SP Systems, Inc. v. United States

OPINION AND ORDER GEORGE W. MILLER, Judge. This is a post-award bid protest action pertaining to a contract for programmatic support services within the Flight Project Directorate of the National Aeronautics and Space Administration (“NASA”) at Goddard Space Flight Center in Maryland. After an initial award to SP Systems, Inc., disappointed bidder ASRC Research and Technology Solutions, Inc. (“ARTS”) filed a protest with the Government Accountability Office (“GAO”). GAO sustained ARTS’ protest, concluding that NASA had used certain labor rates provided in connection with the Request for Proposal (“RFP”) in an unsupportable manner. GAO recommended that NASA conduct a limited re-evaluation of the proposals. NASA did so, and that re-evaluation resulted in an award of the contract to ARTS. SP Systems then filed this bid protest action, alleging various improprieties in GAO’s recommendation and in NASA’s reevaluation process. For the reasons stated below, the Court denies SP Systems’ protest. I. Background NASA currently contracts for the planning and managing of space flight projects, including managing flight hardware, ground systems, and launch vehicles, through a contract known as the Program Analysis and Control (“PAAC”) II contract. Decision of the Government Accountability Office, B-400217, B-400217.2, August 21, 2008 (“GAO Decision”) at 2, AR Tab 44 at 005624. In October of 2007, NASA issued an RFP for the PAAC III contract, which was to be awarded to the offeror “representing the best value to the government, considering three evaluation factors: mission suitability, past performance, and cost.” Id. In evaluating proposals, the mission suitability factor was most important, with cost and past performance given approximately equal weight. Id. at n. 2. A Mission Suitability Mission suitability was graded on a 1,000-point scale consisting of the following: 275 points for understanding the statement of work; 275 points for the technical approach to Representative Task Orders (“RTO’s”), 400 points for the offeror’s management plan, and 50 points for safety and health considerations. Id.; RFP at 116, AR Tab 7 at 000520. The RFP indicated that any plan to capture the incumbent workforce (that is, retain the personnel currently performing these tasks at NASA under the present contract) would be “evaluated for reasonableness and consistency with other parts of the proposal.” RFP at 113, AR Tab 7 at 000517-518. The offerors were required to propose, among other things, their direct labor rates for eighteen specified non-management labor categories, consisting of six primary categories (Scheduling Specialist, Configuration Management Specialist, Documentation Specialist, Information Technology Specialist, Project Support Specialist and Accounting Specialist), with three sub-categories for each (Junior, Intermediate, and Senior). GAO Decision at 3, AR Tab 44 at 005265. The offerors could propose different labor subcategories, but they had to identify the percentage of effort for each, which the Govern [4] ment used to calculate a “weighted” average direct labor rate for each of the 18 labor categories. Id. A reading library NASA provided to offer-ors contained “average labor rates” which “reflect the incumbent’s current direct labor categories and average ... unburdened direct labor hourly rates for those categories being used under the incumbent contract for PAAC II Services.” RFP at 89-90, AR Tab 7 at 000493-94. These average labor rates, referred to by all parties as the “library rates,” were “provided for information purposes only,” id., although the instructions to offerors who proposed to retain the incumbent workforce required them to “clearly explain variances from their proposed unburdened Offeror Direct Labor Rates ... and those of the current incumbent’s unburdened rates.” Id. (¶ L.15). In response to industry questions, NASA explained that the library rates were “straight” rather than “weighted” averages: that is, they “did not take into account the number of individuals in any specific category.” Responses to Industry Questions at 18, AR Tab 6 at 000395. Although mission suitability was to be evaluated separately from cost, NASA reserved the right to adjust an offeror’s mission suitability score based on a lack of cost realism, because this “may be a significant indicator of an Offeror’s understanding and ability to perform.” RFP at 116, AR Tab 7 at 000520; GAO Decision at 4. The RFP indicated in several locations NASA’s concern with retaining the personnel presently working on the contract. RFP at ¶¶ M.l(b), (c), M.4, AR Tab 7 at 000512, 000514. B. Past Performance The RFP also provided that each offeror would be evaluated on past performance, specifically “each offeror’s record (including the record of any significant subcontractors and/or teaming partners) of performing services or delivering products that are similar in size, content and complexity to the requirements of this solicitation.” RFP at 119, AR Tab 7 at 000523. Offerors were to provide information for relevant contracts and subcontracts exceeding $2 million in value that were performed or completed within the past three years. GAO Decision at 4, AR Tab 44 at 005266; RFP at 100, AR Tab 7 at 000504. Past performance was considered both for the substantive reviews given to the offeror’s performance and for the similarity of the contract to the present solicitation. RFP at 119, AR Tab 7 at 000523. On the criterion of similarity, a contract could be “highly relevant,” “very relevant,” “relevant,” “somewhat relevant,” or not relevant, based upon its “similarity in size, content and complexity” to the PAAC III contract. RFP at 119-20, AR Tab 7 at 000523-24. C. Initial Award and GAO Review NASA received four proposals and the Source Evaluation Board (“SEB”) drew the following conclusions: Mission Suitability Proposed Recommended Mission Adjectival Cost Plus Cost Plus Past Offeror_Suitability Rating_Award Fee_Award Fee_Performance SP Systems 914_Excellent $191,086,870.00 $192,448,817.00 Excellent A2_592_Good_$187,014,782.00 $189,426,766.00 Excellent ARTS_864_Very Good $176,914,763.19 $189,331,294.37 Excellent B 382 Fair $191,339,121.20 $196,763,915.36 Very Good GAO Decision at 4; SEB Final Report at 2, AR Tab 37 at 004186. The Source Selection Authority (“SSA”), based upon this eval [5] uation, concluded that SP Systems’ proposal represented the best value to the Government. Source Selection Statement at 17, AR Tab 39 at 004500 (“SP’s overall excellent Mission Suitability proposal coupled with their excellent Past Performance and competitive cost proposal outweighs the slight probable cost advantages proposed by ARTS and DB.”). ARTS received its debriefing on the contract award on May 28, 2008, and filed a bid protest with the GAO on June 2, 2008. Contracting Officer’s Statement of Facts, AR Tab 44 at 005062. ARTS argued (1) that NASA had improperly evaluated its cost proposal and wrongly assigned it a “significant weakness” for its management plan and (2) that NASA had failed to properly evaluate the similarity of SP Systems’ prior contracts to the PAAC III contract. GAO Decision at 5-6, AR Tab 44 at 005267-68. The GAO sustained portions of ARTS’ position on both contentions, for the following reasons. 1. Cost Realism and “Significant Weakness” in Management Plan ARTS’ cost realism and management plan arguments both related to NASA’s use of the library rates in evaluating its proposal. Those rates were as follows: Accounting Documentation MIS Project Support Scheduling Specialist CM Specialist Specialist Specialist Specialist Junior Level $16.46 $18.65 $14.74_$20.47 $19.49_$21.49 Mid Level $17.25 $24.57 $21.31_$26.65 $22.41_$29.83 Senior Level $24,24 $37.38 $29.16 $38.79 $29.91 $52.17 Procurement Reading Library Materials, AR Tab 8, at 00712. Both ARTS and SP Systems had in their proposals contemplated sub-levels beyond junior, intermediate and senior, as permitted by the RFP. GAO Decision at 7, AR Tab 44 at 005269. ARTS denominated the junior positions as level I, the intermediate positions as level II, and then proposed to divide the senior positions into three categories: levels III, IV, and V. Id. SP Systems proposed to divide all three levels (junior, intermediate, and senior) into three sub-levels. Id. Both ARTS and SP Systems then identified in their cost models “the percentage of effort for each sub-level, which was then used to calculate a “weighted’ average labor rate.” Id. It suffices for present purposes to note that although SP Systems created three sub-levels for each seniority category, the weighted averages of those rates matched the library rates. Id. at 8, AR Tab 44 at 005270. ARTS, on the other hand, chose to assume a slightly greater level of effort (that is, number of employees) was required in tiers III and IV than in level V of its senior-level subdivisions of the library rates. Id. at 7, AR Tab 44 at 005269. ARTS’ rates, when averaged without this weighting, were identical to the library rates. Id. at 10, AR Tab 44 at 005272. But when ARTS’ weighted averages were calculated for senior-level non-management positions, assuming a slightly greater number of employees in the lower tiers, those averages were less than the library rates. Id. at 7, AR Tab 44 at 005269. Both ARTS and SP Systems proposed to retain a high percentage of the incumbent workforce, and NASA was understandably concerned about the offerors’ ability to meet the salaries currently being paid by the incumbent contractor. Id. at 8, AR Tab 44 at 005270. But the GAO noted that “[ujnfortu-nately, NASA’s solicitation did not disclose the incumbent workforce’s actual labor rates to offerors, at least not in a meaningful way, and those conducting the evaluation may not have had access to the actual rates.”3 Id. at [6] 8-9, AR Tab 44 at 005270-71. That is, although the offerors knew the straight average of the labor rates paid to the employees in each category, there was no way for them (or the agency’s evaluating officials) to know whether the distribution of salaries in those labor categories was top- or bottom-heavy. But having no better information, NASA officials “relied on the unweighted library rates and treated them as reflecting the rates paid the incumbent workforce.” Id. at 9, AR Tab 44 at 005271. GAO rejected ARTS’ challenge to the use of the library rates in the cost realism analysis, concluding that it was reasonable for the agency to use the library rates to represent the cost of the incumbent workforce. The current labor costs would be constant for all offerors who proposed to retain members of the incumbent workforce (that is, all of the offerors), thus it was reasonable to utilize the library rates as a “plug number” to reflect that constant value, and to adjust ARTS’ cost upwards to meet that standard. Id. at 10, AR Tab 44 at 005272. Although GAO found the library rates had been properly used as a “plug number,” GAO disagreed with NASA’s decision to assign ARTS a significant weakness for the management plan subfactor because its rates were inadequate to capture the incumbent workforce. Id. at 10, AR Tab 44 at 005272. Although ARTS’ weighted average was lower than the library rate, unless one knew the incumbent’s actual weighted average labor rate, there was no way to tell whether ARTS’ weighted average was equal to, more than, or less than that of the incumbent. Id. at 11-12, AR Tab 44 at 005273-74. Specifically, GAO held: [W]e find no reasonable basis for the agency’s assignment of a significant weakness to ARTS’ proposal under the management plan subfactor. Absent more information, there is simply no way for the agency to determine whether the library rates or the weighted averages proposed by any offeror are closer to the incumbent’s direct labor cost. In addition, if one compares the weighted rates proposed by ARTS and SP Systems, it is not possible in many instances to determine whether one firm will be more or less likely to attract the incumbent workforce in any given labor category, since the firms proposed different high and low rates within labor categories, in some instances different numbers of levels of sub-categories, and different percentages of effort for each sub-level. For any given labor category, it may be that, as compared to the incumbent workforce, one offeror’s rates are high, the other’s are low; neither the evaluators nor the offer-ors had any basis to know. Id. at 11, AR Tab 44 at 005273. Because ARTS had elsewhere committed to, at minimum, paying incumbents their current wages, and “[gjiven the meaninglessness of the library rates as a criterion for retaining the incumbent workforce,” GAO found the assessment of a significant weakness against ARTS for risking failure to retain employees of the incumbent to be unreasonable. Id. at 12, AR Tab 44 at 005274. 