THIS OPINION WAS INITIALLY ISSUED UNDER PROTECTIVE ORDER AND IS BEING RELEASED TO THE PUBLIC IN REDACTED FORM ON MAY 18, 1993 ___________________________ DENIED: May 10, 1993 ___________________________ GSBCA 12301-P, 12333-P COMPUADD CORPORATION, and APPLE COMPUTER, INC., Protesters/Intervenors, and INTERNATIONAL DATA PRODUCTS CORPORATION, Intervenor, v. DEPARTMENT OF THE AIR FORCE, Respondent, and ZENITH DATA SYSTEMS CORPORATION, and GOVERNMENT TECHNOLOGY SERVICES, INC., Intervenors. Marie N. Doland and James J. Tierney of Hazel & Thomas, Falls Church, VA, counsel for Protester/Intervenor CompuAdd Corporation. John A. McCullough and Ron R. Hutchinson of Doyle & Bachman, Washington, DC; and Susan Norman of Apple Computer, Inc., Reston, VA, counsel for Protester/Intervenor Apple Computer, Inc. D. Oscar Fuster, Vice President of International Data Products Corporation, Gaithersburg, MD, appearing for Intervenor. Clarence D. Long, III, Joseph M. Goldstein, and Lieutenant Colonel Dwight D. Creasy, Office of the General Counsel, Department of the Air Force, Washington, DC; and Captain Eric J. Marcotte, Gunter Air Force Base, AL, counsel for Respondent. Laura K. Kennedy, Kevin P. Connelly, Grace Bateman, Trisa J. Thompson, Daniel Marino, and William A. Breskin of Seyfarth, Shaw, Fairweather & Geraldson, Washington, DC, counsel for Intervenor Zenith Data Systems Corporation. Richard J. Conway, William M. Rosen, and Joel Fried of Dickstein, Shapiro & Morin, Vienna, VA, counsel for Intervenor Government Technology Services, Inc. Before Board Judges HENDLEY, HYATT, and VERGILIO. HYATT, Board Judge. These protests challenge awards by the Department of the Air Force, to Zenith Data Systems Corporation and Government Technology Services, Inc. (GTSI), of indefinite delivery indefinite quantity (IDIQ) contracts to supply the Department of Defense (DOD) and, on a lesser scale, civilian agencies, with microprocessor-based microcomputer systems and associated software, hardware, user-installable components, and warranties. This acquisition is intended to fulfill general purpose desktop computer needs throughout DOD and is commonly referred to as the Desktop IV (or DTIV) procurement. This is the third round of protests concerning the Desktop IV procurement. For the reasons stated, we deny the protests. Background The Solicitation[foot #] 1 The Desktop IV request for proposals (RFP) was issued by the Air Force on July 3, 1991. The solicitation called for the supply of commercial automatic data processing hardware, software and support services and, in particular, for the supply of up to ----------- FOOTNOTE BEGINS --------- [foot #] 1 A more detailed discussion of the purpose of this procurement and the terms of the RFP is set forth in the Board's decision in the earlier protests of the single award to Zenith Data Systems Corporation. CompuAdd Corp. v. Department of _______________________________ the Air Force, GSBCA 12021-P, et al., 1993 BPD 48 (Dec. 23, _____________ 1992). ----------- FOOTNOTE ENDS ----------- 300,000 microcomputer systems. Protest File, Exhibit 16.[foot #] 2 The Air Force planned, and so stated in the RFP, to effect a streamlined acquisition process with the objective of awarding a contract within six months. This was to be accomplished by using a functional specification and by making award without discussions. Protest File, Exhibit 1. The applications to be supported by Desktop IV microcomputer systems include office automation functions such as word processing, database management, spreadsheet creation and manipulation, and business graphics. Protest File, Exhibit 16. The microcomputer systems to be purchased under the Desktop IV contract were set forth in four contract line items (CLINs): CLIN 0001-Basic System; CLIN 0002-Intermediate System; CLIN 0003- Advanced System; and CLIN 0004-Open Systems Common Applications Environment. Other CLINs and sub-contract line items (SLINs) provided for the purchase of such items as separate software packages (CLIN 0005) and printers (CLINs 0006-0007). Protest File, Exhibit 16. The RFP included a provision allowing the Air Force to award one or two contracts: The Government may elect to award the contract to one or two offerors based on an integrated assessment of each offeror's ability to satisfy the requirements of the solicitation. The Government will not award individual CLINs/SLINs on the contract to any one offeror, but may award all CLINs/SLINs to two offerors. Protest File, Exhibit 16 at 3. Section M of the solicitation further stated that the Source Selection Authority (SSA) would determine whether to make an award to an individual proposal or combination of proposals depending on which offered the "best overall value to the Government." Protest File, Exhibit 16 at 196. In sum, the Government could award either one or two entire contracts, which was understood by all the offerors. The Air Force advised offerors that it would not make two awards to the same firm, however. CompuAdd Corp. v. Department of the Air Force, GSBCA 12021-P, et al., 1993 BPD 48, at 5 (Dec. 23, 1992). The ability to make more than one award was deemed desirable in part because of the need for a hedge against the possibility of an offeror being unable or unwilling to deliver in the quantities ordered by the Air Force. Id. Section M provided that offerors would be evaluated in three specific areas: management; technical; and cost/price. ----------- FOOTNOTE BEGINS --------- [foot #] 2 Upon the filing of the current round of protests, it was agreed that the record from the immediately preceding protests would be incorporated in its entirety as part of the record in the present protests. ----------- FOOTNOTE ENDS ----------- Management and technical were each more important than cost/price. At the same time the RFP stated: Although Cost/Price is listed last, it will be a significant criteria [sic] for award as part of an integrated assessment with the management and technical areas. The Government reserves the right to award a contract(s) at other than the lowest proposed price after consideration of all factors. Protest File, Exhibit 16 at 196. The solicitation contained an equipment modification/ substitution clause that authorized adjustments to equipment offered by the contractor or contractors after award. Protest File, Exhibit 16 at 63. The RFP also made applicable the requirements of the Trade Agreements Act of 1979 (TAA). The Air Force explained to offerors that this provision applied to all individual components of the microcomputer systems, such as monitors, keyboards, and pointing devices, as well as to the central processing units. As such, all of these components were required to be manufactured in a TAA approved country: The basic system on all CLINs includes a system unit, monitor, keyboard, and pointing device. The Government does not interpret the bundling of these items as creating a "new and distinct article of commerce" under the Trade Agreements Act (TAA). We believe that each item listed above retains its independent character and, therefore, must independently comply with the TAA. Note that, unlike the BAA [Buy American Act], the TAA does not embrace a cost based test to determine compliancy. The determining factor is not whether more than 50 percent of the cost of the system is composed of TAA compliant products. The test - if products from several countries are used - is whether component parts have been substantially transformed within an appropriate country. We do not view merely packaging products into a system as a substantial transformation. Protest File, Exhibit 2175. Offerors were required to certify that the products offered under their proposals were in compliance with the requirements of the TAA. Protest File, Exhibit 16. The Prior Protests After evaluating initial proposals submitted in response to the RFP, the Air Force concluded that all but two proposals contained deficiencies that disqualified them from an award without discussions. This left only two offerors, Sysorex Information Systems, Inc. and CompuAdd Corporation, with offers that the Government deemed eligible for award. Awards were made to both offerors on November 9, 1991, at an evaluated price of $1.2 billion over the three year life of the contract. Protest File, Exhibit 11. The proposals selected at that time offered a range of technical solutions and a wide disparity in prices. These initial awards were protested by eight of the remaining disappointed offerors. These protests were settled and dismissed after the Air Force agreed to terminate the initial awards and open discussions. Apple Computer, Inc. v. Department of the Air Force, GSBCA 11161-P, et al., 1992 BPD 9 (Jan. 8, 1992). The Air Force amended the solicitation, conducted discussions, and proceeded again to award, this time selecting a single proposal that had been submitted by Zenith Data Systems Corporation. This award was announced on September 9, 1992. The evaluated price for this award, over a three year period, was $740 million. Once again, a number of disappointed offerors protested. The Board issued a decision granting the protests of CompuAdd Corporation, Electronic Data Systems Corporation (EDS), GTSI, and Apple Computer, Inc. on December 23, 1993. CompuAdd Corp. That decision essentially ruled that the Air Force had not adequately considered combinations of awards in contrast to a single award in determining the best overall value for the Government. In addition, the Board found that the monitors proposed in Zenith's successful offer did not meet the requirements of the Trade Agreements Act of 1979 for substantial transformation in a designated country. After requesting offerors to extend their remaining offers, the Air Force announced two awards, one to Zenith, accepting an alternate proposal, and one to GTSI, on February 2, 1993. Protest File, Exhibits 281, 283, 286. In response to this announcement, EDS filed a motion to enforce the Board's decision dated December 23, 1992. Shortly thereafter, on February 12, 1993, EDS and CompuAdd Corporation each filed protests of the latest awards. These protests were docketed as GSBCA 12300-P and 12301-P, respectively. Apple Computer, Inc. timely intervened in the protests filed by EDS and CompuAdd and subsequently filed its own protest, GSBCA 12333-P, on March 15, 1993, after conducting discovery. On April 19, 1993, on the forty-fifth working day after filing its initial protest of the awards to Zenith and GTSI, EDS filed motions seeking to withdraw its protest and its motion to enforce the decision in CompuAdd Corp.[foot #] 3 On ----------- FOOTNOTE BEGINS --------- [foot #] 3 Recognizing that the motions filed by EDS would cause some delay in the issuance of the Board's decision on the merits of the remaining protests, the Air Force petitioned the Board to partially lift the interim suspension imposed after the (continued...) ----------- FOOTNOTE ENDS ----------- April 27, 1993, the Board dismissed the EDS protest. In the opinion accompanying the dismissal of that protest, the Board indicated that, because of the timely interventions of Apple and CompuAdd in the EDS protest, none of the issues raised would be removed from consideration in the Board's opinion on the merits of the remaining protests. Electronic Data Systems Corp. v. Department of the Air Force, GSBCA 12300-P (Apr. 28, 1993). The issues in this third round of protests have been narrowed considerably. Protesters have challenged these awards on four specific bases: (1) that the Zenith monitor continued to be unacceptable and was not eligible for award; (2) that the Air Force engaged in improper discussions with Zenith; (3) that the Air Force conducted a flawed best value analysis in selecting the Zenith and GTSI proposals; and (4) that evidence adduced in the second and third rounds of protests demonstrated that the Air Force's requirements had changed, requiring amendment of the RFP and another round of BAFOs. Findings of Fact 1. Following issuance of the Board's decision granting the protests in CompuAdd Corp., the SSA convened a meeting of advisors to consider appropriate courses of action available to the Air Force in proceeding with the Desktop IV procurement. Transcript at 146-47, 270-72, 833-34. At this initial juncture, the focus was on whether to cancel the procurement or proceed with an award. Transcript at 833-34. 2. After concluding that it might still be feasible to continue with the procurement, the discussions turned to whether the RFP continued to reflect accurately the Air Force's needs. Transcript at 692. The advisory group determined that it did. Transcript at 692. At the conclusion of this meeting, the SSA asked the chairman of the Source Selection Advisory Council (SSAC) to reconvene the SSAC and determine if an evaluation process could be developed that would meet the guidelines set forth in CompuAdd Corp. Transcript at 813-14. 3. The Air Force next convened a meeting of the SSAC at Gunter Air Force Base. Protest File, Exhibit 278; Transcript at 144-59. The purpose of the meeting was to develop, if possible, a methodology for proceeding with the procurement that would accord with the Board's decision in CompuAdd Corp. Transcript at 144-45. The Air Force considered alternatives ranging from reopening of negotiations, or proceeding with award based upon ----------- FOOTNOTE BEGINS --------- [foot #] 3 (...continued) filing of the initial protests by EDS and CompuAdd. On April 20, after Apple and CompuAdd withdrew their objections, the Board issued an order partially lifting that suspension and permitting the Air Force to begin acceptance testing on the Zenith and GTSI units that will be supplied under their contracts. ----------- FOOTNOTE ENDS ----------- remaining proposals, to canceling the procurement. Protest File, Exhibit 275. After considerable discussion, the SSAC decided to pursue award based upon existing proposals without revising the RFP or conducting further discussions. Transcript at 158. 