ERIC G. BRUGGINK, Senior Judge.
This is a post-award bid protest of the Department of Energy's ("DOE") award of a Blanket Purchase Agreement for Information Technology services to Accenture Federal Services LLC ("Accenture"). Plaintiff, ActioNet, Inc., currently holds an IDIQ task order contract with DOE for IT services. Currently pending is plaintiff's motion for a preliminary injunction. Oral argument was held on April 11, 2019. As announced at the conclusion of that argument, we deny the motion, primarily due to a lack of likelihood of success on the merits.
DOE issued Request for Quotations No. DE-SOL-0008790 ("RFQ" or "solicitation") on November 15, 2017, asking holders of Federal Supply Schedule 70 contracts to submit bids for a single award blanket purchase agreement ("BPA") against which orders could be placed for information technology ("IT"), telecommunications, and cybersecurity services for a period of up to five years (one base year and four option years). The total estimated value for all five years was $2,000,000,000. The BPA as a whole is known as "CBOSS," which stands for CIO Business Operations Support Service. The RFQ included performance work statements ("PWS") for two planned initial orders: one for IT Modernization Strategy Support, known as the "IM-60;" and the other was for Cybersecurity Strategy and Program Management Support, known as "IM-30." In large part, bidders were to tailor their offers to these two task orders.
The agency made the award on a best value tradeoff basis, which was promised by the RFQ, based on the four following factors: 1) Past Performance, 2) Management Approach, 3) Technical Approach, and 4) Price. Past Performance was the most important evaluation factor and was rated for relevance and quality of performance with an overall adjectival rating of either Outstanding, Good, Satisfactory, Marginal, or Unsatisfactory. DOE reserved for itself the right to eliminate offers rated marginal or lower on Past Performance, and it did so for one offeror.
Management Approach, the second most important factor, and Technical Approach, the third in importance, were also rated adjectively from Outstanding to Unsatisfactory. Proposals were assigned strengths and weaknesses for these two factors according to the definitions provided in the solicitation. See Pl.'s Ex. 1 at 699.
Although the RFQ estimated a total value of $2 billion for the BPA across five years, it did not require the agency to evaluate the total price of each offer. Instead, the solicitation stated that the agency would compare labor rates across offers, total prices for the two anticipated initial orders (IM-60 and IM-30), and the load percentage to be charged on other direct costs and materials. The results of the BPA labor rates evaluation was the most important element of the price evaluation. The solicitation also stated that the agency would perform a price realism and reasonableness analysis of all offerors' prices and rates. The agency reserved the right to conduct discussions with offerors but warned that it might make an award without having done so.
The Technical Evaluation Committee ("TEC") performed the non-price evaluation of offers and presented its findings to the Source Selection Official ("SSO"), who frequently met with the TEC and the Contracting Officer ("CO") throughout the evaluation process. The CO conducted the price analysis. The source selection decision was made by the SSO.
DOE received six offers in response to the RFQ. The TEC first evaluated the Past Performance factor by itself, completing that evaluation on September 24, 2018. One of the six offerors was eliminated by the CO because it was rated only marginal for Past Performance. The TEC then completed its evaluation of the other two non-price factors on November 19, 2018. The results for all three factors were as follows (ranked in order):
PX 5 at 9.
The CO completed her price analysis the next day on November 20, 2018. All five remaining offerors' prices were found to be both fair and reasonable overall, but plaintiff's evaluated prices for the two initial tasks orders were found to be unrealistically low. The highest offeror was $25,990,987.33 for the two orders; Accenture was at $21,358,203.81; and ActioNet bid the work for only [ ]. The next lowest offeror, by comparison, was $19,766,505.24.
The CO evaluated the BPA labor rates by comparing the offerors' proposed rates against one another for each category. The mean for each labor category was calculated and then a standard deviation analysis was performed against that mean for each category. Rates within two standard deviations on the low side of the mean and one standard deviation above were considered to be within range. The CO then evaluated what percentage of all of the labor categories each offeror's prices were. The higher the number, the better. Plaintiff was evaluated to have 90.8% of its rates within range while intervenor was found to have 72.3%.