2. SP Systems’ Past Performance Evaluation ARTS also challenged NASA’s evaluation of SP Systems’ past performance on the ground that the agency had not in fact graded past performance based upon the prior contracts’ “similarity in size, content and complexity” to the PAAC III contract. RFP at 119-20, AR Tab 7 at 000523-24. In fact, the GAO found, NASA had merely looked to see whether prior contracts exceeded the $2 million threshold, and deemed all contracts over that amount to be relevant. GAO Decision at 13, AR Tab 44 at 005275. GAO found this to be inconsistent with the RFP, which required NASA “to consider the relative size of offerors’ past performance references in weighing their past performance ratings and assessing whether the references were highly [7] relevant, very relevant, relevant, somewhat relevant, or simply not relevant at all.” Id. at 14, AE Tab 44 at 005276. Because NASA had not conducted this analysis, GAO concluded that the “determination that SP Systems’ past performance was ‘highly relevant,’ particularly given that SP Systems’ references were, in most respects, small fractions of the size of the contemplated PAAC III contract, was unreasonable.” Id. 3. GAO Recommendation The present protest rests to some degree on a challenge to the nature of GAO’s recommendation, which reads in relevant part as follows: We recommend that NASA reevaluate proposals by removing the ‘significant weakness’ improperly assigned to ARTS under the mission suitability factor and readjusting ARTS’ point scores and ratings accordingly, and reevaluating the relevance of SP Systems’ past performance. Based upon this reevaluation the agency should make a new source selection decision. If, after the new evaluation, NASA determines that another firm’s proposal represents the best value to the government, the agency should terminate SP Systems’ contract and make a new award. GAO Decision at 14, AR Tab 44 at 005276. Unsuccessful offeror DB Consulting filed a protest with the GAO on August 28, 2008, protesting its exclusion from what it characterized as reopened competition. AR Tab 45 at 005316 Getter from James H. Roberts III to the Hon. Gary L. Kepplinger dated August 2, 2008). GAO dismissed this protest as “premature and academic.” DB Consulting Mot. for Judgment on the Administrative Record (docket entry 26, Dec. 12, 2008) (“DB Consulting Mot.”) at 4. GAO issued its decision on August 22, 2008. On September 4, 2008, counsel for SP Systems and ARTS (the parties to the GAO proceeding), received permission to share a redacted version of the GAO decision with their clients. Ex. C to Def.’s Mot. For Judgment. The relevant portion of the remedial paragraph of the redacted version read as follows: We recommend that NASA reevaluate proposals by removing the [REDACTED] improperly assigned to ARTS under the mission suitability factor and re-adjusting ARTS’ point scores and ratings accordingly, and reevaluating the relevance of SP Systems’ past performance. Based upon this reevaluation, the agency should make a new source selection decision. If, after the new evaluation, NASA determines that another firm’s proposal represents the best value to the government, the agency should terminate SP Systems’ contract and make a new award. Def.’s Mot. for Judgment on the Administrative Record (docket entry 35, Dee. 19, 2008) (“Def.’s Mot.”) at 22. D. NASA Action Following GAO Decision On Friday, September 26, 2008, Richard McCarthy of NASA sent an email to GAO and counsel for ARTS and SP Systems stating, in its entirety, “Dear All: In the matter of the corrective action in the PAAC III procurement, be advised that NASA hopes to have a new source selection by the end of next week. Thank you.” Exh. C to Def.’s Mot. 1. SEB Addendum to Final Report On September 29, 2008, the SEB completed an addendum to its final report regarding the PAAC III solicitation. PAAC III Addendum to Final Report, AR Tab 44 at 005280. The supplemental report reflected the SEB’s reassessment of both the “significant weaknesses” assigned to three of the offerors for failure to match the library rates and the past performance reviews. As for the management plan, NASA removed the significant weaknesses assigned not only for ARTS, but also for offerors DB Consulting and [* * *] (which is not a party to this proceeding), as a result of their “risk of failing to capture incumbents based on proposing rates below the library rates.” Id. at 2, AR Tab 44 at 005281. SP Systems had no technical weaknesses or significant weaknesses relating to incumbent capture, and so no changes were made to its score. Id. The scoring table for Mission Suitability Subfac [8] tor C-Management Plan thus read as follows]_ SP DB ARTS [***] Significant Strength 2 12 2 Strength_4 3 4_4 Weakness_0 2 1_1 Significant Weakness 0 0 0 2 Deficiencies 0 0 0 0 Addendum to Final Report at 2, AR Tab 44 at 005281. The removal of a significant weakness from ARTS’ evaluation “enabl[ed] the SEB to consider an adjectival rating of Excellent” for ARTS on Subfactor C. Id. To be consistent, the SEB also rescored Subfactor C for both DB Consulting and [* * *]. Id. at 3. But in both instances, the significance of the offerors’ remaining weaknesses in Subfactor C resulted in an unchanged adjectival rating of “Very Good” for DB Consulting and “Fair” for [* * *]. Id. When these revised Subfactor C calculations were combined with the offerors’ prior scores on the other Mission Suitability sub-factors, SP Systems remained the highest scorer, receiving 914 out of a possible 1000 points. Id. ARTS was now close behind, with a score of 912 (as opposed to the 864 awarded in the first evaluation). DB Consulting received 604 points, up from 592, and [* * *] only 414, improved from 382. Id.; SEB Final Report at 2, AR Tab 37 at 004186. In re-examining the past performance ratings, NASA noted the GAO’s recommendation and the terms of the RFP before “carefully reexamin[ing] the size of the SP Systems’ Team’s past performance record.” Addendum to Final Report at 4, AR Tab 44 at 005283. There were six past performance references for SP Systems: five for SP itself, and one for “its much larger team member, Orbital Sciences Corporation.” Id. The SEB observed that “no SP Systems Team member provided past performance data for a contract that was close in monetary value to the solicitation’s maximum ordering value of $200M, or provided past performance data for a previous contract that had close to the estimated 844 employees required for the PAAC III effort.” Id. The SEB did, however, favorably consider five contracts SP Systems performed as an 8(a)4 prime contractor worth $2.5 to $3.5 million, along with a $33.7 million contract performed by Orbital Sciences Corporation. Id. These contracts “demonstrate[ ] success on similar multi-million dollar contracts — one of them valued at 17% of the maximum value of PAAC III — which the SEB considered a significant effort in terms of size.” Id. Ultimately, the SEB concluded that: While the degree of similarity of size between the SP Systems Team’s referenced past contracts and the solicitation is not great in terms of pure dollars, the SEB determined that SP Team members’ successful performance on those lower dollar contracts, first hand as prime contractors, sufficiently reflects the team’s ability to perform on larger contracts, and this demonstrated experience is relevant to this analysis.... Though the SEB has some concerns about the degree of similarity between the size of the SP Systems’ Team’s past performance references and the PAAC III procurement, in the RFP’s overall relevance analysis, which takes into consideration size, content, and complexity, those concerns are not sufficient to offset our determination that the SP Systems Team has highly relevant experience. Id. at 5, AR Tab 44 at 005284. In a footnote, the SEB also observed that in reviewing the materials once more, it had found a discrepancy between the past performance questionnaires and the SEB’s first report regarding the relevancy analysis for ARTS. Id. at 4 n. 1. That is: Specifically, the final report did not reflect the SEB’s complete consideration and evaluation of the questionnaires, in that the narrative did not fully describe all relevancy ratings submitted by the evaluators. While [ARTS] received ‘significantly relevant’ ratings in all areas of the [Statement of Work], [ARTS] also received ‘moderately relevant’ and ‘minimal rele9 vanee’ or ‘did not perform’ ratings in other surveys. Nonetheless, the SEB hereby confirms that its original relevancy evaluation for the [ARTS] contracts was accurate (i.e., that it considered all survey ratings presented in the questionnaires), and that its overall relevancy evaluation for the entire ARTS team remains unchanged. Id. The final past performance ratings for all four offerors were: Overall Offeror Adjectival Rating SP Systems, Inc. Excellent DB Consulting Group, Inc. Excellent ASRC Research and Technology Solutions, LLC [ARTS] Excellent [*•**]Very Good SEB Debrief Package, AR Tab 44 at 005297. 