4. The Air Force interpreted the Board's decision in CompuAdd Corp. to exclude from the competition only the Zenith proposal previously selected for award. The SSAC recognized that monitors offered under the remaining Zenith proposals had the same model number as the monitor offered under the disqualified Zenith proposal. Transcript at 148. The Air Force construed the Board's opinion, however, to relate solely to the manufacturing process described in the prior hearing, and not to the monitor itself. Beyond certifying to compliance with the TAA, the proposals did not describe or otherwise identify the manufacturing process to be used for any product offered. Transcript at 149-51. Based on this, the Air Force decided that only the Zenith proposal selected for the prior award was excluded from further consideration and that it could properly continue to consider the alternative proposals offered by Zenith. Transcript at 152-53. 5. Since proposals had expired as of September 30, 1992, the Air Force concluded that it should provide vendors an opportunity to extend their proposals. The Air Force considered the possibility of requesting that offerors specifically confirm their TAA self-certifications, but rejected this idea as being susceptible to the construction that discussions had been opened. Transcript at 98, 154. The Air Force also determined that extension of the proposals was enough to renew TAA certifications. Transcript at 100. On January 11, 1993, the contracting officer sent letters to each vendor requesting written extensions of proposals. The letter stated that changes to "prices, delivery terms, and other terms and conditions" would not be permitted and that failure to comply with this condition would result in removal of a proposal from further consideration. Protest File, Exhibit 276. 6. A total of thirty-three proposals were extended by the offerors, including six submitted by Zenith, five submitted by GTSI, two each submitted by CompuAdd and Apple and one submitted by EDS. Protest File, Exhibit 277. In addition, Zenith sent a letter to the Air Force acknowledging that it had not been requested to extend its proposal number 3 (the proposal determined to be ineligible for award by the Board in CompuAdd Corp.), but offering to extend this offer anyway. The letter is identical to that extending the other Zenith offers except that it states the following: I assume that you did not request this extension of our offer on proposal 3 based on your analysis of the GSBCA decision. However, as ZDS intends to appeal that portion of the GSBCA decision relating to our proposal 3, I believe that it is appropriate to extend this offer pending resolution of our appeal. Post-Hearing Brief of CompuAdd Corporation, Exhibit 3. The First SSAC Best Value Analysis 7. The Air Force conducted a preliminary analysis of the remaining proposals at the initial meeting at Gunter Air Force Base. Protest File, Exhibit 279; Transcript at 158-66. This was essentially a hypothetical exercise designed to evaluate whether the methodology would work as a practical matter. Transcript at 95. After removing the Zenith proposal deemed ineligible for award under the Board's decision in CompuAdd Corp., the Air Force reviewed remaining proposals using the "single best value" analysis described in the Board's opinion. In conducting this analysis the Air Force reviewed proposals using an electronic spreadsheet that contained the results of technical and management evaluations for each proposal.[foot #] 4 This spreadsheet also reflected present value cost associated with each offer. Protest File, Exhibit 279. 8. Proposals were ranked based upon color codes assigned to management and technical areas.[foot #] 5 After grouping proposals with like color ratings, the Air Force scrutinized proposals in painstaking detail, looking behind the color ratings to the pros and cons of particular features. The SSAC then eliminated from further consideration any lower ranked proposal with a higher or equivalent cost compared to a better rated proposal. After proceeding in this manner, the SSAC eventually identified five proposals that constituted better value for a single award. These five consisted of three Zenith proposals (proposals "M", "AJ", and "Y"), one GTSI proposal ("AQ"), and the EDS proposal ("R"). Protest File, Exhibit 279. 9. The SSAC compared these five proposals and discussed whether superior features of specific proposals justified the cost differences. Protest File, Exhibits 279-80; Transcript at 165, 813. After performing this analysis, the Air Force arrived at the conclusion that proposal AJ, which was priced at approximately the same amount as R, but had a "decided advantage ----------- FOOTNOTE BEGINS --------- [foot #] 4 This was essentially the same electronic spreadsheet described by the Board in CompuAdd Corp., 1993 BPD ______________ 48, at 23-24. [foot #] 5 The color ratings, in descending order, were blue, green, yellow, and red. CompuAdd Corp., 1993 BPD 48, at ______________ 16. Blue signified that a proposal item or area exceeded specified standards and green meant that it met specified standards. Id. __ ----------- FOOTNOTE ENDS ----------- in the technical area," offered the best single value.[foot #] 6 Protest File, Exhibit 280; Transcript at 165. 10. After determining that AJ constituted the best single value, the SSAC turned to developing a procedure to identify the most advantageous pair of proposals. The SSAC determined that the best pair of proposals would most likely be derived from among the "better value" set of single proposals. To supplement this set, which reflected only "Intel chip" technology, the SSAC added into the consideration the "best" of the alternative technology proposals, including an Apple proposal that offered the Motorola chip and another featuring Sun Sparc technology. In addition, the SSAC added a third proposal which offered a markedly different software solution. Protest File, Exhibit 279. 11. After removing proposal pairings from the same offeror,[foot #] 7 these eight proposals yielded twenty-two possible combinations. The SSAC then decided to look at whether it could eliminate some combinations using a methodology similar to the one applied to reduce the field of single better value proposals. In addition, the SSAC decided to consider the alternative technologies only in conjunction with the best of competing technologies. Since the SSA would be responsible for ultimately selecting the best of the proposals offering the Intel chip, the Apple and Sun Sparc proposals were not compared with any Intel-based offers, but only in combination with each other. This reduced the set of pairs to be analyzed by the SSAC to ten. Protest File, Exhibit 279. 12. The SSAC identified three basic benefits it believed could be derived from a dual award: risk reduction; continuing competition; and provision of a range of technology and prices. Protest File, Exhibit 279. Because the SSAC believed that serious management concerns had been eliminated in the better value winnowing process, it deduced that risk reduction would be ----------- FOOTNOTE BEGINS --------- [foot #] 6 The differences between proposal AJ and the disqualified Zenith proposal AX were relatively minor. The two proposals offered identical management solutions. CLIN 0001 offerings were identical under both proposals. Proposal AJ offered a slightly better 486 processor on CLIN 0002 and slightly higher video resolution. CLIN 0003 offerings were identical. The AJ system offered on CLIN 0004 had a larger hard drive than the AX system. Proposal AJ also offered a larger, more expensive laser printer than provided under AX. Otherwise the two proposals were virtually identical. The SSA found that AJ's higher price was attributable largely to the higher quality, more expensive laser printer. Protest File, Exhibit 1103. [foot #] 7 As a practical matter, the SSAC determined that it would not consider a pair of proposals originating from the same offeror, as this would not achieve any risk reduction or continued competition after award. ----------- FOOTNOTE ENDS ----------- achieved with virtually any pair of proposals selected. Similarly, it believed that competition between the two awardees would occur regardless of which vendors were selected. Thus, the SSAC focused on the third factor, achieving a range of offerings. Transcript at 182-83, 840; Protest File, Exhibit 279. 13. The SSAC returned to performing a detailed analysis of the spreadsheets in reviewing remaining proposal pairs CLIN by CLIN. Primary emphasis was given to the microcomputer systems (CLINs 0001-0004), the software (CLIN 0005) and the non-impact printer (CLIN 0007). Transcript at 825-26. These items were deemed to be the major discriminators for technical proposals. The primary management discriminator was warranty. Transcript at 182. 14. Looking at the ten pairs remaining, the SSAC performed another CLIN by CLIN analysis, this time to ascertain which pairs offered the user community the better choices of "management and technical solutions with appropriate price tradeoffs." Protest File, Exhibit 279; Transcript at 813-16, 838. When the SSAC deemed that a significant majority of orders would flow to one vendor, or where one proposal in the pair generally offered better technology at better prices, the pairing was discarded as unbalanced. Protest File, Exhibit 279. 15. This winnowing exercise reduced the better value set of pairs to five. Transcript at 816. The pairs that were eliminated were proposals R/M, R/Y, R/AJ, R/AD, and AQ/AD. All five pairs contained a Zenith proposal; four of the five resulted in discarding a pairing of EDS with another offer. Protest File, Exhibit 279. The SSAC did not consider that the eliminated pairs offered no reasonable choices, diversity of offerings or value. It simply concluded that the remaining pairs were better. Transcript at 815-16. Following this additional winnowing process, the pairs remaining in contention for award were R/AQ, M/AQ, AJ/AQ, Y/AQ, and U/AF (the alternate technology proposals). Protest File, Exhibit 279; Transcript at 815-16. 16. Specifically, in comparing the AJ and R solutions for purposes of determining if these two offers made a logical pair, the SSAC noted that although proposals R and AJ offered similar management solutions, AJ offered a higher technical solution (486 processors) than R (386 processors) on CLINS 0001 and 0002 at a lower price in years one and two, and for only $100 more per microcomputer system in year three. For CLIN 0003, the AJ microprocessor was slightly better, and less expensive in years one and two, while costing only $60 more in year three. The CLIN 0004AA solutions offered by AJ and R were virtually identical, but AJ provided lower prices in all three years. For CLIN 0004AB, AJ offered a slightly better system for a lower price in year one and for only $90 and $40 more in years two and three. Finally, with respect to CLIN 0007, the non-impact printer, AJ offered a somewhat more expensive, but clearly superior, product which the SSAC believed would be the user choice. Protest File, Exhibit 279. 17. The SSAC then developed four cost models for dual awards based on possible ordering patterns. Possible ordering patterns were derived by the conduct of group discussions, drawing from the varied levels of knowledge that the SSAC members possessed concerning likely user demand for products given pricing.[foot #] 8 Some members believed that lower prices would control. Others were aware of users with strong desires to obtain enhanced technology. After considerable discussion, compromise positions were reached in assessing likely ordering patterns. Protest File, Exhibit 279; Transcript at 809- 47. 18. The "low price" cost model depicted the situation in which the buyer would always select the lowest priced item. The "average" cost model represented an even split of orders between the two proposals. The "high tech" cost model represented the situation in which users would always opt for the most advanced technology, regardless of price. Finally, the remaining cost model reflected the ordering pattern developed by the SSAC based on its own estimate of likely order distributions.[foot #] 9 Protest File, Exhibit 279; Transcript at 817-20. 19. After reviewing the pairs and likely ordering patterns, the SSAC developed a consensus view that AJ/AQ, or the best Zenith proposal and the GTSI proposal, represented the best pair, or the best value combination. Transcript at 240, 850. The Second SSAC Best Value Analysis ----------- FOOTNOTE BEGINS --------- [foot #] 8 The SSAC represented a wide variety of user communities in the Air Force. The SSAC members had, collectively and individually, many years of experience in computer acquisitions. Each member brought his or her own particular perspective to the exercise. The varied backgrounds of these individuals resulted in detailed debates over the likely ordering patterns that would be experienced under each remaining pair. Eventually, after much debate, the group would reach a consensus as to what the split of orders would be for a given pair on a given CLIN. This is essentially how the estimated ordering patterns were derived in conducting the best value analysis for combinations. Transcript at 809-47. [foot #] 9 The SSAC, in deriving its own estimates, used three ordering ratios. If one product was deemed to enjoy a large advantage, the SSAC assigned 80 percent of orders to that offeror's product. A slight advantage garnered a 60 percent distribution of sales, and no advantage resulted in an even split of orders. Ratios were allocated to CLINs 0001-0004, 0005 and 0007, as comprising the major cost components of the proposals. Transcript at 844-56; Protest File, Exhibit 280. ----------- FOOTNOTE ENDS ----------- 20. The SSAC convened again on January 22. In the interim period following the first meeting, the SSAC chairman had become concerned with the potential risk with respect to deliveries inherent in Zenith's offers, all of which proposed the same model of monitor deemed noncompliant with the TAA under the Board's decision in CompuAdd Corp. After some discussion, the SSAC decided that the Board's recognition of the large size and strong capabilities of Zenith and its monitor manufacturer permitted the conclusion that although some additional risk might be present, it was not enough to change Zenith's relative ranking. Transcript at 249-55. 