Peterson's LLC

DOCUMENT FOR PUBLIC RELEASE

The decision issued on the date below was subject to a GAO Protective Order. This public version has been approved for public release without redaction.

Decision

Matter of: Peterson's LLC

File: B-424599; B-424599.2

Date: September 25, 2026

Jacqueline K. Unger, Esq., Katherine B. Burrows, Esq., Abigail H. Finan, Esq., and Kristine E. Crallé, Esq., Piliero Mazza, PLLC, for the protester.
Holly A. Roth, Esq., Gregory R. Hallmark, Esq., and Ben R. Smith, Esq., Holland & Knight LLP, for Vantage Point Consulting, Inc., the intervenor.
Brian L. Howell, Esq., Leslie Walter, Esq., and William Moorhouse, Esq., Department of Defense, for the agency.
Raymond Richards, Esq., and John Sorrenti, Esq., Office of the General Counsel, GAO, participated in the preparation of the decision.

DIGEST

Protest challenging the technical acceptability of the awardee's proposal is sustained where the awardee's proposal did not respond to a material solicitation requirement and the agency unreasonably concluded that the proposal met the requirement.

DECISION

Peterson's LLC, a small business of Greenwood Village, Colorado, protests the award of a contract to Vantage Point Consulting, Inc. (VPC), a small business of Portland, Oregon, under request for proposals (RFP) No. H9821026RE002, issued by the Department of Defense, Personnel Readiness Management Agency (PRMA)[1], for academic skills training (AST) for military service members. The protester challenges the evaluation of proposals and the resulting source selection decision.

We sustain the protest.

BACKGROUND

The Defense Activity for Non‑Traditional Education Support (DANTES), a component of PRMA, manages a portfolio of voluntary education programs to help military service members achieve personal and professional career goals. AR, Tab 5.3, Amended Performance Work Statement (PWS) at 441.[2] AST is a college and career readiness program supporting military service members' improvement in math, English, and science skills. Id. The goal of the requirement is to “continue the Defense Department's capacity to deliver AST designed for Service members needing to improve their academic readiness for the purposes of college‑level course placement, qualification for reclassification or advance[d] training, and military advancement.” Id. at 442.

The agency issued the RFP on February 23, 2026, as a small business set‑aside under the procedures of Federal Acquisition Regulation (FAR) parts 12 and 15, seeking proposals for a commercially available, computer adaptive, asynchronous web‑based course that will improve cognitive, metacognitive, and non‑cognitive academic skills for military service members. AR, Tab 4.1, Initial RFP at 289; AR, Tab 5.3, Amended PWS at 443; AR, Tab 5.4, Instructions to Offerors at 466. The RFP contemplated the award of a fixed‑price contract with a 1‑year base period and four 1‑year option periods, to the offeror whose proposal represented the best value to the government, considering the following evaluation factors, shown here in descending order of importance: (1) technical capability; (2) past performance; and (3) price. AR, Tab 4.1, Initial RFP at 291‑292 (period of performance); AR, Tab 5.5, Evaluation Criteria at 475‑476.

The technical capability factor was divided into the following four evaluation elements: technical approach and evaluation methodology; program management; staffing plan; and transition and execution plan.[3] AR, Tab 5.5, Evaluation Criteria at 477‑478. Under the technical capability factor, proposals would receive a factor-level rating on a scale of: outstanding; good; acceptable; marginal; or unacceptable.[4] AR, Tab 5.5, Evaluation Criteria at 478‑479. Under the past performance factor, proposals would be rated on a confidence scale of: substantial; satisfactory; limited; no confidence; or neutral. Id. at 480. Proposals were due on March 30. AR, Tab 6.1, RFP Amend. 2, at 511.

The agency received eight proposals by the due date, including proposals from Peterson's and VPC.[5] AR, Tab 1, Contracting Officer's Statement (COS) at 35‑36. The relevant evaluation results are as follows:

	Peterson's	VPC
Technical Capability	Good	Outstanding
Past Performance	Substantial	Substantial
Price	$14,741,299	$7,882,561

Id.; AR, Tab 13, SSD at 1261.