2. John Wolfgang Memoranda Two memoranda signed by SEB Chairman John Wolfgang appear in the record, dated September 30, 2008, the day after the addendum to the SEB’s final report. AR Tab 44 at 005298-005305. The purpose and import of these memoranda are unclear, and they are the focus of many of the parties’ arguments in this protest. Both memos are titled “Process used by the PAAC III Source Evaluation Board,” the first “to reevaluate Mission Suitability and probable cost in accord with GAO ruling” (hereinafter “Wolfgang Rate Memo”) and the second “to reevaluate Past Performance in accordance with GAO ruling” (hereinafter ‘Wolfgang Past Performance Memo”). The first of these memos, the “Rate Memo,” concerned the rescoring of the management plans (Subfaetor C of Mission Suitability). The memo observed that the SEB removed the significant weakness from ARTS’ proposal as recommended by GAO and “[a]s required by the FAR and NFS, for consistency, the SEB also removed, for the same reason, a Weakness from DB’s findings and a Significant Weakness from [* * *]’s findings.” Wolfgang Rate Memo at 005298. Mr. Wolfgang’s memorandum continued: The SEB was concerned about correctly identifying cost adjustments based on what the GAO had declared to be an improper reference for incumbent capture. Therefore we performed a cost analysis of senior labor rates based on the actual incumbent labor rate for senior personnel from the PAAC II contract 533 data.5 The PAAC II total actual weighted average cost per senior unloaded manpower is $37.16 per hour. The reference Government provided average senior labor rate is $36.34. ARTS bid weighted average unloaded manpower at $34.55 per hour, SP bid weighted average unloaded senior manpower at $37.88 per hour, DB bid weighted average unloaded senior manpower at $37.29 and [* * *] bid unloaded weighted average unloaded manpower at $36.33. Thus ARTS underbid every hour of senior manpower by $2.61 and [* * *] underbid every hour by $0.83. Both DB and SPS overbid the average unloaded labor rates. The SEB decided to strictly follow the recommendations of the GAO ruling and not reassign a Significant Weakness based on actual incumbent manpower cost from the 533 data since these cost deltas were similar to those reported in our original report determination. Wolfgang Rate Memo at 005298. As discussed in detail below, ARTS and the Government argue, inter alia, that this paragraph is nothing more than the internal and unverified musings of a single member of the SEB (albeit its Chairman). They contend that neither the SEB nor the SSA made a decision based upon the incumbent’s “total [10] actual weighted average” labor rates, nor could they have done so without amending the RFP to disclose those evaluation criteria and allowing the offerors an opportunity to revise their proposals accordingly. But NASA could not disclose the incumbent’s weighted average labor rates to the offerors because the incumbent contractor, which was also bidding on the contract, asserted that this information was proprietary. Thus, the defendant and defendant-intervenor argue, NASA had no other option but to follow the recommendation of the GAO and eliminate the weaknesses assigned to offerors for deviating from the library rates. But SP Systems and DB Consulting counter that the memorandum is signed by the Chairman of the SEB and is clearly part of the analysis the SEB undertook. As for the Government’s inability to use the information, they make several counterarguments. At oral argument, the parties appeared to be in agreement that the Government could not have simply used the Wolfgang calculations in its reconsideration of proposals without amending the solicitation to disclose those new criteria and allowing the offerors to revise their proposals in light of that amendment. Oral Arg. Tr. at 12, 58-59. DB Consulting argues that this is precisely what should have occurred, in order to ameliorate this and various other problems DB Consulting sees with the award process.6 Id. at 99-100. SP Systems, on the other hand, asserts that the Government need not disclose all of the criteria it will use to evaluate proposals. Therefore, the Government could (and should) have used the Wolfgang numbers to “verify” that the library rates were a reasonable approximation of the incumbent’s actual weighted average labor rate — that is, create the correlation between the library rates and the incumbent’s actual weighted average labor rates that the GAO found was improperly missing from the earlier record. After having remedied this deficiency in its first analysis, NASA should, according to SP Systems, have assigned a justified significant weakness to ARTS for significantly underbidding the library rates as a proxy for the incumbent’s actual weighted average labor rates. In these slightly differing ways, plaintiff and plaintiff-intervenor each contend that it is arbitrary and irrational for the Government, knowing (as a result of the Wolfgang Rate Memo) that ARTS had underbid the incumbent’s actual labor rates, to simply ignore that information and award the contract to ARTS anyway. To the extent that the Government claims it was bound by the GAO decision to act as it did, SP Systems and DB Consulting argue that the GAO’s remedy was both overly restrictive and irrational. The Court will evaluate the merits of the parties’ competing positions below. The second Wolfgang memorandum partakes of a similarly uncertain and controversial status. SP Systems contends that this memo, which concerns the re-evaluation of the offerors’ past performance records, supports a downgrade in ARTS’ past performance rating. In the memo, Mr. Wolfgang notes that since the GAO ruling the SEB had identified additional past performance information for all the offerors except for ARTS “which has no past performance that the SEB could find.” Wolfgang Past Performance Memo at 005303. Ultimately, he concluded that the ARTS “team has contracts of the size and complexity of the PAAC III work and have an excellent rating for performance, but their relevance to the PAAC III work is only moderate which is problematical. In addition ARTS itself, the prime contractor who will be responsible for over 50% of the work across the contract, at best has neutral management or past performance history.” Id. at 005305. SP Systems argues that the continued rating of “excellent” for ARTS’ past performance is plainly inconsistent with the conclusion of the Wolf [11] gang Past Performance Memo that ARTS’ previous contracts were only moderately relevant to the present solicitation. These arguments will likewise be addressed below. On October 14, 2008, the SSA, George Barth, Deputy Director of Flight Projects for Planning and Business Management, signed an Amended Selection Statement awarding the PAAC III contract to ARTS. AR Tab 44 at 005306. In making his decision, Mr. Barth indicated that he had considered the SEB’s Final Report, the SEB’s Addendum Report, a presentation from the SEB, along with “the views of senior NASA/Goddard personnel who heard the presentation and who have responsibilities related to this procurement.” AR Tab 44 at 005307. This Amended Selection Statement largely echoed the reasoning and conclusions of the SEB Addendum, and made no reference to the Wolfgang memo-randa or the controversial statements therein. Ultimately, Mr. Barth observed that ARTS and SP Systems presented the best proposals, with significantly higher Mission Suitability scores than DB Consulting or [* * *]. ARTS and SP both received Excellent past performance ratings and were otherwise essentially equivalent, except that ARTS proposed the lowest total evaluated probable cost. He thus concluded that ARTS’ proposal “represents the best value to the Government” and selected ARTS as the contract awardee. AR Tab 44 at 005308. SP Systems was advised of the award to ARTS on October 28, 2008, AR Tab 45 at 005398, and received a debriefing on November 3, 2008. AR Tab 45 at 005417. DB Consulting sent an email requesting a debriefing to an incorrect email address, and it was therefore ultimately received too late for it to qualify for a debriefing. AR Tab 45 at 005438, 005452. SP Systems filed its complaint protesting the award of the PAAC III contract to ARTS on December 1, 2008 (docket entry 1). This complaint was filed prior to SP’s receiving the Administrative Record, and thus before it knew of the existence of the Wolfgang memo-randa. That original complaint alleged that: (1) “[t]he reevaluation conducted by NASA, as constrained by the narrow remedy mandated by the GAO decision, was contrary to the terms of the RFP,” which required consideration of compensation levels in section M, Id. ¶¶ 52-62, and, quoting, FAR 15.206,7 complained that “NASA never amended the solicitation to allow offerors to submit revised proposals based on the changed evaluation criteria that NASA used in the second round of the procurement,” Id. ¶¶ 63-64; (2) that the removal of the significant weakness assigned to ARTS was improper, irrational and violated the terms of the solicitation because “[biased upon the best releasable information in its possession, NASA’s decision to spread the labor force equally among the sub-levels was rational and proper” (ie., assume the incumbent’s actual weighted average was the same as the unweighted average), and the original decision was thus correct, Id. ¶ 68; (3) that ARTS did not have a Defense Contract Management Agency approved accounting system, Id. ¶¶ 73-79; (4) that ARTS itself has no experience as a prime contractor managing a contract of the size of the PAAC III, and thus the “excellent” rating ARTS received for past performance was unwarranted, Id. ¶¶ 80-94, and “constrained by the GAO decision and the unreasonably narrow relief recommended by GAO, NASA only reevaluated SP Systems’ past performance, not ARTS’ ”, Id. ¶¶ 95-99. ARTS intervened as a party-defendant on December 2, 2008 (docket entry 13), and DB Consulting intervened as a party-plaintiff on December 10, 2008 (docket entry 24). On December 18, 2008, SP Systems filed an amended complaint that withdrew Count III (relating to DCMA approval of ARTS’ accounting system) (docket entry 31). SP Systems expanded its allegations in Count I, arguing that “NASA ... did not exercise any independent judgment as required by FAR 15.305, but instead mechanically res-cored proposals as GAO required. Had GAO properly allowed NASA the discretion to reevaluate ARTS’ proposal, rather than simply mandating that the Significant Weak [12] ness be removed, NASA would almost certainly have found the Significant Weakness was justified. This conclusion is supported by the fact that NASA’s second cost analysis validated the library rates and the conclusions reached in the first source selection decision.” Id. ¶ 126. It also made further allegations regarding Count IV, namely that ARTS’ partner, [* * *], was to perform only [* * *] percent of the work, but ARTS’ past performance evaluation was primarily based on [* * *]’s history, Id. ¶ 144, and that although the SEB had found upon re-evaluating the data that ARTS’ past performance was only moderately relevant, ARTS nonetheless received an “excellent” rating. Id. ¶¶ 149-153. The amended complaint also added a fifth and sixth count, titled “Failure to Follow Findings of Second SEB Evaluation/Improper Source Selection Decision,” Id. ¶ 155, and “Failure to Provide SSA Information Necessary to Make a Rational Decision/Failure to Consider Information/Failure to Document.” Id. ¶ 170. These counts are based upon the portions of the Wolfgang memoranda discussed in detail above, and the alleged failure of the SEB to use that information or communicate the data contained in those memoranda to the SSA. II. Standard of Review To resolve these cross-motions for judgment on the administrative record, this Court makes factual findings based upon the administrative record. Bannum, Inc. v. United States, 404 F.3d 1346, 