As to the load rates for materials and other direct costs, ActioNet had a range of rates starting at [ ] for materials and subcontractor costs up to [ ] for travel and other direct costs. Accenture offered [ ] for its load across categories and a [ ] load for its teaming partner [ ].
The SSO reviewed the analyses performed by the TEC and CO and ranked the offers for non-price factors. Intervenor was the highest ranked, and plaintiff was the lowest. Accenture was chosen as the most advantageous (best value) quote to the government. The SSO did, however, perform a trade-off analysis between Accenture and each of the four other offerors in her source selection decision. As between the plaintiff and the intervenor, the SSO found that ActioNet had a price advantage in the categories of BPA and load rates, but, due to the unrealistically low prices offered for the two initial offers, the SSO determined that the price factor was not in favor of plaintiff. The SSO's decision document also states that, even if plaintiff's order prices were found to be realistic, she would still award to Accenture due to the technical (including Past Performance) advantage offered by its quote. The trade-offs between intervenor and the other offerors each also ended with the conclusion that the rating for Past Performance was the most important discriminator in choosing Accenture for award.
DOE notifed the offerors of its decision on November 30, 2018, and awarded the first two task orders on December 6, 2018. The period of performance began the next day. Plaintiff filed a protest at the Government Accountability Office ("GAO") on December 10, 2018, however, which resulted in an automatic stay of performance. GAO denied the protest on March 5, 2019, finding both that the agency had a reasonable basis for its evaluation and award decision and that ActioNet lacked standing because it did not challenge the ratings of the offerors other than Accenture rated higher than it. ActioNet, Inc., B-417173, 2019 CPD ¶ 100 (Comp. Gen. Mar. 5, 2019).
Plaintiff filed its complaint in this court on March 13, 2019, along with a request for a Temporary Restraining Order ("TRO") and a preliminary injunction. A status conference was held on March 14, 2019. Accenture intervened on that day as well. Although the agency represented that it had no plans to issue any new orders prior to April 12, 2019 (the date when plaintiff's incumbent contract was set to expire), plaintiff maintained its request for a preliminary injunction. We entered a briefing schedule on March 18, 2019, concluding with oral argument on April 11, 2019.
On March 19, 2019, the government filed a notice of intent to issue six additional task orders on or by March 21, 2019. This prompted plaintiff to file a motion on March 22, 2019, requesting that the court issue interim relief to preserve the status quo until the preliminary injunction hearing on April 11, 2019. After a response from defendant on March 28, 2019, we denied the request for interim relief by order on March 29, 2019, finding that plaintiff had not made a showing of irreparable harm if an injunction were not issued prior to the preliminary injunction hearing. ActioNet, Inc. v. United States, No. 19-388C, 2019 WL 1423096, at *3-4 (Fed. Cl. Mar. 29, 2019). Oral argument on plaintiff's motion for a preliminary injunction was held on April 11, 2019, at the conclusion of which we announced that the court would deny plaintiff's motion and requested the parties to propose a briefing schedule for resolution on the merits of a permanent injunction. We confirmed that result by short order on April 12, 2019. This opinion contains our reasoning for having done so.
We have jurisdiction over challenges to agency actions taken or not taken in connection with a federal procurement. See 28 U.S.C. § 1491(b) (2012). Our review under this section is deferential to agency decision making in accordance with the standard set forth in the Administrative Procedures Act, 5 U.S.C. § 706, which is to say that we review agency action in a procurement for illegality and a lack of rationality. Impressa Construzioni Geom. Domenico Garufi v. United States, 238 F.3d 1324, 1332-33 (Fed. Cir. 2001). So long as the agency's decision was not irrational or otherwise illegal, we will leave it undisturbed.
