Matter of: KCA Corporation File: B-255115 Date: February 9, 1994
B-255115: Feb 9, 1994
Contracting agency's mechanical application of an undisclosed government estimate of minimum staff-hours to evaluate offers was unreasonable where the approach: (1) failed to assess whether a firm's proposed work force was particularly skilled or efficient. (3) awarded the same score for proposals that were both above and below the government estimate. Agency's determination that there was no reason to disqualify awardee was reasonable. KCA argues that the Army's evaluation of proposals was irrational and failed to assess the relative strengths and merits of each offeror's staffing proposal. Was biased in favor of Mitchco because several of Mitchco's key employees previously served in the Army.
Matter of: KCA Corporation File: B-255115 Date: February 9, 1994
1. Contracting agency's mechanical application of an undisclosed government estimate of minimum staff-hours to evaluate offers was unreasonable where the approach: (1) failed to assess whether a firm's proposed work force was particularly skilled or efficient, (2) failed to assess whether because of a unique approach the firm could satisfactorily perform the work with different staffing than estimated by the agency, (3) awarded the same score for proposals that were both above and below the government estimate, and (4) overemphasized deviations in less important functions at the expense of more important requirements. 2. Absent any evidence that awardee's personnel, who had previously worked for the government, had access to information not available to all offerors, agency's determination that there was no reason to disqualify awardee was reasonable.
KCA Corporation protests the award of a contract to Mitchco, Inc. under request for proposals (RFP) No. DABT23-93-R-0024, issued by the Department of the Army for food services at Fort Knox, Kentucky. KCA argues that the Army's evaluation of proposals was irrational and failed to assess the relative strengths and merits of each offeror's staffing proposal. KCA also contends that the Army failed to hold meaningful discussions; failed to recognize that Mitchco's project manager had a conflict of interest; and was biased in favor of Mitchco because several of Mitchco's key employees previously served in the Army, or were employed by Fort Knox.
We sustain the protest in part, and deny it in part.
On July 27, 1993, the Army issued the solicitation for a firm, fixed-price contract for food services at 35 Fort Knox buildings for a I-year base period, with four I-year option periods. The solicitation required prices for dining facility attendant services--primarily janitorial-type services--at 11 buildings; management and food production services--primarily food preparation--at 10 buildings; and full food services--combining janitorial and food preparation services--at 14 buildings.
The RFP provided for award to the responsible offeror whose offer was most advantageous to the government, price and other factors considered. Paragraph M.2 of the RFP stated, in relevant part:
"Offerors are advised that primary consideration will be given to the evaluation of technical proposals rather than price. However, should technical competence between offers be considered approximately the same, price takes on an increased importance as a selection determinant."
The solicitation identified two evaluation factors: comprehension of the RFP requirements, and general management. The comprehension factor included subfactors for staffing/methodology and offeror-furnished items; the general management factor included subfactors for key personnel, quality control, internal control, management decision authority, government property administration, other personnel, management approach, corporate support, and personnel administration.
Offerors were advised that the staffing/methodology subfactor was of substantially greater importance than all the other technical subfactors combined. Specifically, amendment No. 2 to the RFP warned offerors that the staffing/methodology subfactor
"is of such critical importance that a rating of unacceptable will render the entire proposal unacceptable. The other subfactors are of nearly equal importance but the subfactor of [offeror [f)urnished [i]tems is of somewhat greater importance.
In fact, the staffing/methodology subfactor represented 85 percent of all available points.
With respect to evaluating an offeror's comprehension of the critical staffing/methodology subfactor, paragraph L.21 of the RFP indicated that the agency would review each offeror's proposed staffing levels and methods. Thus, offerors were requested to submit individual staffing charts for each dining facility and separate operational use. Offerors were also encouraged to submit "[i]nnovative or unique approaches, if appropriate, toward meeting solicitation requirements." Offerors were advised, however, that such approaches should be explained in their proposals.
The source selection plan (SSP) anticipated the award of a numerical score to each offeror's proposed staffing for each building, and each different operational use of the building. Thus, the different operational uses of the 35 buildings were to be awarded 59 separate scores. The SSP stated that scores should be awarded after comparing the offeror's proposed staffing level (for each operational use) to the independent government estimate (IGE). Evaluators were then required to calculate the percentage of deviation between the proposed staffing level and the IGE. According to the SSP, proposals offering staffing levels within 5 percent of the IGE were to receive numerical scores in the "good" range--i.e., between 56 and 75 points. Proposals that matched the IGE were to receive the highest score, 75; proposals offering staffing levels within I to 5 percent of tee IGE--either above or below--were to receive fewer points down to a score of 56. The SSP provided that a higher score could be given to a proposal that deviated more than 5 percent above or below the IGE, if the proposal provided a fully acceptable rationale for the deviation, and the evaluator concluded that the approach offered a benefit to the installation.