1356 (Fed.Cir.2005). Ultimately, the Court must determine whether the aggrieved offerors have established that the agency’s procurement decision was “arbitrary, capricious, an abuse of discretion or otherwise not in accordance with law.” 28 U.S.C. § 1491(b)(4); 5 U.S.C. § 706(2)(A); Impresa Construzioni Geom. Domenico Garufi v. United States, 238 F.3d 1324, 1332 (Fed.Cir.2001); ManTech Telecomm. & Inf. Sys. v. United States, 49 Fed.Cl. 57, 64 (2001). The Court may set aside an agency’s contract award if “(1) the procurement official’s decision lacked a rational basis; or (2) the procurement procedure involved a violation of regulation or procedure.” Impresa, 238 F.3d at 1332. A challenge, like this one, alleging that the agency’s decision was irrational, requires the Court to examine “whether the contracting agency provided a coherent and reasonable explanation of its exercise of discretion, and the disappointed bidder bears a heavy burden of showing that the award decision had no rational basis.” Id. at 1333-34. This requires a showing that the agency “entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” Keeton Corrs., Inc. v. United States, 59 Fed.Cl. 753, 755 (2004) (quoting Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983)). The agency’s discretion increases where the procurement at issue is negotiated and the still further when the contract is awarded on a “best value” basis. Burroughs Corp. v. United States, 223 Ct.Cl. 53, 617 F.2d 590, 597-98 (1980); Galen Med. Assocs. v. United States, 369 F.3d 1324, 1330 (Fed.Cir.2004); E.W. Bliss Co. v. United States, 77 F.3d 445, 449 (Fed.Cir.1996). Even more deference is due to the agency when the challenge involves the agency’s evaluation of bidders’ past performance. Overstreet Elec. Co. v. United States, 59 Fed.Cl. 99, 117 (2003) (noting the “triple whammy of deference”). This Court will interfere with a Government procurement “only in extremely limited circumstances.” CACI, Inc.-Fed. v. United States, 719 F.2d 1567, 1581 (Fed.Cir.1983). Where there are errors in the procurement process, they must “significantly prejudice” the plaintiff. Alfa Laval Separation, Inc. v. United States, 175 F.3d 1365, 1367 (Fed.Cir.1999). To establish “significant prejudice” a plaintiff must show that there was a “substantial chance” it would have received the contract award but for the errors. Id. at 1367. The GAO has already reviewed the first round of this procurement. This Court is not, of course, “bound by the views of the Comptroller General nor do they operate as a legal or judicial determination of the rights [13] of the parties.” Burroughs Corp., 617 F.2d at 597. But another pertinent principle appears in Honeywell, Inc. v. United States, 870 F.2d 644 (Fed.Cir.1989), where the Federal Circuit addressed this court’s role in reviewing agency action taken in conformity with a recommendation by the GAO. The appeals court observed that “[ajgencies traditionally have deferred to GAO recommendations, and as a general policy have acceded to the views of the GAO even when those views conflicted with the agency’s original position,” id. at 647, and that Congressional policy supported the conclusion that “a procurement agency’s decision to follow the [GAO] recommendation, even though that recommendation differed from the contracting officer’s initial decision, was proper unless the [GAO] decision itself was irrational.” Id. at 648. In every case, there is “a zone of acceptable results.” Id. As articulated in E.W. Bliss Co. v. United States, 33 Fed.Cl. 123, 134-35 (1995), aff'd, 77 F.3d 445 (Fed.Cir.1996), “Honeywell cannot be read to supplant Burroughs, nor to confer on the GAO a degree of deference beyond that delimited in Burroughs.” This Court should not conduct a de novo review of the issues decided by the GAO, but instead may only inquire whether the GAO decision was rational and the agency justifiably relied upon it. Honeywell, 870 F.2d at 647 (“[T]he Claims Court impermissibly undertook what can fairly be characterized only as its own independent de novo determination....”); Centech Group, Inc. v. United States, 79 Fed.Cl. 562, 575 (2007), aff'd, 554 F.3d 1029 (Fed.Cir.2009) (“In deciding whether an agency was justified in following a GAO recommendation, this Court determines whether GAO’s decision was rational. The Court’s review is not de novo.”); Interstate Rock Prods., Inc. v. United States, 50 Fed.Cl. 349, 363 (2001) (“[T]o the extent that an agency chooses to follow the advice of the GAO, courts should only intervene if the advice the agency receives is ‘irrational.’ ”); Firth Const. Co., Inc. v. United States, 36 Fed.Cl. 268 (1996) (“If the GAO’s advice is rational, it is not arbitrary or capricious to follow it” but concluding GAO advice was, in fact, irrational because the “GAO draws a legal conclusion with no principled support”).8 If the GAO recommendation is, for example, plainly contrary to a statutory requirement, that decision is irrational and an agency action is not justifiably based upon it. Grunley Walsh Int’l, LLC v. United States, 78 Fed.Cl. 35, 44 (2007). In order to prevail under Honeywell, therefore, a plaintiff must show “either that the agency corrective action constituted a clear violation of applicable statutes or regulations, or that the [agency] acted arbitrarily and capriciously in following GAO’s decision because GAO’s decision was itself irrational.” Centech, 79 Fed.Cl. at 574. If the agency’s judgment is reasonable, then this Court cannot substitute its judgment for that of the agency, even if the record might support a different result. ManTech, 49 Fed.Cl. at 63. Plaintiff argues that the Honeywell standard of review did not survive the passage of the Administrative Dispute Resolution Act (“ADRA”) in 1996. 28 U.S.C. § 1491(b). Relying primarily on IMS Services, Inc. v. United States, 33 Fed.Cl. 167 (1995), a pre-award bid protest, SP Systems claims that “agencies must perform their own analysis of whether the GAO is correct, and if not correct, follow the proper course in performing subsequent procurement actions.” SP Systems Opposition to Motions for Judgment on the Administrative Record (docket entry 48, Dec. 31, 2008) (“SP Opp.”) at 12. SP Sys14 terns also quotes IMS Services, a decision by Judge Horn, for the following proposition: Although noncompliance with a GAO recommendation may not be the preferred action of the agency, it may be the correct action. Therefore, it is the agency’s responsibility to fully and independently evaluate all recommendations given by the GAO. While this court recognizes that a procurement agency normally will accept the advice of the GAO, it is imperative that the agency perform its own evaluation of the procurement process before making final decisions. If, in its own expertise, the procurement agency determines that the GAO’s recommendation is misguided, it has a responsibility to make up its own mind and to act on its own advice. If the agency acts contrary to the GAO decision, the agency must report the noncompliance to the GAO for report to the Congress, as required by 31 U.S.C. § 3554(e)(1). Ultimately, this court must determine whether or not the Navy acted properly and was not arbitrary and capricious during the procurement, including a review of the agency decision to follow the GAO recommendation and to reopen another round of BAFOs. 33 Fed.Cl. at 184. This Court does not view this language as supporting plaintiffs argument, nor does it regard Honeywell, Burroughs and the ADRA as inconsistent with one another. The cases continue to reaffirm the viability of Honeywell long after the passage of the ADRA. See, e.g., Centech Group v. United States, 554 F.3d 1029, 1039 (Fed.Cir.2009). Honeywell itself is entirely consistent with the notion that the Court reviews the agency action, not the GAO decision itself. 870 F.2d at 647 (“The question before the Claims Court was whether the Army justifiably followed the GAO’s recommendation_”). While Honeywell remains viable and applicable precedent, the Court does not agree with defendant’s argument that in the present case the “award can only be overturned if the GAO’s decision is irrational.” Defendant’s Reply in Support of its Motion for Judgment upon the Administrative Record (docket entry 56, Jan. 9, 2009) (“Def.Reply”) at 2. The question presented is whether the agency’s procurement decision was reasonable based upon the record before it, and the GAO decision is only part of that record. If the GAO makes a rational recommendation and the agency simply implements that recommendation, then the agency action itself has a rational basis. In the rare instance where the agency considers a GAO decision to be so thoroughly wrong as to be irrational (or the court later concludes the GAO decision is irrational), then the agency is not justified in relying upon the decision. Thus, inquiring after the rationality vel non of the GAO decision is, where the agency action is solely based upon that decision, examining whether there exists a rational basis for the agency’s acts. In this case, however, there exist both the GAO decision and further analysis by NASA subsequent to the GAO decision that involved data that had not been before the GAO; the GAO decision and that subsequent analysis together constitute the record upon which the second source selection decision was made. III. Analysis A. Timeliness Although in their papers ARTS and the Government argue that SP Systems’ protest is untimely because it knew or should have known all of the pertinent facts as of September 4, 2008, when the redacted GAO decision was released, see, e.g., Def.’s Mot. for Judgment at 12, 22-25; ARTS Motion for Judgment on the Administrative Record (docket entry 38, Dec. 19, 2008) (“ARTS Mot.”) at 6-7, both parties retreated from that position at oral argument. They were wise to do so. “Section 1491(b) gives this Court jurisdiction to review an agency procurement decision, not the GAO’s review of that agency procurement decision.” Centech v. United States, 78 Fed.Cl. 496, 507 (2007) (“GAO’s actions are not reviewable in this Court.”). A challenge to the terms of a solicitation containing a patent error must be raised prior to the award or the ability to raise such an objection is waived. Blue & Gold Fleet, L.P. v. United States, 492 F.3d 1308, 1313 (Fed.Cir.2007). Where the terms of the solicitation have not changed, and the [15] agency has done nothing other than receive a non-binding recommendation from the GAO, there is no agency action to be protested. On September 4, 2008, the parties had no knowledge of any agency action either accepting or rejecting the GAO recommendation. Indeed, NASA could, theoretically, have determined that it would not follow GAO’s advice, and any protest would have been — as was DB Consulting’s original filing — “premature and academic.”9 DB Consulting Mot. at 4. At argument, ARTS and the Government placed great reliance on the September 26, 2008 group email from NASA reading “Dear All: In the matter of the corrective action in the PAAC III procurement, be advised that NASA hopes to have a new source selection by the end of next week. Thank you.” Exh. C to Def.’s Mot. They contend that this email informed SP Systems that a “new source selection” would be made, which was sufficient to put SP Systems on notice that NASA would “likely” strictly follow the GAO recommendation, Def.’s Mot. at 23, which was enough to require SP Systems to file a bid protest arguing that the GAO recommendation was irrational. The principal case relied upon, Centech, involved quite different circumstances. In that case, the GAO issued a decision recommending that the Air Force “reopen discussions and request and review revised proposals, evaluate those submissions consistent with the terms of the solicitation, and make a new source selection decision.” 