The issue of prejudice is also relevant to our review under this statute. First, a protestor must show that it is an economically interested party, which, in the post-award context, means that it had a substantial chance of award but for the alleged error(s) of the agency. Data Gen. Corp. v. Johnson, 78 F.3d 1556, 1562 (Fed. Cir. 1996). The protestor need not show that it was next in line for award, however, to establish standing, only that there would have been a reasonable likelihood of it being awarded the contract. Id. at 1563 If it cannot make such a showing, it was not prejudiced by the error(s), and jurisdiction will not attach. Even if the disappointed bidder establishes standing, it must also show that the error alleged was prejudicial, which means that, not only an error took place, but that it affected the outcome of the agency's decision. Labatt Food Serv., Inc. v. United States, 577 F.3d 1375, 1380-81 (Fed. Cir. 2009) (holding that the agency's acceptance of bid revisions by email did not prejudice protestor because it had no bearing on the agency's review of final proposals).
Here, we find that plaintiff has established standing. Plaintiff alleges that it was rated too low for each of the non-price factors, alleges that its prices should have been viewed more favorably, and that intervenor was rated more highly for several of the factors. We also know that the agency performed a trade-off between Accenture and ActioNet. We thus conclude that, had plaintiff been evaluated more favorably and/or intervenor been rated less favorably, a different outcome is reasonably likely. Put another way, as both this court and the Federal Circuit have articulated the test, when the protestor shows that it is "`within the active zone of consideration'" by the agency, it has established standing. E.g., Advanced Mgmt. Strategies Group, Inc. v. United States, 139 Fed. Cl. 404, 411 (2018) (quoting Alfa Laval Separation, Inc. v. United States, 175 F.3d 1365, 1367 (Fed. Cir. 1999)). We will consider the separate prejudice question as it pertains to the merits below.
Standing established, we move to the merits of plaintiff's request for a preliminary injunction. We consider four factors when entertaining a request for preliminary injunctive relief: (1) plaintiff's likelihood of success on the merits; (2) whether plaintiff will suffer irreparable harm without the injunction; (3) the balance of the harms between the parties; and (4) the public's interest in an injunction. Bona Fide Conglomerate, Inc. v. United States, 96 Fed. Cl. 233, 239 (2010). This court is generally reticent to enjoin agency procurement action without a showing that the protestor is likely to succeed on the underlying merits of its protest. See Intelligent Waves, LLC v. United States, 135 Fed. Cl. 299, 314 (2017). We begin with the first factor, as it is dispositive in this case.
Plaintiff broadly attacks each and every element of DOE's evaluation of its own procurement, arguing at each step that it should have been more highly rated and that it was treated differently than other offerors. It also argues that intervenor's rating for the Past Performance factor was too high. Although plaintiff's attack is broad in scope, its briefing and oral argument was focused more specifically on certain arguments. This opinion is generally concerned with those more focused inquiries, but we have considered each of plaintiff's arguments and find that none of them support a showing of likelihood of success on the merits.
We begin with the most important evaluation factor, Past Performance. The thrust of plaintiff's argument is that its own past performance examples were irrationally viewed as less relevant than they should have been and that DOE irrationally found ActioNet's performance for several of these contract efforts to have been less stellar than the agency had previously rated those efforts in the past. As to intervenor's rating, the primary argument offered by plaintiff is that Accenture's references for IT migration services were considered more relevant than they should have been. Plaintiff also sees disparate treatment in this regard by DOE as between it and the intervenor. Plaintiff further argues that the agency impermissibly relaxed the requirements to receive an Outstanding rating for Accenture despite not finding each of intervenor's references to have exceeded the scope of the solicitation's requirements, as plaintiff argues is required by the RFQ.
Each offeror was required to provide six narrative examples of past performance. These six were to show similar services performed to the CBOSS PWS and to show similar levels of size, scope, and complexity. Of the required examples, DOE asked, among other things, that two of them be examples of "IT Modernization/Cloud Migration." PX 1 at 700. DOE promised to review the past performance questionnaires for each of the past performance efforts and stated that it might review any available past performance records, such as CPARs. The agency also reserved the right to contact references for each narrative example as well as review other pertinent past performance information ("any other source"). Id. at 702.