Based on the evaluation results, the source selection authority (SSA) determined that VPC's proposal represented the best value to the government. AR, Tab 13, SSD at 1289. In making this decision, the SSA explained that based on VPC's outstanding technical rating (the only offeror to earn the top rating), its substantial confidence past performance rating, and its total evaluated price which was lower than the next highest‑rated proposals (including Peterson's), VPC's proposal was “clearly the superior choice.” Id. On June 12, the agency awarded the contract to VPC. Id.; AR, Tab 2, MOL at 97.

On June 15, Peterson's requested a debriefing. AR, Tab 2, MOL at 97. The agency provided a written debriefing that same day. Id.; see AR, Tab 14, Debriefing at 1290‑1295. On June 22, Peterson's filed the instant protest with our Office.

DISCUSSION

Peterson's contends that the agency should have found VPC's proposal ineligible for award for failing to meet a material solicitation requirement and also generally challenges the evaluation of proposals and the resulting source selection decision. We have thoroughly reviewed all of the protester's challenges and find only one basis to sustain the protest, specifically that the awardee's proposal failed to demonstrate compliance with a material solicitation requirement to provide for offline course access. Accordingly, our decision focuses on the sustained protest ground and does not discuss the other, non‑meritorious challenges.

Where an evaluation is challenged, our Office will not reevaluate proposals but instead will examine the record to determine whether the agency's judgment was reasonable and consistent with the stated evaluation criteria and applicable statutes and regulations. IDEMIA Nat'l Sec. Sols., LLC, B‑421418, B‑421418.2, May 1, 2023, at 5; J&J Maintenance, Inc- dba J&J Worldwide Services, B‑423821.2, B‑423821.3, Apr. 20, 2026, at 4. It is a fundamental principle that a proposal that fails to conform to a material solicitation requirement is technically unacceptable and cannot form the basis for award. HumanTouch, LLC, B-419880 et al., Aug. 16, 2021, at 6; High Noon Unlimited Inc., B‑419268, Jan. 12, 2021, at 4 n.6.

In reviewing an agency's evaluation, we do not limit our review to contemporaneous evidence, but consider all of the information provided, including the parties' arguments and explanations. Netizen Corp., B‑418281 et al., Feb. 21, 2020, at 6‑7 n.5. Although we generally give little weight to reevaluations and judgments prepared in the heat of litigation, see Boeing Sikorsky Aircraft Support, B‑277263.2, B‑277263.3, Sept. 29, 1997, at 15, post‑protest explanations that provide a detailed rationale for contemporaneous conclusions and simply fill in previously unrecorded details will generally be considered, so long as those explanations are credible and consistent with the contemporaneous record. Remington Arms Co., Inc., B‑297374, B‑297374.2, Jan. 12, 2006, at 12.

Peterson's challenges the agency's evaluation of VPC's proposal under the technical capability factor, arguing that VPC failed to meet a material solicitation requirement to deliver offline course access. Protest at 14; Comments & Supp. Protest at 12-15; Supp. Comments at 10‑11. As discussed below, we sustain this protest ground.

Under the technical approach and evaluation methodology element of the technical capability factor, offerors were instructed to demonstrate their technical capability to meet the requirements outlined in the PWS. AR, Tab 5.4, Instructions to Offerors at 471. In turn, the agency was to evaluate proposals to assess “how well the Offeror demonstrated an understanding and ability to meet the PWS requirements” and “the extent of the Offeror's understanding and capabilities to provide an online academic skills training product containing the appropriate subject area content as required by the PWS.” Id. at 477 (relevant criteria for the overall factor and for the element).

Relevant here, the PWS included several requirements, one being for offline course access. AR, Tab 5.3, Amended PWS at 443‑454 (listing requirements that the Contractor “shall” meet or provide). The PWS requirement for offline course access read in full:

The course mobile application shall have the capability to be accessed and used in non‑network environments. This capability includes temporary storage of course progress and user data locally should the device not be connected to the internet. Once network access is available and the user logs into the application, the course progress and data shall automatically upload to and synchronize with the online hosted application. Video lesson content may be adjusted or curtailed below the 50 [percent] threshold described in [PWS] section 4.1.5 in the interest of device storage and capability. There is no requirement to maintain placeholder data for video content viewed in the offline mode.

Id. at 449 (PWS section 4.1.6, Offline Course Access).