78 Fed.Cl. at 501 (quoting GAO recommendation). The Air Force issued an amendment to the solicitation stating that it was “issued in response to the GAO recommendation to re-open discussions with all offerors while mindful” of other GAO comments. Id. Centech, the prior awardee, brought suit challenging the issuance of the amendment. Judge Williams found that the agency’s amending the solicitation constituted a de facto rescission of the previous award to Centech, and Centech’s lawsuit was brought as a preaward challenge to the revised terms of the solicitation. Judge Williams denied the Government’s motion to dismiss the lawsuit and found that the court possessed jurisdiction over the Air Force’s corrective action. In this case we are not dealing with a preaward lawsuit, and at least some of the argument is over the agency’s failure to amend the solicitation. Centech, therefore, does not govern the outcome of this case. ARTS and the Government also cite another GAO decision that is somewhat closer to the present situation (albeit without the involvement of GAO’s timeliness rule, 4 C.F.R. § 21.2(a)(2)). In Domain Name Alliance Registry, B-310803.2 (August 18, 2008), the offeror, DNAR, filed a post-award protest alleging that the agency had improperly failed to hold discussions with it prior to award. The agency offered to re-evaluate DNAR’s proposal, thus mooting the GAO protest. The new evaluation panel sent a letter asking for certain financial information, which DNAR sent. In the middle of April, the agency indicated to DNAR that it would make a new award “near the end of the month.” DNAR wrote to the contracting officer prior to the award expressing its shock that the award would again be made without conducting discussions. DNAR nonetheless waited until after the award to file another protest regarding the failure to hold discussions. GAO concluded that “[o]n this record, we think DNAR knew or should have known that the agency did not intend to hold discussions ... [and] could not reasonably await the agency’s second award decision without raising any challenge.” The Court cannot find that the circumstances here reasonably notified any of-feror that NASA was about to “strictly” implement the GAO decision. Bare notice that a second award decision would be forthcoming demonstrated that the agency intended to do something, but there was no indication whether the agency intended to take the action that GAO recommended, intended to do as SP Systems hoped and “verify” the library rates to affirm its initial decision, or [16] intended to take some other course entirely. The September 26, 2008 email was insufficient to place the offerors on notice of the agency’s future action and did not trigger a requirement that they file a pre-award bid protest. Thus, this post-award protest is timely. B. Mission Suitability As SP Systems has succinctly summarized, “[t]he entire case, with respect to the Mission Suitability evaluation, comes down to whether or not NASA was wrong in eliminating ARTS’ Significant Weakness for Mission Suitability in NASA’s second evaluation and award decision.” SP Opp. at 26. ARTS adds that “this protest is quite narrow. SPS concedes that NASA’s Source Evaluation Board (‘SEB’) and Source Selection Authority (‘SSA’) did exactly what GAO told them to do.” ARTS Reply in Support of Motion for Judgment on the Administrative Record (docket entry 54, Jan. 9, 2009) (“ARTS Reply”) at 9. While that is ultimately true, the route that NASA took in deciding to implement the GAO recommendation was circuitous, and therein lies the problem. To the extent plaintiff and plaintiff-interve-nor contend that the GAO decision was irrational on its face as issued, the Court disagrees. See, e.g., SP Systems Motion for Judgment on the Administrative Record (docket entry 29, Dec. 12, 2008) (“SP Mot.”) at 22 (proposing various alternative actions that GAO might have taken). Even assuming it would have been reasonable of GAO to require NASA to evaluate the proposals with reference to the incumbent’s actual weighted average labor rates, if what the GAO recommended is also rational, then the theoretical possibility of a different rational recommendation does not avail the plaintiff. Keco Indus., Inc. v. United States, 203 Ct.Cl. 566, 492 F.2d 1200 (1974); ManTech, 49 Fed.Cl. at 80 (“Let there be no mistake that if this court, rather than the Army, were constructing the corrective action here, a different plan might be adopted. But that is not this court’s prerogative.”). But given the Government’s representations to the GAO that actual weighted average labor rates were unavailable to be used in evaluating proposals, the GAO’s recommended course of action — eliminating weaknesses assigned due to the failure of the offerors’ weighted average labor rates to match the library rates, when the library rates themselves could not be correlated to the incumbent’s actual weighted average labor rates — was not only rational, but entirely appropriate given the facts known to GAO at the time the decision was made. The crux of the matter lies in a difference between the facts as represented to the GAO and information developed by Mr. Wolfgang following the GAO decision. To the extent that Mr. Wolfgang relied upon or created new data not presented to the GAO — data that the GAO believed was not available — the question arises whether that changes the analysis of NASA’s second source selection decision. Defendant asserts the Wolfgang numbers are puzzling and unclear and confusing. True enough. But they represent a detour from the course of action recommended by the GAO, and that departure suggests a different course of action NASA might have taken. In the words of Mr. Wolfgang, the GAO “declared” the library rates “to be an improper reference for incumbent capture.” Wolfgang Rate Memo, AR Tab 44 at 005298. “Therefore,” he continued, “we performed a cost analysis of senior labor rates based on the actual incumbent labor rate for senior personnel from the PAAC II contract data.” Id. GAO had no reason to know such calculations were possible, given its conclusion that “those conducting the evaluation may not have had access to the actual rates.” GAO Decision at 9, AR Tab 44, at 005271. The result of Mr. Wolfgang’s calculations included a finding that “[t]he PAAC II total actual weighted average cost per senior unloaded manpower is $37.16 per hour.” Wolfgang Rate Memo, AR Tab 44 at 005298. Mr. Wolfgang then went on to compare that rate (whatever it means) to single rates for each offeror, including ARTS, SP and DB. It is unclear, as the Government points out, what this comparison shows, since every offeror proposed at least six rates for senior manpower, not one. It is also uncertain whether the “actual” rate, however it was derived, [17] precisely correlates to the offerors’ proposals, which were based upon job classifications in the RFP that had lower education and experience requirements than the incumbent’s categories. AR Tab 44 at 000703; AR Tab 9 at 00823. Additionally, “[t]he accuracy of the alleged ‘actual incumbent rate’ ... is impossible to measure.” Def.’s Mot. at 31 (citations omitted). After making these computations, Mr. Wolfgang reached the following conclusion, which is only as reliable, of course, as its underlying data: “Thus ARTS underbid every hour of senior manpower by $2.61 and [* * *] underbid every hour by $0.83. Both DB and SPS overbid the average unloaded labor rates.” Wolfgang Rate Memo, AR Tab 44 at 005298. The ultimate question facing the Court is whether the agency’s procurement decision was reasonable based upon the information available to the agency at the time of the second source selection. As discussed above, if the agency simply acts to implement a GAO decision, and that GAO decision is rational, then the agency’s actions have a rational basis. But when the agency does more than implement the decision, and injects into the record new data not presented (indeed, the availability of which was denied) to the GAO, then the GAO decision is only part of the record available to the agency in making its second source selection decision. As applicable here, the question is: given the existence of the Wolfgang Rate Memo, what reasonable options were available to the agency? The existence of one rational course of action does not, of course, preclude the rationality of a second, different, course of action. If there were multiple rational possibilities, was the agency’s choice among those options reasonable? One of the possible alternate courses of action is proposed by DB Consulting. As noted above, the parties concede that NASA could not have used the Wolfgang-derived “actual incumbent” numbers to directly evaluate the offerors’ proposals without amending the solicitation and allowing the offerors an opportunity to amend their submissions. Oral Arg. Tr. at 12, 58-59, 99-100. But the Court cannot conclude that it would have been irrational of the agency to actually proceed to amend the solicitation and allow new submissions. Indeed, DB Consulting’s argument is that the GAO recommendation was irrational precisely because it did not require that course of action. While the Court disagrees with DB Consulting’s contention that the GAO’s recommendation was irrational, amending the solicitation is one of the possible rational alternative courses of action. SP Systems sets forth a different possibility. SP asserts that NASA ought to have relied upon Mr. Wolfgang’s calculations as “verifying” its original use of the library rates as a proxy for the incumbent’s actual weighted average labor rates. The concluding sentence of the paragraph forms the foundation of SP Systems’ argument. It reads, “The SEB decided to strictly follow the recommendations of the GAO ruling and not reassign a Significant Weakness based on actual incumbent manpower cost from the 533 data since these cost deltas were similar to those reported in our original report recommendation.” Wolfgang Rate Memo, AR Tab 44, at 005298. In essence, SP Systems construes that sentence to mean that: (1) Mr. Wolfgang “verified” that the library rates were in fact close enough to the actual weighted average incumbent manpower costs to form a useful basis for comparison; and (2) having so “verified” the library rates, he concluded that the “cost deltas were similar to those reported” in the original award decision. Because those original “cost deltas” were sufficient to support the imposition of a Significant Weakness on ARTS’ proposal, SP asserts, and Mr. Wolfgang confirmed those “cost deltas” to be correct, NASA should logically have reimposed the Significant Weakness on ARTS’ proposal. SP Systems interprets the word “since” in the last clause of