.

The five traps

1. Stale advice. CIO-SP4 was cancelled in January 2026. Alliant 3 was announced the next month and live in March. MAS removed the price reduction clause. FedRAMP replaced its levels with classes. Any vehicle guide older than this year is a risk, including ours if you are reading it a year from now.

2. Treating a schedule as a sales channel. A schedule is a purchasing mechanism. If you have no demand generation, the schedule produces nothing and the minimum sales requirement comes due.

3. The five-year cliff on DoD non-SBIR work. Government purpose rights on noncommercial software typically run for five years from the contract, modification or option that required the development, then convert to unlimited rights. If your product depends on DoD data rights you did not assert, that clock is already running. The productization guide explains the assertion deadline, which is at proposal time.

4. The set-aside illusion. An 8(a), SDVOSB or WOSB vehicle is not a shortcut to a sale. It changes who can compete, not whether anyone has a requirement. STARS III, for example, is an IT services vehicle with a $50B ceiling and an ordering period that runs to July 2029, and it is still a competition at the task-order level.

5. Buying the vehicle before the product. The most expensive order of operations. A schedule modification plus a compliance program plus a build, before anyone has said they will buy, is how a $50,000 idea becomes a $700,000 program. The cost model shows the three scenarios side by side.

The next 30 days

  1. Write down the six inputs. If you cannot name the agency and the budget line, stop and do that first.
  2. Check whether you already hold a vehicle that can carry a product CLIN. Many firms do and have never modified it.
  3. If you have a Phase II, price the Phase III route before any other.
  4. If you are selling a product to civilian agencies, get the current MAS solicitation and read the SIN attachments for the one you would offer under.
  5. If a single agency is most of your market, ask that agency's small business office which vehicle its buyers actually use. The answer is often an IDIQ you have never heard of.

What this page does not cover

Proposal writing, capture management, and pricing-to-win. Those are a different discipline and a different set of costs. This page is about access: which door exists, what it takes to stand in it, and what it costs to stay there.

This page is dated on purpose. Re-check it after 1 November 2026, when SEWP VI ordering starts and NITAAC stops accepting new orders.

Where to start

If you want a first read on which vehicles you already hold and which are idle, that is what the Federal Activation Audit does: a full audit of every prime vehicle and certification with status and expiration dates, a procurement-forecast cross-reference of named open opportunities, and a 30/60/90-day activation roadmap. The published floor is from $12,500. If the question is still whether the product is worth selling, start with the free Signal diagnostic instead. It takes about eight minutes and will tell you if the answer is not yet.

Run the free HARBOR Signal diagnostic

30 questions. 8 minutes. Your firm's HARBOR Readiness Score + a live HARBOR Signal snapshot. No credit card.

Where this fits

This is one step in a larger problem. The Federal Activation Audit is the engagement that covers it. A federal infrastructure activation plan: a full vehicle and certification audit, named open opportunities, a concentration-risk read, and a 30/60/90-day activation roadmap.