In response to the RFP, VPC proposed a product called SPARK. AR, Tab 8.2, VPC Technical Proposal at 634. VPC described SPARK as an artificial intelligence (A.I.)‑powered course authoring, delivery, assessment and management system, tailored through VPC's A.I. software and training development engine for the DANTES AST mission. Id.

In evaluating VPC's proposal, the agency rated it as outstanding under the technical capability factor and assessed it with two significant strengths and five strengths under the technical approach and evaluation methodology element. AR, Tab 10.1, TEB Report at 1085‑1088. No weaknesses or deficiencies were noted. Id.

In its initial protest filing, Peterson's challenged VPC's technical acceptability in multiple ways. Relevant here, the protester argued that SPARK is offered through a VPC division called Solari, and that neither VPC nor Solari have a commercially available mobile application for the SPARK platform. Protest at 14. Peterson's argued that because neither VPC nor Solari have a commercially available mobile application for SPARK, VPC could not have met the PWS requirement for offline course access. Id. Peterson's argued that this, among other alleged flaws with the SPARK product, should have rendered VPC's proposal unawardable. Id.

The agency countered that offering a commercially available mobile application was irrelevant to the PWS requirement for offline course access. MOL at 79. The agency argued that “the PWS requires the contractor to provide a mobile application with offline storage and automatic synchronization capabilities during performance of the contract.” Id. The agency then claimed that VPC's proposal met the PWS requirement because, according to the agency, “VPC states its SPARK technical proficiency provides non‑internet services to [the] Department of Labor [(DOL)] Make America AI‑Ready initiative.” Id. (citing AR, Tab 8.2, VPC Technical Proposal at 638).

The portion of VPC's proposal discussing Make America AI‑Ready appears in proposal section 2.1, labeled “The Learning Science Has Moved.” AR, Tab 8.2, VPC Technical Proposal at 637. This section of the proposal responded to the technical approach and evaluation methodology element and included a discussion of how VPC's approach is in line with “[t]he most successful learning platforms in the world[.]” Id. For example, VPC discussed that successful platforms use short daily practice lessons paired with immediate feedback, adaptive difficulty, and gamified habit formation, and stated that such an approach is included with SPARK's methodology. Id. The proposal then included a blue text box titled “A Federal Agency Already Proved This Works[.]” Id. The blue text box is where Make America AI‑Ready is discussed and is what PRMA cites as evidence that VPC's proposal met PWS section 4.1.6 for offline course access. The text box read in full:

On March 24, 2026, the [DOL] launched its Make America AI‑Ready initiative: a nationwide workforce training program delivered entirely through daily text messages, with no video, no course shell, and no learning management system. Workers complete foundational AI skills training in 10 minutes a day over seven days, on any phone, including basic models, without internet access. The program was built through a public‑private partnership with an AI‑powered learning platform and is aligned to the [DOL's] AI Literacy Framework.

A cabinet‑level federal agency, operating under the same compliance and accessibility requirements as the Department of [Defense], has demonstrated that interaction‑based, non‑video learning reaches working adults at scale making use of flexible learner interactions like [HTML5[6]] simulated learner interactions that adapt to each and every learner. SPARK applies the same model to AST preparation; purpose‑built for the service members DANTES exists to serve.

Id. at 638.

Peterson's maintains that VPC's proposal did not demonstrate the required offline course access capability and therefore should have been assessed with a deficiency and found technically unacceptable.[7] Comments & Supp. Protest at 12‑15. Here, Peterson's makes three relevant points.

First, the protester argues that the offline course access requirement was a material solicitation requirement and notes that another proposal was evaluated as technically unacceptable for failing to meet it. Id. at 13 (citing AR, Tab 13, SSD at 1278).[8] Second, the protester notes that an individual technical evaluator assessed VPC's proposal with a deficiency for failing to address offline course access but the deficiency did not appear in the TEB report and the contemporaneous record did not explain why the deficiency was not included in the TEB report. Id. at 14‑15. Third, the protester argues that the portion of VPC's proposal cited by the agency as demonstrating VPC's offline course access capability did not state what the agency claimed it stated. Id. at 13. In this regard, the protester argues that VPC's reference to DOL's Make America AI-Ready initiative did not demonstrate VPC provided any services to the DOL whatsoever, nor did it demonstrate that the SPARK platform met the requirements of PWS section 4.1.6 for offline course access. Id. at 13‑14. In response, the agency maintained its position that it reasonably evaluated VPC's proposal. Supp. MOL at 7‑8.