the sentence as implying that only the GAO’s instructions prevented NASA from taking this, the most reasonable, course of action.10 [18] Let us assume, arguendo, that the agency had, in fact, discovered a true, documented, and verifiable correlation between the library rates and the incumbent’s actual weighted average labor costs. What was it to do? Plaintiff argues that this analysis would have corrected the error identified by the GAO by “verifying” that the library rates sufficiently reflected the incumbent’s actual weighted average labor costs. The agency could then simply have reinstated its earlier analysis, reimposed a significant weakness on ARTS, and again awarded the contract to SP Systems. The Court disagrees. It is well settled that a Government procurement decision must be based upon the “source selection criteria in the solicitation.” FAR 15.308; 15.305(a) (“An agency shall evaluate competitive proposals and then assess their relative qualities solely on the factors and subfactors specified in the solicitation.”); 10 U.S.C. § 2305(b)(1) (“The head of an agency shall evaluate sealed bids and competitive proposals and make an award based solely on the factors specified in the solicitation.”). The Government may not “evaluate[ ] the proposals received on a significantly different basis than announced in the solicitation.” Hydro Eng’g, Inc. v. United States, 37 Fed.Cl. 448, 471 (1997); CACI Field Servs., Inc. v. United States, 13 Cl.Ct. 718, 728 (1987). If the Government changes any evaluation criterion after issuing the solicitation, it must amend the solicitation and notify the offerors of the changed requirement. FAR 15.206(a); Candle Corp. v. United States, 40 Fed.Cl. 658, 663 (1998).11 Plaintiff is correct that the Government is clearly within bounds to “comparte] ... previous Government and commercial contract prices with current proposed prices for the same or similar items, if both the validity of the comparison and the reasonableness of the previous price(s) can be established.” 48 C.F.R. 15.404—1(b)(2)(ii); id. at (b)(2)(v) (permitting “Comparison of proposed prices with independent Government cost estimates.”). Where the RFP discloses that a particular subject is “an area of evaluation,” then the Government may use “an undisclosed reasonable estimate” in grading proposals. DynCorp, et al., B-257037, 95-1 CPD ¶ 34 (Dec. 15, 1994); OMV Medical, B-281490, 99-1 CPD ¶ 38 (Feb. 16, 1999) (“[T]here is nothing improper in an agency’s comparing proposed prices with an undisclosed government estimate.”). There are limitations upon the Government’s use of such an undisclosed estimate, however. The agency “must be able to demonstrate the basis for the estimate.” Nutech Laundry & Textile, Inc. v. United States, 56 Fed.Cl. 588, 594 (2003) (remanding for recalculation of unsupported estimate). If an of-feror’s proposal is significantly different than the undisclosed estimate, the agency should “conduct discussions” with the offeror to determine whether the proposal is nonetheless realistic. P.E. Sys., Inc., B-249033.2, 92-2 CPD ¶ 409 (Dec. 14, 1992) (“[W]here an agency uses such an undisclosed government estimate, the agency should conduct discussions with an offeror whose proposal substantially deviates from the government estimate in order to learn the reasoning behind the offeror’s particular approach and to determine whether the offeror’s proposed staffing can in fact satisfy the government’s requirements.”); Doss Aviation, B-275419, 97-1 CPD ¶ 117 (Feb. 20, 1997) (“[I]t is inappropriate to determine the acceptability of proposals by the mechanical application of an undisclosed estimate.”). Furthermore, where a solicitation contains estimates that offerors reasonably interpret in different ways and price their bids according to those differing interpretations, the agency may not “remedy perceived (or supposed) deficiencies in solicitation-stat [19] ed estimates by evaluating bids under criteria different from those upon which the bids were solicited.” Northern Virginia Van Company v. United States, 3 Cl.Ct. 237, 240 (1983). Northern Virginia Van approved cancelling an ambiguous solicitation. The solicitation at issue here was not ambiguous, in that it stated the library rates were for information purposes only and the Government clarified in response to industry questions that the library rates were unweighted averages of the incumbent’s labor rates. The GAO recommended that the agency remove an unjustified weakness assigned on one sub-factor due to the agency’s unreasonable use of the library rates and then rescore the proposals. That does not require the use of the library rates in a manner inconsistent with the RFP’s representation that the library rates were unweighted averages provided for information purposes only.12 Given the foregoing principles, NASA faced some difficulties in merely using the Wolfgang numbers to “verify” the library rates and reimpose its earlier award. First, it had to be able to “demonstrate the basis” for the Wolfgang numbers. The record does not provide a sufficient basis for the Wolfgang calculations that the purported “total actual weighted average cost per senior unloaded manpower” could have been properly used as an undisclosed estimate. Thus further work would have had to have been done. Second, the parties conceded that the Wolfgang numbers could not have been used to evaluate the proposals, but SP Systems nonetheless argues they could be used to “verify” the library rates. This merely seeks to accomplish indirectly what cannot be done directly. Third, the Government might have had to conduct discussions with any offeror whose labor rates substantially deviated from the estimate. Fourth, the offerors had based their proposals on the representations in the RFP — as verified by the GAO — that the library rates were unweighted averages to be used for information purposes only. During the ARTS bid protest, the Government argued that reliance on the library rates for comparison to the offeror’s weighted average labor rates was reasonable, and the GAO disagreed. ARTS’ offer reflected its belief that the unweighted average library rates were not correlated to the incumbent’s actual weighted average rates, and GAO agreed. There was nothing in either the RFP or the information presented during proceedings before the GAO that contradicted that belief. It would represent a blatant bait and switch for NASA, post-protest, to simply justify its original use of the library rates, in the face of the GAO decision, without disclosing its “verification” of the library rates to the offerors and providing them an opportunity to amend their proposals. L-3 Comm. EOTech, Inc. v. United States, 83 Fed.Cl. 643, 652-53 (2008); Def.’s Mot. at 36 (“[T]he incumbent’s actual labor rates were not included in the solicitation and had no place in the evaluation of ARTS’ or any other offeror’s proposal.”). Given the offerors’ reliance on the disclosed straight-average library rates in preparing their proposals, a full, fair procurement decision that utilized actual weighted average labor rates, even simply to “verify” the library rates’ correlation to the actual weighted averages, would have required amending the solicitation and providing the offerors an opportunity to amend their proposals. FAR 15.305(a). To do otherwise would be a “quintessential example of procurement conduct that lacks a rational basis.” L-3 Comm. EOTech, 83 Fed.Cl. at 653 (quoting Hunt Building Co., Ltd. v. United States, 61 Fed.Cl. 243, 273 (2004) (“Technical evaluations should be consistent with the factors, subfactors, and procedures outlined in the solicitation.”)). The cases upon which the plaintiff relies in arguing to the contrary do not involve disclosed estimates that were then either changed or used for a purpose other than the one revealed in the solicitation. DynCorp; Lockheed Support Sys., Inc.; VFB Joint Venture; Ogden Allied [20] Eastern States Maintenance Corp., B-257037, et seq., 95-1 CPD ¶ 34 (December 15, 1994), Sea-Land Service, Inc., B-253068, et al., 93-2 CPD ¶ 84 (Aug. 5, 1993); OVM Medical, Inc., B-281490, 99-1 CPD ¶ 38 (Feb. 16, 1999). These cases certainly do not authorize taking the library rates, which were identified in the first phase of the procurement process as the incumbent’s straight average labor rates, and using them to evaluate proposals as though they represented the incumbent’s weighted average labor rates, particularly after the GAO had stated that the library rates were not properly treated as the equivalent of weighted average rates. The Court therefore concludes that merely using the Wolfgang Rate Memo or further research to “verify” the library rates and reinstate the first award was not an option open to NASA. The agency thus faced a choice between two possibilities. It could either (1) amend the solicitation in some form and allow revised submissions (e.g., to indicate the changed use of the library rates); or (2) strictly follow the GAO decision.13 NASA chose to strictly follow the GAO decision and the Court must decide whether it was irrational in doing so. Plaintiff and plaintiff-intervenor argue that the decision to proceed with the award to ARTS was unreasonable in light of the information NASA possessed at the time the second source selection was made, specifically the calculations in the Wolfgang Rate Memo. They allege that those calculations undermined the GAO’s conclusion there was no proof ARTS had significantly underbid the actual labor rates, therefore opening a “Pandora’s box” that NASA could not shut. Consol. Eng’g Serv., Inc. v. United States, 64 Fed.Cl. 617, 628 (2005). The relevant inquiry in examining how an agency implements a GAO recommendation, however, is whether the agency has “stated a reasonable basis” for its actions. SDS Int’l, Inc., B291183.4, 2003 CPD ¶ 127 (April 28, 2003) (‘While, as the protester suggests, a more comprehensive method of implementing our recommendation could have been chosen by the Air Force, including reopening discussions and obtaining revised proposals and reevaluating past performance, the agency has stated a reasonable basis for not performing these actions.”). The Government asserts that it was rational in choosing to strictly follow the GAO’s recommendation instead of amending the RFP for several reasons.14 First, the cost realism analysis for this cost-reimbursement contract was undertaken separately, with all offerors’ costs adjusted upward to meet NASA’s estimate of the actual costs. Thus, the Court cannot conclude it was irrational for NASA to determine ARTS’ proposal posed little risk to the Government. Cubic Applications, Inc. v. United States, 37 Fed.Cl. 345, 359 (1997) (in cost-reimbursement contract, “normalization of the proposed compensation costs suggests that the Army intended to pay the costs necessary to maintain a sufficient workforce, regardless of which contractor prevailed”). Further, NASA had stated that retention of the incumbent workforce would be evaluated based on offerors’ Total Compensation Plans, AR Tab 10 at 000833, and ARTS’ Total Compensation Plan had been rated as “excellent.” AR Tab 44 at 005273-74. That evaluation included the factors in Section M.l of the RFP that SP Systems asserts were not considered in the second