21. The SSAC then proceeded to retrace the steps it had taken in the earlier meeting, to derive a better value set of pairs in order to perform a dual award analysis. One set of pairs in particular generated considerable discussion. These were the AJ/AQ or Zenith/GTSI and R/AQ or EDS/GTSI pairings. Under the ordering patterns developed by the SSAC, the R/AQ pair was predicted to have a lower cost than AJ/AQ by some $61 million. The SSAC was concerned about this because the Zenith and EDS proposals, standing alone, were both evaluated at virtually the same cost. GTSI was a constant in the pairings. Upon examination, the SSAC determined that the price difference was attributable to the availability of lower priced, higher technology offerings under the Zenith proposal, which caused a shift in ordering patterns away from GTSI when the Zenith and GTSI proposal were paired. EDS's prices and products were not attractive enough to cause the same shift; users would purchase more heavily from GTSI, causing the overall price to be lower. Transcript at 819-22; Protest File, Exhibit 279.[foot #] 10 22. After performing its analysis, the SSAC concluded that an award to the AJ/AQ combination provided sufficient advantages over a single award to AJ to warrant a dual award to the pair. Transcript at 820, 846, 850. 23. On January 26, 1993, the SSAC chairman and a majority of SSAC members met with the SSA and repeated the analysis. Transcript at 850. The SSA concurred with the SSAC's consensus and requested assistance in drafting a source selection document. The source selection decision was signed on February 2, 1993. Protest File, Exhibit 280. 24. The SSA, like the SSAC, first conducted a single award best value analysis, reaching the conclusion that Zenith's proposal AJ constituted the single best value among proposals. He then turned to a review of the "logical pairs" identified by ----------- FOOTNOTE BEGINS --------- [foot #] 10 This theory was confirmed by calculating that when the percentage of buys from GTSI is kept constant, the price of the R/AQ and AJ/AQ pairs is approximately the same. Transcript at 713. ----------- FOOTNOTE ENDS ----------- the SSAC and also considered AJ paired with the Apple proposal (AF) and the Sun Sparc-based proposal (U) that had been added into the better value set by the SSAC.[foot #] 11 After reviewing the various combinations he determined that AJ/AQ, the lowest price combination containing AJ, the single best value proposal, consistently yielded a better value than other pairs, including R/AQ. Protest File, Exhibit 280. 25. The SSA noted the following in describing the advantages of the AJ/AQ pairing: This combination of proposals provides a solution of processors ranging from the low end to the high end of the Intel 386 and 486 processor technology. These two proposals provide a broad selection of 386 and 486 based processors at consistently competitive prices in all three purchase years of the contract. The capabilities offered by this pair on CLINS 1 and 2 allow the user a broad choice of technology and price to meet a range of operational requirements. CLIN 3, although the same processor on each proposal, offers differing memory and disk sizes to meet the user's differing needs. CLIN 4 offerings provide a choice of software development environments and runtime software environments on comparably configured systems. This combination also provides users a choice in printer support. Proposal AJ provides a heavy duty laser printer for highly networked environments while proposal AQ provides a laser printer suited for lighter duty in smaller offices with less printer demands. These proposals offer delivery commitments in excess of the total DOD requirement and the combination of the proposals would reduce the risk of any order "ramp up" or surge difficulties adversely affecting the user community. The paired combination also offers users a ----------- FOOTNOTE BEGINS --------- [foot #] 11 In considering these alternate technology pairs, the SSA concluded that they did not offer a best value to the Government when compared to the AJ/AQ combination. He reached this conclusion on the ground that the AJ/U (Sun Sparc) pair was considerably more expensive than the combination of Zenith and GTSI, for roughly the same level of performance. Although the Apple (AF) proposal was stronger than the GTSI proposal from a technological and management standpoint, the SSA was of the view that the Zenith proposal, with equivalent technology and lower prices, would be favored, and thus concluded that the AJ/AF pairing was imbalanced and significantly more costly than AJ/AQ. For the same reason, he determined that the U/AF combination, which would make available all three technologies, but at higher prices for systems with capabilities similar to those offered in conjunction with the Zenith/GTSI combination, did not present the best overall value to the Government. Protest File, Exhibit 280. ----------- FOOTNOTE ENDS ----------- selection of warranties with proposal AQ offering a 1 year warranty and proposal AJ a 3 year warranty. Cost for this combination is $685 million (SSAC Price estimate - Table 3.1). Protest File, Exhibit 280. Expert Testimony on the Air Force's Best Value Analysis 26. The challenge to the best value analysis conducted by the Air Force is based largely on the testimony of an expert witness. The expert agreed that the initial winnowing process, which reduced the number of pairs to be considered to ten, was reasonable and necessary to reduce the analysis to manageable proportions. Transcript at 325-26. He took issue, however, with the second winnowing, in which the Air Force eliminated from further consideration three of the four pairs that involved EDS. The need to achieve manageability no longer applies when only ten pairs are left, and, in his opinion, eliminated three pairs that might be very good best value candidates. He disagreed with the decision of the Air Force to eliminate pairings that, in the view of the SSAC, did not offer sufficient diversity of products and pricing to merit further consideration for a dual award. Transcript at 325-27. He concluded that as a result of this winnowing, there were some solid values that the Air Force never really evaluated. Transcript at 352. 27. The expert, reviewing the written text of the SSAC report, assigned ordering patterns to the discarded pairings, most of which involved a pairing of the single EDS proposal with that of another vendor. Protest File, Exhibit 1151; Transcript at 329. He generally agreed with the approach of using 80/20, 60/40 and 50/50 splits to predict ordering patterns. Transcript at 329. Based on the ordering patterns he believed the SSAC would have assigned had it considered these pairs, the expert concluded that the eliminated pairs in fact were fairly diverse in the sense that both offers garnered significant sales. Transcript at 334-35. 28. Ultimately, the expert concluded that the pairing of EDS's proposal, "R", with an alternate Zenith proposal, "Y", produced the "best value." He concluded this because, under his calculations, the evaluated price of the R/Y combination is lower than that of AJ/AQ and, based on color ratings, the R/Y combination appeared to him to be slightly stronger than the pairing of AJ/AQ. Transcript at 350-51. In his opinion, the combination of the higher, and more evenly, ranked proposals was a plus, because there would be a better alternative choice should one offeror not be able to deliver. Transcript at 351-52. TAA Compliance 29. Shortly after the dual awards were announced, in anticipation of protests, the contracting officer wrote to Zenith requesting a description of the manufacturing process applicable to the monitors scheduled for delivery to the Air Force for acceptance testing. Protest File, Exhibit 287. Zenith responded promptly, explaining that after issuance of the Board's opinion in CompuAdd Corp., Zenith had expended considerable efforts to redefine the manufacturing process for monitors to ensure that the "substantial transformation" test for compliance with the TAA would be met. Zenith further explained that a ruling from the U.S. Customs Service, informing as to whether the monitors would be considered to be substantially transformed in the United States, had been requested. Protest File, Exhibit 288. 30. The contracting officer testified that the Air Force was concerned about the heavy reliance on foreign sourcing for the DTIV procurement which was manifested by virtually all offerors. Transcript at 158. As required under applicable regulations, however, she relied solely on the certifications as submitted to determine compliance from this standpoint.[foot #] 12 The applicable regulations preclude seeking information about the manufacturing processes to be employed in producing these commercial items. The Air Force had no visibility into the manufacturing processes used by offerors for the products offered. It could not know, one way or the other, how products were manufactured or assembled. Changes to the assembly process that did not affect the model number were not part of any offeror's proposal. Transcript at 149-55. The Zenith Monitors 31. Following issuance of the Board's decision in CompuAdd Corp., Zenith representatives met with representatives of its monitor supplier. The monitor supplier indicated its desire to work with Zenith to develop a manufacturing process that could meet TAA requirements. Transcript at 633, 659. Two alternative assembly processes were developed, referred to by Zenith as scenarios II and III (with scenario I being the approach described in CompuAdd Corp.). Zenith's supplier agreed to perform either one to ensure TAA compliance. Transcript at 633. 32. Ultimately, Zenith selected scenario III, and that is the process proposed to be used in assembling the monitors to be supplied in conjunction with the present award. This process entails, among other things, importing the cathode ray tube (CRT) directly from [a TAA designated country] and assembling the integrated tube component (ITC) in the United States. The printed circuit boards (PCBs) used in the monitors would also be ----------- FOOTNOTE BEGINS --------- [foot #] 12 Zenith's certifications stated: "The Offeror certifies that the following supplies qualify as U.S. Made End Products. . . ." Protest File, Exhibits 34, 253. ----------- FOOTNOTE ENDS ----------- populated in the United States.[foot #] 13 The manufacturing process for the monitors has remained essentially the same; what has been modified is the location where some of the steps in the process are performed and the sourcing of certain parts. The model number of the monitors remains unchanged. Transcript at 566-67, 635-36, 652-53, 672. 33. The process adopted in scenario III is cumulative. That is, all of the procedures performed in the United States under scenario I (the process addressed in CompuAdd Corp.) will be performed under scenario III and others, formerly performed overseas, will be added to the group of procedures to be performed in the United States. Transcript at 652-54; Protest File, Exhibit 288. 34. Zenith's monitor supplier, in a request dated January 28, 1993, applied for a definitive ruling from the U.S. Customs Service concerning whether the processes contemplated for assembly of the monitors satisfied the statutory requirements for substantial transformation. Zenith filed a similar request, seeking an exception from country of origin marking, on February 3, 1993. Protest File, Exhibit 288. 35. In requesting these rulings, Zenith and its monitor supplier actually requested that Customs consider some six scenarios. The first three described various processes to be performed in the United States. Scenario I was the process addressed in the Board's decision in CompuAdd Corp.; scenario II was a process which added significant steps to scenario I; scenario III added yet more steps (basically building of the PCBs) to the process. In a supplemental request, filed on February 16, Zenith's supplier requested a ruling on scenarios IV, V, and VI. These scenarios are identical to scenarios I, II, and III, respectively, except that the assembly process would be performed in Singapore, a designated country under the Trade Agreements Act. A ruling on these latter scenarios was requested because Customs has taken the position that it is not authorized to render final determinations concerning whether an article is a product of the U.S. for purposes of Title III of the Trade Agreements Act. 36. In Ruling HQ 734977, issued on April 2, 1993, the Customs Service determined that Scenarios V and VI, which involve the identical procedures described in scenarios II and III, clearly constituted substantial transformation in a designated country, Singapore, under the TAA. As expected, the Customs Service did not address the scenarios involving assembly processes in the United States. Finding 34. ----------- FOOTNOTE BEGINS --------- [foot #] 13 "Populating" the boards refers to the placement of the individual components on the raw circuit boards. Transcript at 567. ----------- FOOTNOTE ENDS ----------- 37. As of January 7, 1993, Zenith's monitor supplier had affirmed that it was "ready, willing, and able" to manufacture monitors using the procedures outlined in scenarios II and III. Transcript at 633. By January 20, 1993, Zenith had decided to direct the monitor supplier to pursue scenario III as opposed to scenario II. Transcript at 674-75. The monitor supplier immediately began preparations to perform accordingly. Transcript at 578-79. 