Plaintiff provided five of its six references from the incumbent IDIQ DOE contract. The sixth was for its [ ]. Past performance questionnaires were available for all six references, and DOE reviewed the CPARs ratings going back as far as 2012 for all offerors' provided references. DOE also contacted, or attempted to contact, technical points-of-contact ("POC") for each reference.
The result for plaintiff was that its CPARs ratings were very good (none were obtainable for the [ ] contract); the questionnaires were largely good and ranged from [ ] to [ ]; the interviews of POCs, however, were mixed and contained some information that DOE viewed negatively, or as the agency put it "a significant amount of highly unfavorable feedback." PX 3 at 51. The agency thus found some risk attached to plaintiff's Past Performance information but found that the information was generally favorable. It also found that some of plaintiff's references were not as relevant as the agency sought. It thus awarded a Satisfactory rating for plaintiff.
Plaintiff attacks the agency's conclusion in general by comparing its CPARs ratings, which were favorable across the board, with DOE's finding that there was significant unfavorable information present for ActioNet, which DOE found to present some risk of poor performance going forward. In plaintiff's view, its high CPARs ratings should have afforded it a higher overall rating. We disagree.
The agency reviewed what it said that it would and came to a documented conclusion that is rationally related to the information that it received and reviewed.
Plaintiff also argued more particularly that its two narrative examples for IT modernization and migration should have been viewed by the agency to be more relevant than they were. Much of the briefing and oral argument was spent on this allegation, the crux of which is that the agency used a metric of 13,000 users migrated from legacy systems to cloud-based systems to determine relevancy for these examples. The agency found that plaintiff did not show that it met this requirement for either example, but it found that intervenor's examples did. Plaintiff argues that this was an unstated evaluation criterion because the agency morphed the 13,000 "user" requirement to mean "seat type users with distinct credentials," which it views as the only way that the agency could have found it not to have met the requirement. PX 3 at 41. Plaintiff also argues that it was treated unequally in this regard because it views intervenor's two examples as insufficient to meet the test.
The allegation that the agency applied a different meaning of "users" than that stated in the RFQ must be rejected because plaintiff has not shown how it was prejudiced by the error. This criteria was applied equally across all offerors, and to the extent that it was a mystery prior to bid submission, that mystery prejudiced all offerors equally, which is to say that it prejudiced none of them as regards the outcome.
We also do not view this allegation as otherwise having merit. The fact that the agency cited a more specific definition of a general term from the solicitation is not per se irrational nor application of an unstated criteria. Defendant and Intervenor provided several citations to the record where the agency clearly intimated what it was looking for regarding migrating "end users" and that the agency was looking for a "managed seat service environment." PX 1 at 253 ("end users"); id. at 383-84 ("managed seat service environment" and "seat service end users."). We think the offerors were on notice of the agency's bent in this regard.
ActioNet also believes that it should have been found to have met the 13,000 user requirement and that Accenture should not have, which is in its view unequal treatment. Plaintiff's first cited example of cloud-base migration involved the moving of accounts for Office 365, a cloud-based suite of applications. Plaintiff provided that it moved [ ] users and [ ] generic mailboxes to these cloud-based programs. For its second cited example, plaintiff offered its migration of DOE's website from a self-hosted solution to a cloud-based one, which now has [ ] of monthly page views but only [ ] of DOE users on the inside managing the content of the website. The agency found neither of these examples to have met the requirement. For the first example, [ ] is not 13,000, and the agency did not consider the additional [ ] generic mailboxes to be users. For the second example, the DOE website, the agency again found that website viewers were not the same thing as the industry standard of a "user," which is a "credentialed user." PX 3 at 43. We are in no position to disagree with these conclusions.
Plaintiff also alleges that Accenture too should have been found not to be compliant with the 13,000 user requirement. The first example provided by intervenor to the agency for this requirement was its work for the[ ]on the [] website, which has millions of users and hundreds of thousands of concurrent users daily. Plaintiff argues that this example is inapposite because Accenture did not migrate the [ ] website to the cloud; it alleges that was done by a prior contractor. Thus, even assuming the proper metric was used, plaintiff should not have been credited with this work.