source evaluation, SP Sys. Mot. at 27-28, and the Total Compensation Plan evaluation was not disturbed during the GAO prot [21] est.15 Neither ARTS nor SP Systems were assessed a penalty in either phase of the source selection for their cost realism proposals. All that is at issue in this protest is the propriety of a cost realism adjustment to the evaluation of Mission Suitability Subfactor C-Management Plan. RFP at 116, AR Tab 7 at 00520. Second, and importantly, the meaning of the calculations in the Wolfgang Rate Memo is unclear, and their accuracy is “impossible to measure.” Def.’s Mot. at 81. As noted above, plaintiff claims that the Wolfgang Rate Memo “verified the accuracy of the library rates,” SP Mot. at 9, but the record simply fails to support that contention. Mr. Wolfgang made calculations of some sort based on data not within the record, and came to less-than-elear conclusions that do not correlate either to the library rates or to any rate proposed by the offerors (e.g., a single rate rather than six rates). It is possible that if Mr. Wolfgang had done a complete and well-documented analysis based upon reliable data that corresponded to the estimates offerors provided in response to the RFP, it could have supported plaintiff’s argument; then again, maybe not. Plaintiff’s contention that the Wolfgang Rate Memo itself mandates a decision in its favor is without merit. The best-case scenario for plaintiff is that the Wolfgang Rate Memo somehow created an obligation on the part of file agency to further investigate the possibility that these preliminary calculations might, if completed, verify the library rates’ correlation to the actual weighted average labor rates of the incumbent contractor and that in turn would create a requirement that the agency incorporate them into its source selection decision. Keco Indus., 492 F.2d at 1202-03 (“[T]he very most that can be said, on this record, is that the interested Air Force officials had access to data which, if they had considered it (which they did not), might well have raised questions about the technical feasibility of Acme’s proposal.”). Plaintiff points the Court to no authority supporting any such chain of obligation, and the law supports the agency’s discretion in methods of implementing a GAO recommendation. Consol. Eng’g, 64 Fed.Cl. at 628; SDS Int'l, Inc., B-291183.4, 2003 CPD ¶ 127 (April 28, 2003). NASA could not, as noted above, have actually made any source selection decision based upon the incumbent’s actual direct weighted average labor rates without amending the RFP and permitting revised proposals. The upshot of plaintiffs position is that NASA could rationally have used the incumbent’s actual direct labor rates in some manner, and the calculations in the Wolfgang Rate Memo were clear enough for NASA to have relied upon them. See SP Mot. at 13-14 (arguing that the SSA “could have ration [22] ally concluded that ARTS’ proposed labor rates did not support its goals of incumbent retention,” “could have rationally concluded that ARTS had failed to explain how it would achieve those goals” and “a reasonable reevaluation could have determined that it was appropriate for the SEB to assign ARTS a Significant Weakness”). SP’s argument appears to be an attempt to shift the burden of proof to the Government. The law is otherwise, however. SP Systems was required to demonstrate that NASA’s action was irrational; that is, show the Court that the Wolfgang Rate Memo was relevant and accurate and that NASA was required to use it in the re-evaluation or conduct further investigation. Career Training Concepts, Inc. v. United States, 83 Fed.Cl. 215, 220 (2008) (concluding preponderance of the evidence is the appropriate burden of proof upon protestor); Hawaiian Dredging Constr. Co., Inc. v. United States, 59 Fed.Cl. 305, 317 (2004) (“In a bid protest the Government does not have the burden of proof to show that the contracting officer acted reasonably in rejecting a bid. The burden is entirely plaintiffs.”). Even if plaintiff were correct that NASA could rationally have used the information in the Wolfgang Rate Memo, that is insufficient to meet plaintiffs burden to demonstrate that NASA was irrational in not relying upon the numbers contained in the memo. The Court does not find the agency’s decision to strictly comply with the GAO recommendation, rather than amending the solicitation and allowing revised proposals, to be irrational or unreasonable. SP Systems argues that the award decision might have been different if NASA had used the information contained in the Wolfgang Rate Memo. But a bid protestor cannot meet its burden merely by demonstrating that the record might have supported a different result. ManTech, 49 Fed.Cl. at 60. If NASA’s corrective action was rational and lawful, hypothetical corrective action that NASA could have but did not pursue is irrelevant. Id. at 80; Cambridge Sys., B-400680, B-400680.3 (Jan. 8, 2009). The GAO recom-mendation itself, as stated earlier, was also rational. Thus, the agency possessed a reasonable basis for its second source selection decision and the agency’s corrective action was “reasonable under the circumstances,” DGS Contract Serv., 43 Fed.Cl. 227, 238 (1999), and “appropriate to remedy the impropriety,” Rockville Mailing Service, Inc., B270161, 96-1 CPD ¶ 184 (April 10, 1996). SP Systems also argues that the second source selection decision is fatally flawed because the information contained in the Wolfgang Rate Memo was not contained in the SSA’s Source Selection Statement. Because the RFP had not been amended, however, those numbers (whatever they connoted) were not properly considered in making the final source selection decision.16 The SSA documented his determination to follow the GAO recommendation, and his statement provided a rational basis for that decision. The Court cannot conclude, based upon the Administrative Record, that it was unreasonable for NASA to decide that the Wolfgang Rate Memo was too slender a reed upon which to rest a deviation from the GAO’s recommendation. Thus, the Court rejects SP Systems’ challenge to NASA’s determination to follow the GAO recommendation by removing, in the second round of the PAAC III procurement, the significant weakness assigned in the first round to ARTS in grading Mission Suitability Subfactor C (the management plan) and then rescoring the proposals. C. Past Performance NASA considered offerors’ past performance both for the substantive reviews given to the offeror’s performance and for the similarity of the contract to the PAAC III contract. RFP at 119, AR Tab 7 at 000523. On the criterion of similarity, a contract could be “highly relevant,” “very relevant,” “relevant,” “somewhat relevant,” or not relevant, based upon its “similarity in size, content and complexity” to the PAAC III contract. RFP at 119-20, AR Tab 7 at 000523-24. In other [23] words, there were two criteria and three sub-criteria set forth in the RFP: (1) substantive reviews of performance and (2) similarity of experience. Similarity of experience was then subdivided into (a) size, (b) content, and (c) complexity. ARTS challenged NASA’s evaluation of SP Systems’ past performance by alleging that the agency had failed to evaluate the size of SP Systems’ prior contracts in assigning SP Systems a past performance rating of “excellent.” GAO Decision at 13, AR Tab 44 at 005275. The GAO concluded that there was no indication in the record that NASA had considered the size of the contracts in determining whether they were, for example, highly relevant, moderately relevant, or not relevant. Following this instruction, the agency revisited SP Systems’ previous contracts and concluded that “any size concerns are not sufficient to offset the highly relevant content and complexity of the experience offered by the SP Systems Team.” AR Tab 44 at 005307. SP Systems was thus reassigned a past performance rating of “excellent.” That is, although SP Systems was not particularly strong on criterion (2)(a), size, it was strong enough on criteria (2)(b) and (2)(c) to merit an “excellent” rating. SP Systems now contends that in the reevaluation, ARTS’ past performance rating should have been reduced below “excellent.” With respect to ARTS, the Wolfgang Past Performance Memo observed that “their team has contracts of the size and complexity of the PAAC III work and have an excellent rating for performance, but their relevance to the PAAC III work is only moderate which is problematical.” AR Tab 44 at 005305. SP Systems interprets this sentence as stating that ARTS’ past performance was, overall, only “moderately relevant,” and NASA improperly failed to take that moderate relevance into account. Although the sentence is not as clear as it might be, SP Systems’ reading is nonetheless not the fairest interpretation. The more reasonable reading of the Wolfgang Past Performance Memo is that ARTS is excellent on criterion (1), substantive reviews, and also does well on criteria (2)(a), similarity of size, and (2)(c), similarity of complexity. ARTS did not, however, fare as well on criterion (2)(b), similarity of content (that is, “relevance to the PAAC III work”). Just as with SP Systems’ own evaluation, although one of the similarity subfactors was less strong than the others, the agency chose to assign ARTS an “excellent” rating for past performance. The Court sees nothing unreasonable about that decision. Consol. Eng’g, 64 Fed.Cl. at 637 (past performance evaluation “will not be disturbed unless it is unreasonable or inconsistent with the terms of the solicitation or applicable statutes or regulations”). The agency acted reasonably given the information it possessed at the time of the second source selection decision. This is not a case where the agency simply ignored information known to it; NASA analyzed the past performance reviews and made a considered decision about the rating to assign. SP Systems also argues that ARTS’ past performance evaluation is disproportionately and unreasonably based upon the history of one of its teaming partners, which is proposed to perform only [* * *] percent of the work on the PAAC III procurement. The RFP states, however, that “The Past Performance factor will evaluate each offeror’s record (including the record of any significant subcontractors and/or teaming partners).” RFP at 119, AR Tab 7 at 000523. A “significant subcontractor” is “any subeon-tract[or] and/or teaming partner whose proposed subcontract value exceeds $2M of the total proposed cost.” RFP at 100, AR Tab 7 at 000504. The team partner whom SP Systems challenges, [* * *], was to perform services worth more than $2 million, and qualified as “significant” under the RFP. The RFP does not indicate that the past performance evaluation will be weighted based upon the percentage of the work to be performed, or whether the subcontractor’s work is central to the overall contract. While the FAR states that a past performance evaluation “should take into account past performance information regarding ... subcontractors that will perform major or critical aspects of the requirement when such information is relevant to the instant acquisition” this language does not impose a mandatory requirement to weight the evaluation by the “criti24 cality” of