38. In proposal AJ, Zenith agreed to deliver 20,000 microcomputer systems in month one of the contract and 25,000 systems per month thereafter. Protest File, Exhibit 246. Under the contract, the Air Force must first issue a delivery order for acceptance test equipment, to be delivered within ten days. Testing of the equipment should take approximately two weeks. Transcript at 705. Taking into account the internal procedures required for ordering off the Desktop IV contracts, the Air Force estimates that it would be in a position to issue the first purchase orders to Zenith around thirty days after successful completion of acceptance testing. Transcript at 679, 705-06. The contract provides for acceptance of a purchase order within two weeks after receipt by the contractor. The first delivery is required to be made within twenty-one days after order acceptance. Protest File, Exhibit 16. All told, Zenith calculates that it would have at least three months, after issuance of a decision permitting the Air Force to proceed with awards, before it would be required to deliver under the contract. It thus estimates that the earliest it would be required to be able to deliver 20,000 systems would be by July 19, 1993. 39. On March 12, 1993, Zenith issued a purchase order to its supplier for incremental delivery of 20,000 monitors to be completed by July 5, 1993. The purchase order specifies that the monitors be manufactured in accordance with scenario III. Transcript at 619. The monitor manufacturer has already issued purchase orders to U.S. circuit board suppliers to ensure an adequate supply will be available so as to satisfy this obligation to Zenith. Protest File, Exhibits 3073, at H01021, H01022, H01024, H01025; Transcript at 639-40. The circuit board suppliers are well established companies in the business. Should they have any problem meeting the demand, however, there is an overabundance of circuit board suppliers located in the Silicon Valley area of California such that Zenith's monitor supplier anticipates that there could be no conceivable difficulty in obtaining an adequate supply of PCBs. Transcript at 636-47. 40. Zenith's monitor supplier is one of the largest manufacturers of computer monitors in the world. It has a long- standing business relationship with Zenith and is committed to expend the resources necessary to meet its obligations in connection with Desktop IV deliveries. Transcript at 555, 593, 622, 635, 681. Discussion The Board does not decide the current protests with a tabula rasa. Although no pre-award protest of the terms of the solicitation was ever filed, this is the third set of post-award protests involving this procurement. We must resolve the issues before us here not only on the basis of the record presented in this round, but also in the context of the record developed, and the decision issued, in connection with the second round of protests. In CompuAdd Corp., the Board found that the monitors proposed by Zenith under the offer selected for award did not meet the requirements of the TAA and therefore held that the award to Zenith was invalid. It also found that the Air Force had not conducted an appropriate dual award analysis that comported with the criteria set forth in the RFP. The Board granted the protests and revised the Air Force's delegation of procurement authority: Keeping in mind these policy considerations, we conclude that the award to Zenith cannot stand. The Air Force may proceed with award based on an analysis of remaining proposals if it believes these proposals can satisfy its needs. CompuAdd Corp., 1993 BPD 48, at 53. The Board added that if the Air Force pursued this course of action, it should conduct a best value analysis that gave "due consideration to combinations of proposals, which, given likely ordering patterns, may lead to the best overall value to the Government, assuming no loss of technical advantages and the ability to save money." Id. The Board also recognized that the Air Force might prefer to revise or clarify the solicitation and permitted this alternative as well. Finally, the Board recognized that other approaches might also be available and authorized the Air Force to proceed in any other fashion that accorded with applicable statute and regulation. Id. Following the announcement of the dual awards to Zenith and GTSI, EDS filed at the Board a motion seeking issuance of an order to show cause, enforcement of the Board's order, and suspension of the procurement. That motion argued that the Air Force had violated the Board's order by leaving the alternate Zenith proposals in the competition despite the fact that all the monitors offered were the same model number. Specifically, EDS construed the Board's decision as eliminating Zenith altogether from further consideration absent a re-opening of discussions leading to another round of BAFOs. After filing its protest of the agency's award decision, EDS requested that the Board treat its motion to enforce as a dispositive motion under its protest. The Board deferred ruling on this motion, and on other dispositive motions filed by the parties, however, because factual disputes, directly relevant to resolution of the issues presented, required development of a record. Most of the arguments raised in the motion to enforce were subsumed within the subsequent protests. Accordingly, these arguments remain for resolution on the merits, Electronic Data Systems Corp. v. Department of the Air Force, GSBCA 12300-P (Apr. 28, 1993), and will be addressed below. One point asserted in that motion deserves comment at the outset, however. This concerns the allegation that the Board's decision in CompuAdd Corp., holding that the AX proposal was noncompliant with a requirement of the solicitation, a priori required the exclusion of all Zenith proposals from further consideration unless the Air Force conducted discussions and called for another round of BAFO's. The Board's decision addressed only proposal AX; it did not look at other Zenith proposals. Zenith as a competitor was not disqualified. The Board found only that its AX proposal was not compliant. Thus, the decision did not automatically preclude the Air Force from considering the alternate Zenith offers in the manner that it did. Trade Agreements Act Compliance In CompuAdd Corp., the Board examined the compliance of Zenith's proposed monitors under the Trade Agreements Act because the protesters alleged that the parts for the monitors were manufactured in a non-designated end country and that the process by which the monitors would be assembled in the United States was not sufficient to constitute substantial transformation. After the Board concluded that the monitors offered by Zenith, at least as demonstrated at the time of the hearing, were not compliant with the TAA, Zenith considered ways in which to rectify this problem with respect to other proposals. Noting that the Board had recognized in its decision that it should look to Customs law, Zenith decided that the appropriate mechanism to ensure that its monitors would not be disqualified would be to seek a ruling from the U.S. Customs Service. The ruling was requested pursuant to the authority of the U.S. Customs Service to make determinations as to whether a product is "substantially transformed" within the meaning of the TAA. Specifically, on January 28, 1993, Zenith's supplier submitted a request for a country-of-origin final determination relating to a Government procurement. A few days later, on February 3, Zenith submitted a request for an exception from the country-of-origin marking requirements of 19 U.S.C. 1304 (1988). Several weeks later, Zenith's supplier amended its request for a ruling, asking the Customs Service to determine also whether the identical three scenarios would constitute substantial transformation if performed in Singapore. On April 2, 1993, the Customs Service issued a final determination concluding that the second and third scenarios, as performed in Singapore, would constitute substantial transformation in Singapore.[foot #] 14 The U.S. Customs Service is the administrative agency vested with authority to issue binding determinations with respect to the origin of an article under the Government procurement provisions of Title II of the Trade Agreements Act of 1979. See 19 U.S.C. 2515(b) (1988). The TAA authorizes the issuance of advisory rulings and final determinations. A final determination is judicially reviewable by means of an action commenced by any party-at-interest. 28 U.S.C. 2631(e) (1988). The Customs Service has developed considerable expertise in its day-to-day application of the substantial transformation concept in various customs law proceedings. See, e.g., 19 CFR 10.14(b), 10.177(a)(2). Zenith argues that a final determination is binding upon all concerned until such time as it is set aside in the judicial review process. As such, Zenith contends that the Board is bound by the Customs ruling and, at the very least, must accord it substantial deference. The Board agrees that, at a minimum, the determination rendered by the Customs Service respecting the assembly process in Singapore deserves exceptional weight. Although Zenith proposed to deliver U.S. made end products, the evidence in the record demonstrates that the process addressed in the Customs final determination as to Singapore is identical to the process that would be employed in the United States. The rationale of the Customs Service determination should clearly apply equally to the same process implemented in the United States. The Board has held that products of the United States that meet the TAA's rule of origin must be treated as exempt from the exclusion authority of that Act. International Business Machines Corp., GSBCA 10532-P, 90-2 BCA 22,924, 1990 BPD 125. Accordingly, in light of the evidence of record concerning the operations to be performed in the United States, and the specific determination, issued by the Customs Service, addressing these precise operations, we find that the monitors as proposed to be supplied by Zenith under proposal AJ are fully compliant with the solicitation's TAA requirement. A second prong of this argument advanced by protesters is that even if the monitors, under the process currently contemplated, meet the TAA requirements, Zenith was not, and will not be, in a position to supply the quantities it is obligated to provide under the contract. Again, protesters point to the fact that the monitor assembly process will be a start up operation in the United States for Zenith's supplier. The Board effectively rejected these "production capacity" arguments in its decision in ----------- FOOTNOTE BEGINS --------- [foot #] 14 As noted in Finding 35, the Customs Service does not consider that it has the authority to issue a binding determination on the issue of substantial transformation with respect to processes that take place in the United States. ----------- FOOTNOTE ENDS ----------- CompuAdd Corp. and does so again here. There is ample evidence in the record that Zenith and its supplier, both of which are large corporate entities, are fully capable of meeting "worst- case" production requirements. That is, even assuming the Air Force orders the maximum number of systems it is entitled to order as soon as it is able to do so, Zenith will still be able to meet its obligations.[foot #] 15 Findings 38-40. Moreover, the SSAC took into account evidence adduced in the earlier protest to the effect that the monitor assembly process would be a start-up operation in the United States, and concluded that Zenith's relative ranking would not be affected. Finding 20. The Extension of Proposals and Improper Discussions Protesters contend that the Air Force improperly conducted discussions with Zenith, but with no other offeror, to the prejudice of other offerors. According to protesters, the Air Force was lacking essential information it needed to confirm the acceptability of Zenith's remaining proposals and the request for extensions of proposals gave Zenith an opportunity to communicate to the Air Force that its proposals had somehow been made acceptable.[foot #] 16 ----------- FOOTNOTE BEGINS --------- [foot #] 15 There is ample evidence in the record that the assembly process itself is fully developed, and can readily be performed in the United States; that Zenith's supplier is one of the largest such manufacturers in the world; and that it is fully committed to making whatever efforts are necessary to ensure timely delivery of the quantities required. We noted in CompuAdd ________ Corp. that other offerors had similarly relied on representations _____ from their suppliers that the quantities would be supplied as necessary without having in place firm orders for those items. CompuAdd Corp., 1993 BPD 48, at 34 n. 15. Given the RFP's _____________ requirement that acceptance testing be passed before any orders could be placed, we did not find that Zenith had misrepresented its abilities to deliver, but observed only that this additional risk might have affected its evaluation had the Air Force been aware of it. As we note above, the Air Force considered this information in making this award. [foot #] 16 A the hearing, protesters also stated that telephone "communications" taking place between Zenith personnel and the chairman of the SSAC also gave Zenith the opportunity to convey this information. Zenith notes that these communications were in the nature of "innocent conversations" lamenting the outcome of the Board's decision before Zenith saw even limited portions of that decision. Transcript at 56, 62. Although the Air Force and Zenith conceded, in responses to interrogatory questions, that these conversations occurred, no testimony concerning these conversations was elicited from the chairman of the SSAC or from Zenith personnel. Protesters have not, in the (continued...) ----------- FOOTNOTE ENDS ----------- The Federal Acquisition Regulation (FAR) defines what constitutes discussions in the context of a negotiated procurement: "Discussion," as used in this subpart, means any oral or written communication between the Government and an offeror (other than communications conducted for the purpose of minor clarification), whether or not initiated by the Government, that (a) involves information essential for determining the acceptability of a proposal, or (b) provides the offeror an opportunity to revise or modify its proposal. 