Defendant rebutted this point at oral argument by showing that intervenor did not claim credit for the initial move of the website to the cloud nor was the agency confused about this fact. In fact, Accenture's proposal represented that it migrated three critical applications used on the website to the cloud; the three applications have [ ] users. See id. at 38 (stating that Accenture had "migrated three core systems" to the cloud). We thus find no basis to question the agency's views of these contracts due to this distinction.
The second cited example of cloud migration by Accenture was its migration of the [ ] to a cloud-based application. Both the proposal and DOE's evaluation of it cite over [ ] credentialed users of this system. We thus find a rational connection between the requirement of the migration of 13,000 end users to this cited effort and find no reason to call into question the agency's evaluation in this regard.
The last avenue of plaintiff's challenge to DOE's past performance evaluation concerns whether intervenor was eligible to receive an Outstanding rating per the terms of the solicitation. The RFQ states that a proposal will be considered to be Outstanding for Past Performance if the quoter shows that its
PX 1 at 697. At oral argument, plaintiff argued that the fact that only two of the references of intervenor were shown to involve greater size, scope, magnitude of effort, and complexity shows that it should have been considered Good at best. We disagree.
The term "involved" can be read several ways. It could mean that each example of Past Performance exceeded CBOSS requirements, or it could be read to mean that, of all of the examples provided by an offeror, some must involve a scope greater than CBOSS. The agency applied the second definition. That is not unreasonable nor was it applied unequally to different offerors. In sum, we have considered all of plaintiff's offered arguments regarding Past Performance, including those not specifically referenced above, and find them unavailing. Plaintiff has not shown a likelihood of success on the merits for Past Performance.
For the Management Approach factor, plaintiff's argument centered on how the agency evaluated the relevant experience of ActioNet's offered key personnel. Admittedly, the TEC's report mistakenly stated a higher number of years for the relevant categories than the solicitation required. There is no dispute that this mistake was made, but we agree with GAO that the particular number of years was not the basis for plaintiff's weakness under this factor. ActioNet, Inc., 2019 CPD ¶ 100 at 9-10. It was instead the fact that ActioNet's proposal for its key personnel lacked sufficient detail to determine whether its personnel had the requisite experience (even if the number listed for necessary years of experience was too high). See PX 4 at 32-33 (Management Approach evaluation of supporting documentation for key personnel). We have reviewed plaintiff's Attachment F and conclude that the agency was not irrational in finding it to be lacking sufficient detail for DOE to satisfy itself that plaintiff's personnel had the required years of experience.
The Technical Approach factor was not discussed during oral argument, but we briefly consider it here because it is plain from the record that the agency had a reasonable basis for its evaluation of this factor. DOE considered the proposed approach for the two initial task orders under this factor. Plaintiff was assigned two weakness for its IM-60 approach and one weakness for the IM-30 order. Plaintiff challenges all of them.
The TEC found plaintiff's approach for this initial task order to be problematic because the timeframe proposed was, in the agency's view, too short to achieve the level of satisfactory performance it desired. One of the overarching goals of the contract was to obtain a modernization and updating of DOE's IT infrastructure and cybersecurity. The IM-60 task order was anticipated to be an agency wide survey of DOE's IT infrastructure and environment and to identify future needs across the agency. The anticipated period of performance for this task order was six months with an additional six-month option.
Plaintiff proposed to complete its [ ] agency-wide assessment in[ ]. PX 7 at 5 (timeline); id. at 6-9 (narrative description). DOE assigned a weakness for this timeframe, finding it to be evidence that plaintiff did not understand the requirement. Plaintiff views that conclusion as factually wrong, arguing that the agency misread its proposal. Plaintiff urges that the [ ].