To further develop the record in this area, our Office conducted a hearing where the TEB chairperson and the contracting officer (CO)/SSA testified regarding their contemporaneous evaluations and understanding of VPC's proposal. The hearing was testimonial in nature where the agency directly examined the witnesses, and the protester cross‑examined the witnesses. Hearing Transcript (Tr.) at 7.

Relevant to the first point raised by the protester, the TEB chairperson confirmed that the PWS requirement for offline course access was considered a material requirement. Tr. at 61 (cross‑examination); see also id. at 132 (CO/SSA testifying on cross-examination that offline course access was a requirement). Relevant to the second point raised by the protester, the TEB chairperson confirmed that another proposal was assessed a deficiency for failing to address offline course access. Id. at 29 (direct examination). The TEB chairperson also confirmed that in their capacity as an individual evaluator they initially assessed VPC's proposal with a significant weakness for failing to address offline course access and that the significant weakness was later elevated to a deficiency. Id. The TEB chairperson explained that ultimately, the deficiency was not included in the TEB report because another evaluator directed the TEB chairperson to the portion of VPC's proposal discussing DOL's Make America AI‑Ready initiative as evidence of VPC's offline course access capability. Id. at 29‑30 (direct examination).

Relevant to the third point raised by the protester, both the TEB chairperson and the CO/SSA testified that they concluded VPC's proposal met the offline course access requirement based on VPC's reference to the Make America AI‑Ready initiative and the language in the text box. See Tr. at 28‑33 (direct examination of TEB chairperson), 66 (cross‑examination of TEB chairperson), 109‑110 (direct examination of CO/SSA), 138‑139 (cross‑examination of CO/SSA). The TEB chairperson elaborated that they read VPC's proposal as stating VPC offered a product comparable to what was used at DOL under the Make America AI‑Ready initiative and that VPC's product demonstrated offline course access capability because the Make America AI‑Ready discussion mentioned the phrase “without internet access[.]” Id. at 30‑32 (direct examination), 69 (on cross‑examination, testifying: “We're going to do what they did. It's going to work offline. That's how we interpreted it.”).

The CO/SSA articulated a similar understanding of the proposal language. See id. at 139 (on cross‑examination, testifying: “they're providing the same model that was done with [DOL] . . . without internet”). The agency's post‑hearing comments confirm that the agency believed VPC provided services to DOL for Make America AI‑Ready, and that VPC's proposal demonstrated SPARK could function offline, as required, based on the discussion of Make America AI‑Ready. Agency Post‑Hearing Comments at 18. Thus, the record demonstrates that the agency's basis for finding VPC's proposal offered offline course access was the language in the text box referencing Make America AI‑Ready.

On this record, we find the agency's evaluation of VPC's proposal unreasonable. VPC's proposal did not address the material PWS requirement for offline course access, and the agency made unreasonable assumptions about what was being offered by VPC. Accordingly, we sustain this protest ground.

First, the agency concedes that proposing a product with offline course access was a material requirement. Tr. at 61; see also id. at 132. The importance of this requirement is further demonstrated by virtue of another proposal being found technically unacceptable for failing to address offline course access. See AR, Tab 10.1, TEB Report at 1085; AR, Tab 13, SSD at 1278.

Second, VPC's proposal did not address the PWS requirement for offline course access. See generally AR, Tab 8.2, VPC Technical Proposal. Thus, on its face, VPC's proposal did not meet a material solicitation requirement.

Third, the agency's conclusion that VPC's proposal, by virtue of its discussion of Make America AI‑Ready, met the material solicitation requirement, is unreasonable. Nothing in VPC's proposal states that VPC offered a product to DOL for Make America AI‑Ready or that VPC was involved with that initiative in any way. See id. at 638. The proposal states only that Make America AI‑Ready involved “a public‑private partnership with an AI‑Powered learning platform[.]” Id.