the subcontractor’s work. FAR 15.305(a)(2)(iii); Singleton Enters., B-298576, 2006 CPD ¶ 157 (Oct. 30, 2006). The consideration of [* * *]’s past performance history was required by the terms of the RFP, and the weight to be given that history was within the discretion of the agency. Weidlinger Assocs., Inc., B-299433, 2007 CPD ¶ 91 (May 7, 2007) (“Regarding the relative merits of offerors’ past performance information, this matter is generally within the broad discretion of the contracting agency, and our Office will not substitute our judgment for that of the agency. A protester’s mere disagreement with the agency’s judgment does not establish that an evaluation was improper.”) (citation omitted). The Court therefore rejects SP Systems’ challenge to NASA’s determination that ARTS’ past performance merited a rating of “Excellent.” IV. Entitlement to Injunctive Relief Because neither SP Systems nor DB Consulting has achieved success on the merits, the Court “need not examine the other factors for injunctive relief.” Centech, 79 Fed.Cl. at 577. CONCLUSION For the foregoing reasons, the motions of ARTS and the Government for judgment on the administrative record are GRANTED and the motions of SP Systems and DB Consulting are DENIED. The Clerk is directed to enter judgment in favor of defendant and defendant-intervenor. Some information contained herein has previously been designated by the parties as protected information subject to the protective order entered in this action on December 2, 2008 (docket entry 15). This opinion shall therefore be filed under seal. The parties shall review the opinion to determine whether, in their view, any information should be redacted prior to publication in accordance with the terms of the protective order. The parties shall file, within 10 days of the filing of this opinion, a joint report identifying the information, if any, they contend should be redacted, together with an explanation of the basis for their proposed redactions. IT IS SO ORDERED. Footnotes [2] “A” is offeror DB Consulting, which is a party to this case as plaintiff-intervenor, but did not participate in the ARTS bid protest at the GAO. [3] GAO Decision at 9 n. 8, AR Tab 44 at 005271 ("Absent further explanation from NASA, we fail to understand why weighted average labor rates would be considered proprietary to the incumbent and could not have been shared with offer-ors."); Agency Supplemental Memorandum at 7-8, ARTab 44 at 005144-45 (“Furthermore, more detailed data on incumbent pay, such as their weighted direct labor rates, reflecting each of the incumbent’s own current wage sub-categories. [6] cannot be published because it is proprietary to the incumbent (SGT). SGT would not release its weighted direct labor rates, especially here, where SGT competed as a subcontractor to keep the contract. Therefore, the incumbent's current straight average labor rates were the agency's best available and most objective standard against which each offeror was initially compared.”). In the PAAC III procurement, Stinger Ghaffarian Technologies (“SGT”) was a subcontractor to offeror DB Consulting. [4] Because the PAAC III solicitation was an 8(a) set-aside, AR Tab 7 at 000452, the minimum amount for relevant contracts was reduced during the pre-solicitation phases from $3 million to $2 million. Responses to Industry Questions at 3, ARTab 6 at 000380. [5] NASA regulations require the submission of "533" reports from contractors, referring to Form 533 M, which "provides monthly data on actual and planned costs and labor hours, short-term cost projections, estimates to complete, and contract values” and Form 533 Q, "which provides quarterly time-phased cost and labor hour estimates." NPR 9501.2D. 533 data was requested by offerors during the Industry Questions phase but denied on the ground that the data was proprietary. Responses to Industry Questions, AR Tab 5 at 000374. Counsel for the Government represented that the 533 reports did not contain the "total actual weighted average cost per senior unloaded manpower” referred to by Mr. Wolfgang, but that he instead performed some manipulation of the 533 data to arrive at the rate set forth in his memorandum. Oral Arg. Tr. at 9-10, 49, 105. [6] DB Consulting makes a number of other arguments about alleged defects in the award process that were not raised in the earlier bid protest or in plaintiff's complaint in this court. Because DB has decided to stand on SP Systems' complaint, DB Consulting Mot. at 2, we will not consider claims that were not also raised by SP Systems. Casa de Cambio Comdiv. S.A. de C.V. v. United States, 291 F.3d 1356, 1366 (Fed.Cir.2002) (holding that claims not raised in plaintiffs complaint are waived). [7] FAR 15.206 is titled “Amending the Solicitation,” and reads "When, either before or after the receipt of proposals, the Government changes its requirements or terms and conditions the contracting officer shall amend the solicitation” (emphasis added). [8] See also MTB Group, Inc. v. United States, 65 Fed.Cl. 516, 524 (2005) ("Defendant argues that an agency is required to consider a GAO decision when making procurement determinations. Therefore, it extrapolates, the reasonableness of an agency’s decision to follow the GAO decision should set the metes and bounds for judicial review. This latest approach to limiting the court’s exercise of its bid protest jurisdiction would confine review to determining whether an agency rationally relied on and followed a GAO decision. This defies logic: If such were the scope of review, then any agency action based on a GAO decision would qualify as rational based solely on the GAO decision. In fact, while the Court of Federal Claims affords deference to a GAO decision and does not conduct review de novo, the court’s charge is to determine, based on the record before the contracting officer, whether an agency’s procurement decision was reasonable.") (internal citations omitted). [9] The same flaw undercuts the Government’s argument that DB Consulting's protest seeking to reopen competition is untimely because the GAO informed it, in the September 8, 2008 decision dismissing DB Consulting’s protest, that GAO had only recommended a limited re-evaluation of ARTS' and SP Systems’ proposals. Def.’s Mot. at 25-27. [10] While the Court finds the memorandum difficult to understand, SP’s proposed interpretation would require Mr. Wolfgang to have misunderstood the basis for the GAO decision, which did not express any opinion about whether the cost deltas were sufficient or insufficient to impose a Significant Weakness. The problem was, instead, an invalid basis for calculating the delta: [18] as indicated in the first sentence of the paragraph, that the library rates were "an improper reference for incumbent capture.” Wolfgang Rate Memo, AR Tab 44, at 005298. Furthermore, the phrase "decided to” (rather than, as the Plaintiff argues, "was constrained to,” SP Mot. at 19) indicates that NASA knew it had the option to either follow or not follow the GAO's recommendation, and it chose its path willingly. [11] Indeed, as noted above, SP Systems' original complaint included an allegation that NASA had changed the criteria without amending the solicitation and notifying the offerors. Compl. ¶¶ 63-64. [12] The GAO concluded that "there is nothing in the solicitation in the section addressing the technical evaluation expressly indicating that the library rates would be used as a benchmark for evaluation purposes,” in finding that ARTS was not challenging the solicitation itself but instead "the reasonableness of NASA’s evaluation and conclusions, which stem from its reliance on the library rates as an evaluation tool." GAO Decision at 5 n. 6, AR Tab 44 at 005267. [13] SP Systems also argues that for NASA to simply “disregard” the library rates, that is, effectively write them out of the RFP, it would have had to issue an amended solicitation. SP Mot. at 32. The facts clearly indicate that the library rates were not "disregarded" — in fact, GAO approved their use as the “plug number” in the cost realism evaluation. GAO Decision at 10, AR Tab 44 at 005272. The GAO also noted that “there is nothing in the solicitation in the section addressing the technical evaluation expressly indicating that the library rates would be used as a benchmark for evaluation purposes.” Id. at 5, AR Tab 44 at 005267 n. 6. Thus, no amendment to the RFP would be required to allow NASA to refrain from using the library rates as a benchmark in the technical evaluation. [14] The Court has chosen not to rely upon the allegation that NASA could not amend the RFP because it was unable to disclose the incumbent's actual weighted labor rates due to the claim by the incumbent that they were proprietary. [15] The Court is not persuaded by the contentions that the GAO recommendation forbade the agency to consider the offerors' labor rates or related portions of the RFP in violation of the terms of the RFP. SP Mot. at 20. With regard to NASA's initial evaluation, GAO invalidated only one aspect — the imposition on ARTS of a significant weakness — because it was based upon improper use of the library rates, and therefore was insufficient to support the imposition of the significant weakness assigned to ARTS. GAO required NASA to remove the significant weakness and rescore the proposals, without relying on the library rates in an unwarranted manner. The analysis of each offeror's entire proposal remained from the initial evaluation, and only this one flaw was to be corrected. Although plaintiff is unhappy with the manner in which the probability of incumbent capture was considered in the second source selection, to the extent that the analysis changed it was simply because NASA had not demonstrated that the agency's record contained a sufficient basis for the original conclusion. The plaintiff relies extensively upon RFP Section L.15, which instructed offerors who proposed to retain incumbent personnel to "clearly explain variances from their proposed unburdened Offeror Direct Labor Rates ... and those of the current incumbent’s unburdened rates.” See, e.g., SP Mot. at 27. Contrary to the plaintiff's assertion. Section L of the RFP contains instructions to offerors, not NASA. RFP Section L ("Instructions, Conditions and Notices to Of-ferors”), ¶ L. 15, AR Tab 7 at 000493-94. Section L.15 did not direct NASA to do anything, and the GAO decision did not change the terms of the solicitation. Furthermore, the information attained through the GAO protest made the reason for the difference between ARTS’ proposed unburdened rates and the incumbent's unburdened rates abundantly clear: ARTS’ average rates were weighted and the incumbent's were not. At bottom, SP Systems’ arguments rely upon the propriety of using the library rates (which were straight rather than weighted averages), to reflect the weighted average labor rates of the incumbent, a proposition the Court rejects. [16] The amending of a solicitation is done by the contracting officer. FAR 15.206. The contracting officer for the PAAC III contract was Donna Broderick. See, e.g. AR Tab 9 at 000777. Ms. Broderick was a member of the SEB, and she signed the Addendum to Final Report immediately after the Chairman, Mr. Wolfgang. AR Tab 44 at 005285.