Five proposals were received on August 26, and were submitted to the evaluators for initial scoring. When the evaluators provided their scores to the chairman of the Source Selection Evaluation Board (SSEB), he determined that the evaluators had not scored the staffing/methodology subfactor in accordance with the 55P. The SSEB chairman reached this conclusion when he learned that the evaluators had scored the proposals subjectively, and without reference to the IGE. As a remedy, the SSEB chairman created a table--which was added to the 55P as a new page--to direct rescoring of the proposals under the staffing/methodology subfactor. This table showed the percentage deviation between a proposal's staffing level and the IGE, and the corresponding score to be awarded. A portion of the table, showing the scores for the first I0 percentage points of deviation from the IGE, is set forth below:
0.00 percent 75 1.00 percent 71 2.00 percent 67 3.00 percent 63 4.00 percent 59 5.00 percent 56 6.00 percent 55 7.00 percent 53 8.00 percent 51 9.00 percent 49 10.00 percent 47
After rescoring the staffing/methodology portion of the initial proposals using this table, the revised scores were substituted for the initial evaluation scores. Reviewing the results of the amended evaluation, the Army included all five offerors within the competitive range, provided a list of concerns to each offeror, and conducted oral discussions by telephone.
The Army received best and final offers (BAFO) on September 21. KCA and Mitchco proposed the following prices and hours, compared to the IGE:
KCA 494,629.5 hours $ 39.50 million Mitchco 493,862.5 hours $ 39.99 million IGE 472,807.0 hours
The BAFOs were submitted to the evaluators for rescoring in all areas except the staffing/methodology subfactor. In this area, staffing changes were reevaluated by the SSEB chairman using the table discussed above.
After BAFOs were reevaluated, Mitchco received the highest rating--"good"--based on its total score of 6,138.75 points, 5,015 of which were awarded under the staffing/methodology subfactor. KCA received the second-highest rating--"fair"--based on its total score of 5,291 points, 3,952.5 of which were awarded under the staffing/methodology subfactor. In view of the relatively slight price difference between Mitchco and KCA, and Mitchco's significantly higher score, the Army awarded the contract to Mitchco on September 24, and this protest followed.
In its initial protest, KCA argued primarily that the Army was biased in favor of Mitchco because several of Mitchco's key employees previously served in the Army, or were employed by Fort Knox. In addition, the agency's rescoring of initial proposals aroused KCA's suspicion. KCA now argues that the Army's evaluation of proposals was irrational and failed to assess the relative strengths and merits of each offeror's staffing proposal. In this regard, KCA contends that the Army's evaluation approach fails to consider the added value of an offer that proposes to use more staff-hours than the IGE by awarding the same score regardless of whether an offeror exceeds or falls short of the IGE. In addition, KCA claims that the Army failed to hold meaningful discussions, and failed to recognize that Mitchco's project manager had a conflict of interest.
After reviewing KCA's pleadings, the agency's initial report and subsequent filings, the evaluation documents, and the offerors' proposals, and after holding a hearing at which each of the evaluators testified regarding how the proposals were scored, we conclude that there was no bias in favor of Mitchco. For the reasons set forth below, however, we conclude that the scoring method used failed to provide any meaningful assessment of which offeror's staffing levels provided the greatest advantage to the Army.
The Evaluation Methodology
The Army's determination that Mitchco submitted the superior proposal here was based almost entirely upon Mitchco's higher score under the staffing and methodology subfactor. As shown above, although KCA proposed to perform these services using a higher number of staff-hours than Mitchco (494,629 versus 493,862) at a lower fixed price ($39.5 million versus $39.99 million), Mitchco received a higher score because Mitchco's proposed staff-hours were closer to the IGE than KCA's proposed staff-hours. After the Army completed its scoring, the source selection official determined that Mitchco's higher score justified its slightly higher price.
An illustrative example of the Army's evaluation methodology is set forth below. For building 297, a full food service facility, the Army developed an estimate of the staff-hours required for each position it believed was needed to perform these services. For weekday food service for building 297 these estimates were: first cook, 12.5 hours; second cook, 22 hours; rations clerk, 4 hours; server, 9 hours; building clerk, 4 hours; and dining facility attendant, 27.5 hours.
Since Mitchco proposed first cook staffing at 13 hours, a level 4 percent above the government estimate, it received a score of 59 points for this position, in accordance with the table shown above. Similarly, Mitchco received scores for each of the other positions based on the percentage deviation from the estimate. These position scores were then combined to yield a composite score of 51.7 points for weekday food service in building 297. These scores were added to Mitchco's other staffing scores--for example, its score for weekend service for building 297; Monday to Saturday staffing in the Reserve Officers' Training Corps (ROTC) buildings; Sunday staffing in the ROTC buildings; and Sunday to Saturday staffing in building 5917. For the 35 buildings, and total of 59 different operational levels, Mitchco received an average score of 59 points, which was multiplied by 85 to generate a staffing/methodology score of 5,015 points.