More salient, even if VPC was involved with Make America AI‑Ready,[9] VPC's text box discussing Make America AI‑Ready did not address the PWS requirement for offline course access in any fashion. The relevant proposal language appears in the context of evolving learning science and how, in VPC's opinion, short daily practice sessions with immediate feedback produce better learning outcomes than platforms that use long video lessons and static course models. Id. at 637‑638. The discussion of Make America AI‑Ready was largely focused on platform content and delivery methods. Id. at 638. To the extent that offline capabilities were discussed at all, it was a passing reference that included no context or elaboration as to how VPC's proposal would meet the detailed requirements of PWS section 4.1.6. Id. Thus, the agency's assumption that VPC's product could meet the offline course access requirement based on the discussion of Make America AI‑Ready was unreasonable.

To the extent the agency argues VPC did not offer the same product used for Make America AI‑Ready but rather, a product based on the “same model” used at DOL, we find the evaluation unreasonable. See e.g., Tr. at 31‑32 (suggesting that VPC offered a product in the same model as Make America AI‑Ready). As discussed previously, the Make America AI‑Ready reference did not include detail on how that product met any of the specific requirements of the instant RFP for offline course access. Furthermore, as the protester points out, if VPC proposed a product modeled on DOL's product, there would be other issues with VPC's technical acceptability. See Protester Post‑Hearing Comments at 17. For example, the instant RFP requires a “web‑based course” that “shall be available as an online course” whereas the Make America AI‑Ready product was “delivered entirely through daily text messages[.]” AR, Tab 5.3, Amended PWS at 443 (course requirements); AR, Tab 8.2, VPC Technical Proposal at 638. Thus, to the extent it was raised, we reject the argument that VPC's proposal met the requirement for offline course access because VPC's product was modeled on DOL's product.

In conclusion, PRMA unreasonably evaluated VPC's proposal for compliance with PWS section 4.1.6, which required offline course access capability. VPC's proposal did not address the requirement, the contemporaneous record does not explain how PRMA determined that VPC met the requirement, and PRMA's post‑protest explanation is unreasonable.

We also find that Peterson' s was prejudiced by the agency's unreasonable evaluation. Competitive prejudice is an essential element of every viable protest, and we will not sustain a protest unless the protester demonstrates that, but for the agency's error, it would have had a substantial chance of receiving award. Supreme Foodservice GmbH, B-405400.3 et al., Oct. 11, 2012, at 13‑14; see Tribalco, LLC, B-414120, B‑414120.2, Feb. 21, 2017, at 12 n.7 (discussing prejudice where an offeror failed to meet a material requirement and the agency did not waive the requirement).

Under a reasonable evaluation, there is a substantial possibility that VPC's proposal would have been found technically unacceptable for failing to meet a material requirement. See e.g., AR, Tab 13, SSD at 1278 (finding another proposal technically unacceptable for failing to address offline course access); HumanTouch, LLC, supra (a proposal that fails to conform to a material solicitation requirement is technically unacceptable and cannot form the basis for award.). Based on the standing of the remaining six technically acceptable proposals and the RFP's terms which established technical capability as more important than past performance, and the combined non‑price factors as significantly more important than price, there is a substantial possibility that, but for the agency's unreasonable evaluation, Peterson's would be in line for award. See AR, Tab 13, SSD at 1261 (evaluation results of all proposals); AR, Tab 5.5, Evaluation Criteria at 476 (relative weight of evaluation factors). Accordingly, we sustain the protest.

RECOMMENDATION

As detailed above, we conclude that PRMA's evaluation of VPC's proposal was unreasonable and that Peterson's was prejudiced by that unreasonable evaluation. We recommend that PRMA conduct a new evaluation of VPC's proposal under the technical capability factor consistent with the terms of the RFP, applicable procurement law and regulation, and this decision. The agency should adequately document the results of the reevaluation of VPC's proposal. If VPC's proposal is found technically unacceptable, we recommend the agency terminate VPC's contract for the convenience of the government, conduct a new best‑value tradeoff decision considering the remaining technically acceptable proposals, and make a new source selection decision.

We also recommend that Peterson's be reimbursed the reasonable costs of filing and pursuing its protest, including attorneys' fees. 4 C.F.R. § 21.8(d)(1). Peterson's should submit its certified claims for costs directly to the agency within 60 days after receipt of this decision. Id. § 21.8(f)(1).

The protest is sustained.

Edda Emmanuelli Perez
General Counsel

---