48 CFR 15.601 (1992). In essence, protesters urge that having been "educated" by the prior protests and the Board's decision, the Air Force was required to assume that the remaining Zenith proposals were also ineligible for award by reason of the failure to offer products that were compliant with the Trade Agreements Act. Specifically, the Air Force knew that Zenith had offered the same model monitor in all seven of its proposals. The protesters thus urge that the Air Force had no alternative but to deduce that all of these offers suffered from the same defect -- that is, there was no reasoned basis for concluding that Zenith planned to use a different assembly process for other proposals. According to protesters, the Air Force's only choices were either to exclude all Zenith proposals from further consideration or to reopen discussions in order to determine that the proposals were actually compliant. In response, the Air Force and intervenors Zenith and GTSI emphasize that the underlying assembly process is nowhere described in any of the proposals submitted in response to the solicitation. The Air Force had no insight into how any of the many products of foreign origin were going to be manufactured or made compliant with the requirements of the Trade Agreements Act nor, under a functional specification for commercial equipment, would it ordinarily be permitted, let alone obligated, to inquire into these matters. Defense Federal Acquisition Regulation Supplement (DFARS) 211.7004-1(d), 48 CFR 211.7004-1(d) (1991). According to the Air Force, the contracting officer properly relied on each offeror's self-certification, which she considered to have been reaffirmed by the extension of that offeror's proposal. ----------- FOOTNOTE BEGINS --------- [foot #] 16 (...continued) post-hearing briefs, relied on these conversations to support the assertion that improper discussions occurred. We find that these conversations do not constitute discussions. ----------- FOOTNOTE ENDS ----------- The regulations implementing the provisions of the TAA include FAR 25.407, which provides that the solicitation shall include two clauses, one providing for a certification, and further provides that the contracting officer "shall rely on the offeror's certification as submitted." Although the FAR's TAA- compliance requirements as applicable to DOD procurements are varied through substitution of a modified DFARS contract clause also providing for certification of TAA compliance, which was used by the Air Force in this solicitation,[foot #] 17 the FAR's requirement that the contracting officer rely on the certification is not affected. In addition, respondent, Zenith, and now GTSI, once again argue that this subject matter is one primarily related to contract administration, and not to contract formation. Thus, the Air Force contends that it is entitled to rely completely on the certifications supplied by offerors and that to the extent an offer may not be compliant this should be an issue to be resolved by the agency during performance of the contract. The time for compliance with the TAA to be determined, these parties urge, is at the time of performance, and not at the time of formation. We cannot agree that this issue should be so readily dismissed. We do agree that the cited regulation explicitly requires the contracting officer to rely on the certification in determining compliance for the purposes of conducting evaluations and selecting candidates for award. The Board's decision in CompuAdd Corp. did not establish that the Air Force had or has an obligation to verify the accuracy of a self-certification prior to award. Thus, we find that the contracting officer proceeded in a proper manner in accepting the certifications at face value until after award had been made. At the same time, once that certification has been challenged, the Board has recognized that this is also a formation issue requiring the Board to look beyond the four corners of the certification to whether the product offered is in fact compliant. Rocky Mountain Trading Co., GSBCA 10894-P, 91-1 BCA 23,619, 1990 BPD 430. See Sysorex Information Systems, Inc., GSBCA 10642-P, et al., 90-3 BCA 23,181, 1990 BPD 193. We look at these issues because an agency cannot achieve full and open competition if all offerors are not required to propose compliant products and are not evaluated on that basis. Having received such a challenge in CompuAdd Corp., the Board conducted a de novo review and held only that, based on the evidence and legal analysis presented to it in that proceeding, the monitors to be supplied under the Zenith proposal selected for award did not meet the TAA requirement for substantial transformation in a designated end country. The Air Force was under no obligation, based on the ----------- FOOTNOTE BEGINS --------- [foot #] 17 Pursuant to DOD regulations implementing and supplementing the FAR, the RFP contains DFARS 252.225-7006, 48 CFR 252.225-7006 (1991), which, like the pertinent FAR clause, provides that offerors must certify to TAA compliance. ----------- FOOTNOTE ENDS ----------- Board's decision, to assume that the other proposals were noncompliant and to reject those proposals or reopen discussions. Protesters also make much of Zenith's unsolicited letter extending its proposal AX and adding a paragraph indicating its intent to appeal. Finding 6. They argue that this additional paragraph somehow alerted the Air Force that the monitors were now TAA-compliant. The Board is hard pressed, from perusing this paragraph, to ascertain just how this communication can be construed, even read in light of the other letters, to have provided any essential information to the Air Force. It simply extended Zenith's offer pending resolution of the planned appeal. This is not the type of communication that is within the purview of the regulation. On balance, protesters' argument that the request for extensions of proposals constituted discussions is not persuasive. It is predicated upon the implication that the Air Force really required extensions only from Zenith in order to make the quantum leap of faith that the TAA certifications had been validated. In point of fact, the extensions were necessary across the board since all proposals had expired. The contracting officer's reliance on the extensions and her conclusion that the certifications had been revalidated in the case of all offers, was not unreasonable, and was not a conclusion she reached based on any express communications, but on her expectation that offerors would not extend their proposals if they could not, as of the date of extension, comply with all material terms of the solicitation.[foot #] 18 Cf. Wickman Spacecraft & Propulsion Co., B-219675, 85-2 CPD 690. Protesters suggest that discussions occurred by innuendo. We cannot discern that any impropriety occurred. No specific information was communicated to the Air Force by Zenith that was not also communicated by all other offerors. Zenith did not need to change any of its proposals in any way; the monitor assembly process simply was not part of any proposal. Zenith, and all other offerors, were free to make such adjustments at any time, so long as the product proposed did not change its salient features.[foot #] 19 ----------- FOOTNOTE BEGINS --------- [foot #] 18 Moreover, since many offerors relied heavily on foreign sourcing, in fact, the "opportunity" to make adjustments to manufacturing processes in order to ensure TAA compliance was equally extended to all offerors who certified that the systems offered complied with the TAA. [foot #] 19 We are similarly unpersuaded by the argument that post-BAFO discussions occurred. The evidence establishes only that the contracting officer, in response to EDS's challenge to the compliance of the monitors, properly requested that Zenith elaborate on the basis for its self-certification of compliance. This did not constitute improper post-BAFO discussions. Zenith (continued...) ----------- FOOTNOTE ENDS ----------- The Best Value Analysis The solicitation provided that the Air Force would make an award to the proposal or combination of proposals which offered the "best overall value to the Government." Protesters assert that the analysis conducted by the Air Force in arriving at the conclusion that the Zenith/GTSI proposals, in combination, offered it the best value was fatally flawed. The gravamen of this argument is that the analysis lacked logic and disregarded certain combinations which the expert suggests might alternatively have afforded the best value. At the outset, we note that the solicitation accords the Air Force considerable discretion in evaluating proposals and in conducting a best value analysis. As the Board observed in CompuAdd Corp., in many respects this acquisition resembles a schedule contract for purely commercial off-the-shelf equipment. Nonetheless, the award criteria specify that the selection decision will be based on an integrated assessment of which offer or combination of offers provides the "best overall value to the Government." The only parameters established are that technical and management factors are more important than cost. The RFP does not provide any weights by which the significance of these factors, as compared to cost, is to be gauged. No offeror protested the terms of the RFP prior to the closing date for receipt of proposals. It is too late to do so now. What constitutes an appropriate, or adequate, best value determination by a Government agency has, in recent years, formed the subject of considerable protest litigation. The standards which have evolved, however, make clear that it is not enough merely to show that the Government's analysis could have been conducted differently or may have contained flaws: Our task in reviewing [a best value] analysis is not to substitute our judgment or the judgment of expert witnesses for that of the agency officials. The legal standard against which we review best value determinations is not one of perfection or even accuracy, but rather one of reasonableness. Lockheed Missiles & Space Co. and International Business Machines [Corp.] v. Department of the Treasury, GSBCA 11776-P, 11777-P (TMAC II), 1992 BPD 155, at 20. Indeed we have upheld a best value determination where "some elements of the best value analysis lack a rational basis." Computer Sciences Corp., GSBCA 11497- P, 92-1 BCA 24,703, at 123,297, 1992 BPD 6, at 32 [1991]. There is no formulaic methodology for conducting a best value determination; what matters is ----------- FOOTNOTE BEGINS --------- [foot #] 19 (...continued) was not afforded any opportunity to modify its proposal as a consequence of this communication. ----------- FOOTNOTE ENDS ----------- that award is consistent with the terms of the solicitation and that any price premium is justified by specific technical enhancements. TMAC II, 1992 BPD 155, at 20. Grumman Data Systems Corp., GSBCA 11939-P, 1993 BPD 2, at 56 (Nov. 20, 1992) (JSAN II). The Government is permitted to exercise considerable discretion in conducting a tradeoff analysis to determine what constitutes "best value." Under a solicitation like that used in this procurement, the Government is not bound to adopt any particular method of evaluating "best value." The Board has observed in this regard: Although the analysis that the Working Group undertook undoubtedly has numerous assumptions and some margins of error, we do not lose sight of the basic principle that, in making its analysis, the agency is essentially exercising its business judgment, albeit one involving taxpayers' money, and not conducting a definitive or all-inclusive study. Lockheed Missiles & Space Co. v. Department of the Treasury, GSBCA 11776-P, et al. (TMAC II), 1992 BPD 155, at 20-21 (June 2, 1992). Here, as in TMAC II, the solicitation offered considerable leeway to the agency in conducting a tradeoff. The mere fact that other approaches might have been adopted, or that other factors could have been taken into account as well, does not render a tradeoff analysis inconsistent with the evaluation criteria of the RFP, or otherwise cause it to be fatally flawed. Id., 1992 BPD 155, at 20. The Zenith proposal selected for award in this round (AJ) was not significantly different from the proposal selected previously (AX). The management proposals were identical. From a technical standpoint, proposal AJ offered basically the same processors on CLINS 0001 through 0004 and the same software. The major difference in the two Zenith proposals was the higher quality, more expensive laser printer offered under proposal AJ. The Air Force used the same procedure to identify proposal AJ as the "single best value" proposal in this round as it had applied in the previous round. In CompuAdd Corp., 1993 BPD 48, at 46, the Board noted that the steps taken by the Air Force in identifying proposal AX as the "single best value" proposal were not in issue. The Board could not find that the Air Force had fully followed its stated award criteria, however, because the record did not reflect the same thorough consideration of the alternate possibility provided for -- that of a dual award to a combination of proposals. Following the issuance of CompuAdd Corp., the