We agree with GAO that DOE had a rational basis for its conclusion in this regard. See ActioNet, Inc., 2019 CPD ¶ 100 at 12. Even considering the [ ] nature of the report to be produced in [ ], it was reasonable for the agency to consider that too short a period of performance for even a draft, not to mention that plaintiff's [ ] was much accelerated compared to the agency's anticipated 6-12 month period. A [ ] report would necessarily be comprehensive in scope, meaning that the majority of fact gathering, and evaluative work would be complete in that period. That is a reasonable basis on which the agency could conclude as it did.
The second weakness assessed to plaintiff's proposed effort for the IM-60 order was for ActioNet's lack of details for its risk management and mitigation plan. Plaintiff takes aim at this weakness by citing its [ ]. The TEC concluded that this high-level approach was insufficient as it was unrelated to the execution of the IM-60 effort. We find no irrationality with this conclusion.
The second initial task order was aimed at providing immediate and critical support to DOE's cybersecurity efforts across the agency. The TEC found that the [ ] full-time equivalents ("FTEs") proposed to complete this task order were insufficient for the level of service sought and even the level of service proposed by ActioNet. The agency found this [ ] level of effort to be a risk to the successful execution of the task order.
Plaintiff argues, first, that this was a mechanical application of an unstated labor floor, and, second, that plaintiff was not given credit for the []. Defendant and intervenor respond that the agency did not mechanically apply a labor floor but instead found that the number of hours proposed by plaintiff to be problematic. We agree.
There is no support in the record for the idea that the agency had an unstated minimum number of hours that it was looking for offerors to propose for this task order. The TEC reviewed the contract requirements and the proposed effort of ActioNet and concluded that the number of FTEs offered by plaintiff to be inconsistent with both. Plaintiff has not shown how it was otherwise irrational for the agency to conclude as it did. The idea that the agency should have read into plaintiff's proposal [ ]. We find that DOE reasonably reached the conclusion that it did in this regard.
Plaintiff raises two primary issues as it relates to the CO's price evaluation and the SSO's use of that evaluation in her award decision. The first is that the agency failed to give plaintiff proper credit for the price advantage that it offered for BPA labor rates and load rates. The second issue is with the agency's finding that plaintiff's price for the IM-60 and IM-30 task order was too low (unrealistic).
As to the first issue, we find no error to have occurred. The agency conducted a standard deviation analysis of all proposed rates and then compared those rates across all offers to figure out what percentage of overall rates fell within the range that the agency set. Plaintiff does not take issue with this approach in general. It argues instead that its higher percentage of rates within the range, when contrasted to Accenture's lower level, should have resulted in a more favorable rating for it. The same basic argument is made for the non-labor load rates, i.e., that plaintiff's lower rates for this category vis-à-vis intervenor should have made some difference to the agency's ultimate conclusion.
The award decision makes clear that the SSO was aware of ActioNet's advantages as to these two considerations. It plainly stated so. The SSO did not, however, find that those two advantages would overcome the technical superiority of intervenor's proposal, especially given the unrealistic prices for the two task orders proposed by plaintiff. We find no irrationality in these conclusions.
Even were we to agree with plaintiff on the merits, plaintiff would not otherwise be entitled to preliminary injunctive relief. Although there is a case to be made for plaintiff suffering some harm at this juncture without an injunction, the balance of harms does not support an injunction. The parties submitted a series of competing declarations regarding the harm to be suffered absent an injunction on plaintiff's part and because of an injunction on intervenor's and the agency's part. Without detailing each, it is sufficient to say that the agency's desire to begin modernizing its IT infrastructure by moving services to the cloud and its desire to obtain additional cybersecurity services as quickly as possible rises to, at a minimum, equal the harms to be suffered by plaintiff's loss of revenue and potential loss of personnel, both of which seem less than irreparable given that the awardee of this contract is looking at approximately $2 billion in work over five years.
On the whole, we find that plaintiff's disagreements with the agency are just that, a difference of opinion on the relative merits of the proposals the agency reviewed. Such disagreements are insufficient to establish a likelihood of success on the merits. That fact in combination with the relative balance of the harms alleged make a preliminary inunction inappropriate. Thus, we denied plaintiff's motion for a preliminary injunction.