SSAC developed a specific methodology by which it would undertake to evaluate this alternative possibility. That methodology essentially consisted of performing an initial single award best value analysis in virtually the same manner as the process which led to the selection of proposals AX. This reduced the group of better value proposals to five. After comparing these five proposals, the SSAC concluded that Zenith's proposal AJ represented the best single value. Given the technical superiority offered by proposal AJ, at relatively reasonable prices, we find that this conclusion was well justified on the basis of the underlying evaluations. The SSAC then turned to the task of assessing pairs of proposals, with the intention of deriving the combination of proposals that offered the best overall value to the Government. It proceeded in much the same manner as it had adopted in determining the best value individual proposal. Where protesters argue the process was flawed was in the decision to eliminate certain pairs without performing the exercise of assigning the probable ordering patterns which enabled the SSAC to quantify the estimated costs that would be experienced under various combinations of proposals. Specifically, protesters maintain that the SSAC's decision to discard these pairings as not offering sufficient "diversity" was arbitrary and unreasonable. In addition, they claim that the dual award analysis was flawed by reason of the failure to justify selection of the higher priced AJ/AQ solution over the R/AQ solution. Admittedly, the arguments made by the expert with respect to the discarded pairs have some theoretical merit. The arguments advanced, unsuccessfully, by protesters in TMAC II and JSAN II showed that it is invariably possible, with the benefit of hindsight, to find weaknesses in the tradeoff process and to concoct alternative means by which an analysis of this nature might or could be conducted. This is a far cry, however, from establishing that the process employed by the Air Force lacked a reasonable or rational basis. The Air Force explained its rationale for eliminating these pairs in the SSAC and SSA reports. That explanation was reasonable and simply showed the exercise of sound business judgment on the part of the SSAC and the SSA. The SSAC, which represented a broad cross-section of Air Force users, reviewed in a detailed, methodical fashion the management, technical, and pricing options offered by such combinations as R/Y and R/M (the pairs that the expert alleged could or should have been selected as presenting the best overall value to the Air Force)[foot #] 20 and determined that there were reasons why it did not make sense to pursue these combinations to the point of attempting to quantify likely ordering patterns. In assessing the combinations offered by pairing proposals in the "better value" group arrived at under a single award ----------- FOOTNOTE BEGINS --------- [foot #] 20 R/Y and R/M represented pairings of the EDS proposal with two alternate Zenith proposals. Finding 15. ----------- FOOTNOTE ENDS ----------- analysis, the Air Force was looking for particular advantages to be derived from a dual award -- these included a wider variety of technical solutions, reduced risk of meeting needs, and the promotion of competition between awardees throughout the life of the contract. The two latter advantages would flow from any pairing of the better value proposals. Thus, the major factor that was looked for in winnowing pairs was the choice of technical and management solutions and pricing advantages offered. With respect to both R/Y and R/M, the SSAC, after conducting a detailed comparison, concluded that these pairs did not offer sufficiently diverse technical and pricing solutions to result in balanced ordering patterns. Similarly, in comparing the pairing of the AJ/AF (Zenith/Apple) solution with the AJ/AQ (Zenith/GTSI) solution, the SSA concluded that the considerable additional cost associated with Apple's alternative technology was not justified. The fact that protesters disagree with the assumptions used and conclusions reached in conducting this winnowing process does not make them unreasonable. Protesters' allegation that the judgments of the SSAC and the SSA lacked a rational basis simply cannot withstand scrutiny. They complain that the SSAC never defined the term "diversity." Both the SSAC report and the testimony of the Air Force officials who participated in this process explained the winnowing approach satisfactorily. The SSAC was examining pairs of proposals, looking on a CLIN by CLIN basis at technical and management solutions and considering price. If a combination of proposals was geared toward offering high end technical solutions, and it appeared to the SSAC that one proposal would be heavily favored by users as a result of pricing considerations, that pair was eliminated. There is nothing fundamentally unsound about the SSAC's premise that the best overall value for the Government under a dual award would be achieved by providing users with the ability to procure a range of technology at a range of pricing, thus, for example, serving high end users desiring 486 processors, as well as more budget-constrained users whose needs could be met by GTSI's more inexpensively priced 386 processors. Nor is this premise inconsistent with the solicitation's stated award criteria. The SSAC was not required to perform the further exercise of quantifying ordering patterns for the discarded pairs to reach a reasoned judgment as to which combination offered the best overall value. Finally, we address the contention that the price "premium" associated with the selection of the Zenith/GTSI pairing over the EDS/GTSI pairing was not adequately justified.[foot #] 21 The record on this point is to the contrary. The SSAC recognized that this was a "glitch" in ----------- FOOTNOTE BEGINS --------- [foot #] 21 We recognize that this argument is primarily of concern to EDS but we address it in the context that protesters are challenging whether the best value analysis was rational as conducted. ----------- FOOTNOTE ENDS ----------- the sense that both the Zenith and EDS proposals were approximately equal in price and the GTSI proposal was a constant. It thus explored the reasons underlying its determination that the Zenith/GTSI combination would be higher priced, by some $61 million. It concluded that this would occur because users would opt to purchase more Zenith systems, which offered superior processors at lower prices than were offered under the EDS proposal, at least with respect to CLINs 0001 and 0002. The RFP states that technical superiority will be accorded more weight in the selection process than cost. Here, payment of the price premium has been sufficiently supported in the record. Finding 21. Based on the written record and testimony adduced at the hearing, the award decision is amply justified. The fact that other possible selections might also have been justified had the SSAC adopted different criteria in winnowing pairs does not render the decision it made irrational, unreasonable, or otherwise inconsistent with the terms of the solicitation. That decision is well-documented and fully supported by the testimony of Air Force officials who participated in the selection process. Protesters have presented no basis on which to overturn the tradeoff analysis performed by the Air Force. Changed Requirements Finally, we turn to the arguments presented by CompuAdd and Apple concerning the changed requirements of the Air Force. Protesters contend that in the previous hearing Air Force officials made a number of statements indicating a distinct preference for obtaining technically superior 486 processors under CLINS 0001 and 0002 to the virtual exclusion of the products proposed in the vast majority of offers, which for these basic and intermediate user CLINs were predicated on 386 processor technology. Protesters argue that this preference was not reflected in the solicitation and should have been. They also point to the Board's expressed concerns about this issue in CompuAdd Corp. In deciding the prior protests the Board did state its concern that, to the extent the Air Force would not under any circumstances consider an award that included 386 technology, the solicitation might not have properly informed offerors of this requirement. The issue arose largely because the Air Force, in defending its selection decision to make only one award, rather vigorously championed the 486 processors as the unit of preference for any user. Despite protesters' claims to the contrary, the dual award decision made by the Air Force puts to rest any claim that 486 technology was unduly favored outside the reasonable parameters of the solicitation. The solicitation does contain broad discretion to make an award in which technical and management factors are weighted more heavily than cost. The Air Force is explicitly permitted under the solicitation to favor an enhanced technical solution. Certainly nothing in the RFP precludes an award to an all-486 processor solution, so long as the overall best value is achieved. The solution offered by Zenith in proposal AJ offered higher level technology and management at a reasonable cost. The selection of this proposal occurred here in the context of a dual award that included an array of 386 and 486 machines, in much the same manner as did the first set of awards, to vendors offering a range of 286 and 386 processors. The current award decision belies the notion that the Air Force was determined to avoid any award that included the less powerful 386 machines on the basic or intermediate CLINs no matter what other advantages might be derived from such a proposal. Moreover, the Board's decision in CompuAdd Corp. expressly allowed the Air Force to pursue this approach. Consequently, this ground of protest is also denied. In closing, we feel compelled to address to some extent the dissenting opinion. A significant portion of that opinion is addressed to perceived defects in the solicitation, the terms of which have never been protested, either in the instant case or in the two prior protests arising under this procurement. The Board's Rules and well-settled case law, like that of the Comptroller General, preclude protesting the terms of a solicitation after the closing date for receipt of proposals. A major purpose of this rule is to discourage offerors from gambling in the "lottery" and then complaining about the conduct of the lottery after some other offeror wins the "pot of gold." Much of what the dissent complains of is simply and properly beyond the purview of these protests. Our Rules and precedent do not permit us, at this stage in the game, sua sponte to invalidate the solicitation. Although it might, at any point in a procurement, be appropriate to invalidate a solicitation that would lead to the award of a contract that would be void or patently illegal, we are not faced with such a situation here. Decision The protests of CompuAdd Corporation and Apple Computer, Inc. are DENIED. The suspension of the Government's procurement authority lapses by its terms. _______________________________ CATHERINE B. HYATT Board Judge I concur: ___________________________ JAMES W. HENDLEY Board Judge VERGILIO, Board Judge, dissenting. I respectfully dissent from the decision of the majority-- two independent bases compel the granting of the protests. The agency should have deemed unacceptable the Zenith Data Systems Corporation proposal which underlies the award. Zenith misrepresented its ability to deliver monitors complying with the Trade Agreements Act. Further, apart from the errors surrounding Zenith's ability to deliver monitors, the structure of the procurement fatally tainted the selection determination. Given the solicitation and the offers, the agency was not in a position to determine that a single or dual award was most advantageous to the United States. The solicitation did not encourage competition of the actual requirements the agency sought to satisfy. Although offerors were not restricted in what they could propose, and could have offered the systems most desired by the agency, the solicitation does not put offerors on notice of the agency's true preferences. Rather, the solicitation is akin to a lottery, with offerors gambling on foreseeing the technology the evaluators will view as representing the "best value." Such a solicitation does not comply with requirements for developing specifications and for competition, and does not lead to an award necessarily most advantageous to the United States. Further, the solicitation and the source selection authority anticipate that the actual awardee(s) would alter pricing and technology available under the contract. This would occur beyond simple upgrades to hardware and updates of software; the solicitation envisions a contractor adding line items to the contract without requisite competition for those now undefined requirements. This broad approach goes beyond the bounds of legitimate contract practices as it undermines the solicitation's stated requirements and the bases of evaluation and selection. Findings of Fact 1. The agency issued a solicitation to obtain various commercial items under a firm, fixed item-price, indefinite delivery, indefinite quantity contract(s). Protest File, Exhibit 16 at 11 ( B-1), 127 ( L-3). The solicitation was originally issued in July 1991, with award(s) anticipated within six months, without discussions. In particular, through one or two contracts awarded under the solicitation, the agency intended to satisfy the Government's "requirement for microprocessor-based microcomputers, applications software, including upgrades and updates to all delivered software, printers, backup storage devices, removable storage devices, user-installable components, and Contractor-provided support services to operate primarily in a networked office environment." Id. at 29 ( C-1). The systems are to be used worldwide in the Department of Defense (DoD) and other Government agencies. Id. at 29 ( C-2). The contract is non-mandatory for DoD and other federal agencies. Id. at 66 ( H-18). 2. Contract line items (CLINs) include microcomputer systems (basic, intermediate, and advanced), software, hardware, extended warranties, and maintenance training. Protest File, Exhibit 16 at 11 ( B-1), 13-27 ( B-3, B-tables); CompuAdd Corp. v. Department of the Air Force, GSBCA 12021-P, et al., 1993 BPD 48, at 7-9 (Dec. 23, 1992). Any contract is to be for one base year, with two one-year options, plus an additional two option years for user-installable components and extended warranty. The "minimum amount" (i.e., the guaranteed dollar value of delivery orders) for a single contract is $9.5 million, and for each dual award is half of that. Protest File, Exhibit 16 at 11 ( B-2). Any award would be for all CLINs (including sub-CLINs); that is, awards were not to be based on or solely for a particular CLIN or group of CLINs. Id. at 12. 3. A contract clause expressly permits equipment modification and substitution after award. "The Government may allow equipment alterations, modifications, etc., when in the interest of the Government to do so." Protest File, Exhibit 16 at 63 ( H-12). The contract treats such modifications and substitutions as distinct from contractor/original equipment manufacturer sponsored modifications, id. at 64 ( H-13), and from contractor furnished software upgrades/updates, id. at 61 ( H-10). The solicitation also contains the Changes-Commercial Items (May 1991) clause from the DoD Federal Acquisition Regulation Supplement (DFARS), 48 CFR 252.211-7002 (1992). Id. at 85-86 ( I-10). The chairman of the source selection advisory counsel (SSAC) has testified that with two awards, he "would anticipate the market place competition to probably cause one or the other offeror to readjust the proposal that had been provided . . . to offer changes in prices, maybe change the technology." CompuAdd Corp., 1993 BPD 48, at 7 ( 11). The source selection authority (SSA) characterizes the dual award as stimulating "continued competitive pricing and technology upgrade." Protest File, Exhibit 280 at 13. 4. With specified deviations, the acquisition is conducted under Air Force Regulation (AFR) 70-15, Formal Source Selection for Major Acquisitions. The solicitation establishes three areas of evaluation: management, technical, and cost/price. Management and technical areas are of equal importance, and each of greater importance than cost/price. However, "Although Cost/Price is listed last, it will be a significant criteria for award as part of an integrated assessment with the management and technical areas. The Government reserves the right to award a contract(s) at other than the lowest proposed price after consideration of all factors." Protest File, Exhibit 16 at 213-14 ( M-4). 5. A part of the cost/price analysis utilizes a cost evaluation model provided to offerors with the solicitation. The model calculates a most probable total contract cost (MPTCC) over the contract life utilizing prices proposed and specified estimated quantities. Protest File, Exhibit 16 at 224 ( M- 7.c(3)), 225 ( M-7.c(3)(d)), 161-90 (Tables L-1 through L-6). Neither the cost evaluation model nor the solicitation indicates a methodology to evaluate the MPTCC to the Government for a pair of contracts. 6. The solicitation contains the DFARS clause, 252.225- 7007 Trade Agreements Act (Dec. 1991). Protest File, Exhibit 16 at 99-102 ( I-21). This clause provides that offers "will be evaluated in accordance with the policies and procedures of Part 225 of the DFARS except that Offers of U.S. made end products shall be evaluated without the restrictions of the Buy American Act or the Balance of Payments Program." Id. at 102. The solicitation also contains the DFARS clause, Buy American Act- Trade Agreements Act-Balance of Payments Program Certificate (Dec. 1991). Id. at 124-25 ( K-16). 7. An express element of the proposal and evaluation relates to production and delivery capacity. A specific section of the management proposal is to "[d]escribe the current capability to produce and deliver the ordered systems and components in a timely manner . . . . Describe the current production and delivery of microcomputer systems." Protest File, Exhibit 16 at 144-45 ( L-21(1)(iii)); CompuAdd Corp., 1993 BPD 48, at 13 ( 30), 17-20 ( 43-51). 8. The solicitation notifies offerors that "a survey team may contact the Offeror's facility to confirm the Offeror's capability to perform. Current financial statements and other pertinent data should be available for Government review at that time if not already on file with the office having cognizance over the Offeror's facility." Production capability is identified as an area that may be investigated or evaluated. Protest File, Exhibit 16 at 136 ( L-18). 9. Subsequent to the Board determination in CompuAdd Corp., the agency requested offerors to extend, by January 15, best and final offers which had expired. Zenith, for example, extended its proposals on January 13. Protest File, Exhibits 276, 277. 10. In its best and final offer, Zenith certified that its systems, including monitors, "qualify as U.S. Made End Products." Protest File, Exhibit 248 at 217-18. In the proposal on which this protested award is based, Zenith obligated itself to deliver 20,000 microcomputer systems in month one of the contract and 25,000 systems per month thereafter. Id., Exhibit 246 at 22, 79. 11. A specific strength of this selected Zenith proposal was the minimum delivery commitment of at least 25,000 systems per month. Protest File, Exhibit 250. The proposal received the highest possible rating for production and delivery--a blue. Id. at 17. This evaluation paralleled that previously done and reviewed regarding the alternate Zenith proposal at issue in the prior Desktop protest. CompuAdd Corp., 1993 BPD 48, at 41-42. 12. The contracting officer accepted at face value the certification contained in the best and final offer extended by Zenith, and did not further probe Zenith's ability to satisfy its proposed delivery schedule. This action was taken knowing that the monitors here proposed involved similar make and model numbers and proposed delivery schedules to those reviewed by the Board in CompuAdd Corp. Protest File, Exhibits 30 at 109, 248 at 132. 13. In evaluating offers, the agency was to keep in mind the evaluation criteria, and the distinctions among three of the principal configurations to be offered: A. The basic system, to be used primarily to prepare correspondence using the offered office automation (O/A) software package. B. The intermediate system, in addition to functions and capabilities of the basic system, to be used to develop more complex documents and perform more advanced functions including building documents incorporating graphics and text from other application packages, building presentations with a graphics package, building moderately large spreadsheets that may involve complex mathematical operations and creating/maintaining larger and more complex data bases. C. The advanced system, in addition to functions and capabilities of the intermediate system, to be used to create and manipulate large data bases, run simulations, recalculate large spreadsheets, create and manipulate advanced graphics, use statistical packages and run functional programs. Protest File, Exhibits 16 at 218 ( M-7(b)(1)(a)), 279 at 5. 14. The proposals revealed variety by CLIN (and sub-CLIN) in terms of technology and price. For example, offers for the basic and intermediate systems (CLINs 1 and 2) utilized configurations with 386SX, 386DX, and 486SX, as well as other, technology. Protest File, Exhibit 279 at 46-58. For the advanced system (CLIN 3) proposed systems were configured with 386DX, 486SX, and 486DX, as well as other, technology. Id. 15. The SSAC reviewed and evaluated the best and final offers. It deemed five proposals to represent the "best value set from which a single best value may be determined." Protest File, Exhibit 279 at 15-26. It also reviewed pairs of these proposals along with others, as it conducted a dual award analysis. Id. 26-35. In conducting this paired analysis, the SSAC projected ordering patterns under paired proposals considering the management features, technology and prices offered--as stated by the SSAC, the SSAC "applied their combined judgment to determine if the combinations provided the users with a choice of management and technical solutions with appropriate price tradeoffs." Id. at 29. 16. The source selection decision compares offerings of 486-based systems for each of the first two CLINs (the basic and intermediate systems, Finding 13) with those offering 386-based systems. Protest File, Exhibit 280 at 3-9; Exhibit 279 at 10-11, 15-26 (SSAC Analysis Report). For example, In the technical area, the CLIN 1 486SX/20 based hardware system offered is far superior in terms of speed, capability, and expansion than the 386SX/20 based system offered. The 486SX offered will run the offered software more efficiently and effectively. Additionally, offerors are required to provide newer versions of software as they are released during the first three years of the contract. The more powerful 486 based platform is better suited to run innovations and upgrades in software that are likely to become available in the foreseeable future. Protest File, Exhibit 280 at 5 (references to particular proposals omitted). The decision does not indicate how the 486- based system's superior speed, capability, and expansion, will affect the preparation of correspondence--the stated use of the basic system, Finding 13. 17. The analysis of the SSAC and the SSA did not consider all pairs of acceptable proposals. Rather, generally, proposals were eliminated based on scoring and costs, such that the agency eliminated from further consideration those proposals with a total price (based upon the MPTCC, Finding 5) greater than a proposal in a more desirable color group (i.e., a better overall rating in terms of management and technical). Protest File, Exhibit 279 at 18, 26-28, Exhibit 280 at 2, 9-10. The agency did not consider that certain ordering patterns may have produced viable, beneficial pairings with these eliminated proposals. 18. The SSA concluded that a proposal offering a variety of 486-based systems for CLINs 1, 2, and 3 was the single proposal representing the best value to the Department of Defense. The SSA concluded that the proposal "clearly offered a superior Technical and Management solution at a moderate cost." Protest File, Exhibit 280 at 8-9. The analysis does not address how, if at all, the added costs associated with the various line items for the 486-based systems may preclude certain users from purchasing under the contract. 19. The SSA concluded that the best single value proposal in combination with another proposal (offering a 386-based technology for CLINs 1 and 2, and 486-based technology for CLIN 3) was the dual award pair representing the best value for the Department of Defense. Protest File, Exhibit 280 at 13. The range of processors from the low to high end of 386 and 486 processors was an important factor in the determination. Id. at 10-13. 20. The SSA concluded that the dual award, which would reduce certain risks, "stimulate continued competitive pricing and technology upgrade," and offer "users a choice in technology at attractive pricing to satisfy individual requirements," represented the best value to the Department of Defense. Protest File, Exhibit 280 at 13-15. On February 2, 1993, the agency awarded contracts to Zenith and Government Technology Services, Inc. (GTSI).[foot #] 22 21. After award, on March 12, 1993, Zenith issued a purchase order to its supplier for incremental delivery of 20,000 monitors to be completed by July 5, 1993. Transcript at 619. The record demonstrates that Zenith could not deliver systems with the proposed monitors in accordance with the Zenith-proposed delivery schedule which was, for Zenith, a beneficial factor in the evaluation and selection processes, Finding 11. Based upon the record developed, at the time Zenith extended its best and final offer, it lacked a reasonable basis to represent that it could comply with its proposed delivery schedule of Trade Agreement Act compliant monitors. Discussion Zenith monitor The protesters contend that the Zenith monitor continued to be unacceptable and was not eligible for award. The agency and Zenith assert that the proposed monitors would comply with Zenith's Trade Agreement Act certification and that Zenith could deliver compliant monitors without altering its proposal. In the previous round of protests, the Board stated: The real issue underlying this count of the protest, then, is not that Zenith's proposal ----------- FOOTNOTE BEGINS --------- [foot #] 22 A matter overlooked or downplayed by the agency: The winning Zenith proposal received a yellow score, Protest File, Exhibit 16 at 214 ( M-4a.(4)) ("Fails to meet evaluation standards; or has low probability of satisfying the requirement; or has significant deficiencies but correctable"), for commerciality, the fourth of five items under management, for which an overall score of blue was obtained. Protest File, Exhibit 279 at 17. ----------- FOOTNOTE ENDS ----------- contains intentional misrepresentations, but that it did not inform the Government of the fact that the monitors were not presently being produced in the United States, and would not be until after contract award, since this is a factor that could add considerably to the risk that Zenith might not, at least at the outset, be in a position to deliver as many units as the Government might order. . . . As such, we must conclude that knowledge of the full circumstances would have affected the risk analysis and resultant scoring of Zenith's winning proposal. CompuAdd Corp., 1993 BPD 48, at 41-42. Subsequent to that decision, Zenith extended its best and final offer leading to the award underlying these protests. In its proposal, Zenith commits itself to delivering monitors qualifying as United States made end products under the Trade Agreement Act,[foot #] 23 and to deliver 20,000 such monitors (as part of microcomputer systems) in month one of the contract, and 25,000 such monitors per month thereafter. The record demonstrates, however, that Zenith has no commitments which would enable it to obtain 20,000 compliant monitors until early July--that is, after five months of the contract have elapsed. Zenith benefitted in the evaluation and selection process by overstating its delivery capabilities. The agency's conclusions and selection were based upon incorrect information supplied by Zenith. The evaluations and selection determinations were tainted, and showed no regard for solicitation provisions noted in Findings 7 and 8. Zenith's winning proposal was not acceptable. The agency improperly made an award. Therefore, I would grant this basis of the protests. Zenith lacked a reasonable basis to continue to represent its ability to deliver monitors complying with its representation. In light of the Board's earlier decision, Zenith's simple extension of its best and final offer leading to the award here, amounted to a misrepresentation by Zenith of its delivery capabilities. For that reason, I would eliminate Zenith from participating in this procurement as a potential contractor. Regarding delivery, the majority recognizes that in best and final offers offerors committed to a given delivery capacity for each month of the contract, and the agency conducted its evaluations based upon those commitments. However, the majority ----------- FOOTNOTE BEGINS --------- [foot #] 23 For purposes of this discussion, I assume that monitors will be substantially transformed so as to comply with the Trade Agreements Act. ----------- FOOTNOTE ENDS ----------- does not interpret the delivery commitment as actually applying to each month of the contract. Rather, the majority construes the solicitation as establishing a delivery schedule for future months of the contract after protests have been resolved. Such a conclusion is not based upon any language in the solicitation or factors in the evaluation process. (The solicitation does not state that month one of the contract will begin in July 1993. The agency did not conduct its evaluations with that assumption.) Had the agency truly intended to begin requiring deliveries five months after award, it should have so notified all offerors before best and final offers were requested. The effect of the majority's interpretation is to permit an offeror to make commitments in its proposal based upon the assumption that a protest (or protests) or other causes will delay actual ordering under a resulting contract. Best value analysis The agency's approach to this procurement significantly deviates from the traditional indefinite delivery, indefinite quantity contract where a need for items is projected but the exact quantity or precise delivery dates are not known. The agency did not establish specific line items for different systems based upon technology (e.g., CLIN 1: 386SX; CLIN 2: 386DX; CLIN 3: 486SX; CLIN 4: 486DX); the solicitation does not suggest preferences for Intel as opposed to other technologies. Instead, it sought offers of commercial products satisfying largely functional specifications, without specifying particular interests of the potential users (whether they be of the agency, the rest of the Department of Defense, or other federal agencies) or dollar limitations of potential users in placing orders. The agency did specify, however, that it would make a best value determination with management and technical factors more important than cost/price, although cost/price would be a significant criterion. The agency's actions suggest that the solicitation did not accurately portray the requirements underlying the procurement. The analysis of the initial proposals for single best value should have alerted the agency to the fact that it lacked a common basis for soliciting and evaluating the offers. In its analysis of the single proposal representing the best value, the agency settled upon a solution which eliminated significant distinctions between principal line items representing the basic and intermediate systems. The agency had a true (as evaluated) preference, albeit unexpressed in the solicitation, for 486-based systems which would more readily enable a basic system to function as an intermediate system. This blurred the distinctions contained in the solicitation in terms of technology and estimated quantities for line item purchases. This error was exacerbated and made more apparent when the agency conducted its pair-wise analysis leading to the dual award. The projected ordering patterns utilized in the evaluation and selection processes for dual awards are not identified in the solicitation or otherwise substantiated as consistent or realistic. Again, the desire for 486-based systems drove the best value analysis. Viewing the Zenith proposal as the single best value distorts the actual solicitation. The proposal, priced higher than other acceptable solutions, provides a variety only of 486 microprocessor-based solutions for the basic, intermediate and advanced systems. The agency has failed to explain rationally how the 486 solution for the basic system, at a price premium over some 386 solutions, is (1) consistent with the three levels of systems represented as the solicited requirements, and (2) in the best interests of the Government. After receipt of initial proposals, if not earlier, the agency should have recognized the shortcomings of the solicitation and issued an amendment to focus the competition on true requirements. It was unreasonable and improper for the agency simply to proceed through award. 10 U.S.C. 2305(a)(1)(A) (1988); 48 CFR 15.606 (1992). The solicitation is for commercial items. Had the solicitation expressed true preferences and requirements, offerors could have competed on a common basis. Given the offers received, the agency's preferences, and the loosely drafted source selection criteria, the agency was not in a position to make a reasonable determination that a particular award or awards were in the best interests of the United States. Uncompeted technology and pricing The solicitation and the selection authority both anticipate permitting the contractor(s) to alter the technology and pricing available under the contract(s)--apart from simple upgrades to hardware and software. Findings 3, 20. There is an explicit recognition that the ultimate contracts against which orders will be placed need not resemble the proposals. In awarding the two contracts underlying these protests, the agency has created its own variety of schedule contracts which foster competition, but only between two awardees, each of which may substitute and add products without otherwise being subjected to competition from other potential offerors. The agency's approach is flawed in two principal respects. The contract changes in technology and prices will not be subjected to full and open competition; statute is therefore violated. 10 U.S.C. 2305. See Amdahl Corp. v. Department of Health & Human Services, GSBCA 11998-P, 1992 BPD 325, at 11 (Oct. 30, 1992), motion for reconsideration denied, 1992 BPD 416 (Dec. 23, 1993) ("In effect, what [the agency] would award under this solicitation, as it currently stands, is the privilege of negotiating the sale of computers to the agency, on a sole- source basis, over a period of nearly four years."). At the same time, the agency's approach makes illusory a "best value determination"--the agency is not in a position to determine what items may be added, and at what prices, and how those items may affect the ordering projected in the selection analysis. Relief Given the agency impropriety in accepting the winning Zenith proposal, the selection determination is tainted. Apart from that, however, the mechanism for the selection determination was fundamentally flawed because the solicitation failed to place the agency in a position to make a reasonable determination that any award would be consistent with the terms of the solicitation and in the best interests of the United States--competition for the underlying actual requirements was not attained. The solution here is not to permit continued action under this procurement which began in 1991 with an agency-anticipated award within six months without discussions. I would revoke the specific delegation of procurement authority. This would require the agency to determine how best to satisfy its actual requirements, and permit the General Services Administration to focus on the scope of an appropriate specific delegation of procurement authority it may choose to issue. 40 U.S.C. 759(b) (1988). Majority As is apparent, the majority takes a view of certain facts and law at times quite at odds with my view. Some matters demand specific comment. In Finding 14 of the majority opinion, it is recognized that the agency discarded pairings as unbalanced where "a significant majority of orders would flow to one vendor." However, that type of unbalanced ordering is not inherently contrary to the best interests of the United States. The agency's analysis fails to reconcile projected orderings with true underlying requirements (in terms of technology and price). The agency's rejection of such a pair is not reasonable or rationally based. In Finding 17 (including footnote 8) of its opinion, the majority recognizes that the SSAC--representing a wide variety of user communities in the Air Force--projected ordering patterns for dual awards. However, the Air Force is not the sole potential user under the contract; other elements of DoD have committed dollars to the minimum guaranteed purchases. Protest File, Exhibit 2. The interests of the users, the Air Force, DoD, and the United States, may not all coincide. The agency's projected ordering patterns are not necessarily representative. Regarding the Trade Agreement Act certification, I believe that the majority incorrectly concludes that the contracting officer was precluded from looking behind the certification included in Zenith's extended best and final offer. The solicitation permitted the agency to review pertinent data relating to an offeror's production capability. Finding 8. Although regulation does state that the "contracting officer shall rely on the offeror's certification as submitted," 48 CFR 25.407(b) (1992), regulation also provides: "No contract shall be entered into unless the contracting officer ensures that all requirements of law, executive orders, regulations, and all other applicable procedures, including clearances and approvals, have been met." 48 CFR 1.602-1(b) (1992). If the contracting officer knows or has reason to know that a certification is, or may be, inaccurate, it is unreasonable for the agency not to look behind the certification. Sysorex Information Systems, Inc., GSBCA 10642-P, et al., 90-3 BCA 23,181, at 116,372, 1990 BPD 193, at 22 ("where a representation made by an offeror is called into question, the agency should look behind that representation, and if it does not do so, we will, as part of our de novo review"), aff'd sub nom. SMS Data Products Group, Inc. v. Austin, 940 F.2d 1514 (Fed. Cir. 1991); Chesapeake & Potomac Telephone Co., GSBCA 10331-P, 90-2 BCA 22,883, at 114,927-28, 1989 BPD 385, at 19- 20. If the certification is not meant to be enforced as a term of the solicitation, then inclusion of the provisions misstates the agency's requirements. To the extent that the regulation pertaining to certification is inconsistent with these duties of the contracting officer, the regulation is overreaching and void. 40 U.S.C. 759(f)(1) (1988). Further, on page 26 of its opinion, the majority states: In CompuAdd Corp., 1993 BPD 48, at 46, the Board noted that the steps taken by the Air Force in identifying proposal AX as the "single best value" proposal were not in issue. The Board could not find that the Air Force had fully followed its stated award criteria, however, because the record did not reflect the same thorough consideration of the alternate possiblity provided for -- that of a dual award to a combination of proposals. The validity of the single award analysis was a specific issue in that round of protests. CompuAdd Corp., 1993 BPD 48, at 37 ("Moreover, protesters argue that the best value analysis as performed shows that the Air Force de facto changed its requirements after receipt of BAFOs without notifying offerors."). Having found that the single award to Zenith was improper, the Board did not reach the issue regarding the propriety of the underlying analysis and allegedly changed requirements, even though the hearing judge commented on such matters. CompuAdd Corp., 1993 BPD 48, at 46 n.22. Thus, while the majority recognizes that it decides the case with other than a blank slate, I part company with the majority regarding the scope of the Board's earlier decision. Finally, as does the majority, I recognize that none of these offerors protested the terms of the solicitation--all participated in the procurement and attempted to obtain an award under the given evaluation and selection language. However, I conclude that in applying the terms of evaluation and selection preferences to the offers actually received the agency was in violation of statute and regulation for the reasons stated above. Furthermore, I do not believe that it was apparent on the face of the solicitation's language that this necessarily would have occurred. Rather, the violations are predicated on the agency actions taken in light of offers actually received. _________________________ JOSEPH A. VERGILIO Board Judge