In our view, this evaluation scheme failed in several ways to meaningfully assess the proposals. First, the evaluations mechanically assumed that the IGE represented the ideal staffing, and points were deducted from offerors wherever they failed to meet the government estimate, regardless of whether they proposed too few hours or too many. As a result, an offeror who proposed a cushion of staffing above the minimum received a lower score than an offeror who merely matched the staffing estimate, with no regard to whether the higher staffing benefitted the government .
While an agency may rely on its own estimates of the manning levels necessary for satisfactory performance when negotiating a fixed-price contract, absolute reliance on estimates can have the effect of arbitrarily and unfairly penalizing an innovative or unusually efficient offeror. Teledyne Lewisburg et al., B-183704, Oct. 10, 1975, 75-2 CPD Para. 228. In this regard, it is inappropriate to determine the acceptability of proposals by the mechanical application of an undisclosed estimate. Allied Cleaning Servs., Inc., 69 Comp.Gen. 248 (1990), 90-I CPD 7 275 (mechanical application of a staff-hour estimate found unreasonable). See also The Jonathan Corp.; Metro Mach. Corp., B-25l698.3; B-25l698.4, May 17, 1993, 93-2 CPD 7 174, aff'd, Moon Eng'q Co., Inc.--Recon., B-25l698.6, Oct. 19, 1993, 93-2 CPD 7 233 (mechanical application of agency cost estimate found unreasonable). Such an approach fails to assess whether a firm's proposed work force is particularly skilled and efficient, or whether because of a unique approach, a firm could satisfactorily perform the work with different staffing than estimated by the agency. See Kinton, Inc., 67 Comp.Gen. 226 (1988), 88-I CPD 9[ 112.
In addition, our review shows that the Army's evaluation methodology overemphasized deviations in the less important functions at the expense of more important requirements. For example, an offeror who underestimated staffing in the critical cook area might receive a higher score than another offeror who overestimated staffing for servers, but by a larger amount. This could be the case even though the estimated requirement for cooks was 12 times the estimate for servers, and even though the second offeror proposed sufficient staffing overall. Further, the equal weighing of all proposed staffing plans created even more anomalies--for example, deviations from the IGE for a Sunday-only staffing plan in a small building were accorded the same weight as deviations from a Sunday to Saturday plan in a major building. The resulting score, therefore, bore no relation to how close the offeror came to meeting the Army's overall food service needs at Fort Knox.
The record here also shows that the Army made no attempt to account for the particular approaches of the individual offerors. Once the SSEB chairman developed the chart for scoring staffing/methodology, there was essentially no further communication with the evaluators, who had the necessary expertise to assess the innovative and unique approaches that the solicitation invited. The evaluators were not asked to review the proposals for innovative or unique approaches, were not consulted concerning the framing of discussion questions, and were not asked to reevaluate the staffing proposed in BAFOs. In short, the premium here was not on preparing and supporting an estimate, but on guessing the Army's estimate. Accordingly, we find the evaluation unreasonable and inconsistent with the RFP's request for innovative proposals and its stated intent to make award to the proposal offering the greatest value to the government.
Since we conclude that the evaluation failed to reasonably assess the relative merits of the proposals here, we need not review in detail KCA's claim that the Army failed to hold meaningful discussions. Discussions cannot be meaningful unless an offeror is advised, in some way, of the weaknesses, excesses, or deficiencies in its proposal that must be addressed in order for the offeror to be in line for award; an agency may not mislead an offeror, through framing discussion questions, into responding in a manner that does not address the agency's concerns. E.L. Hamm & Assocs., Inc., B-250932, Feb. 19, 1993, 93-1 CPD Para. 156. Here, because the evaluation methodology failed to rationally identify areas of relative weakness in the proposals, it was inevitable that the protester was misled into focusing on areas that were of lesser importance in terms of the impact upon its own evaluation. Thus, we find that the errors in the evaluation ultimately impacted the agency's ability to properly advise offerors of the weaknesses and deficiencies in their proposals.
Conflict of Interest
KCA alleges that Mitchco should be disqualified from consideration for award of this contract because the awardee's project manager was, in KCA's view, a procurement official for this contract, in violation of the procurement integrity provisions of the Office of Federal Procurement Policy (OFPP) Act, 41 U.S.C. 7 423 (1988 and Supp. Ill 1991). In addition, KCA claims that Mitchco's contract manager and office manager were former employees of Fort Knox who had access to the IGE, and that Mitchco's owner is a retired procurement official with ties to source selection officials and members of the SSEB.
As an initial matter, while we recognize that the evaluation methodology used here would make knowledge of the government estimate especially useful, there is no evidence that anyone at Mitchco had impermissible access to procurement sensitive information. In addition, to the extent that KCA began its protest by claiming that the agency was biased in favor of Mitchco, we have found no evidence of any such bias. In fact, while we find the evaluation scheme unreasonable, its application was so mechanical as to preclude any attempt to favor one offeror over another in the evaluation process.
With respect to Mitchco's project manager, there is no evidence that this individual violated any of the procurement integrity restrictions set forth in the OFPP Act. While Mitchco and the Army admit that Mitchco's project manager participated personally and substantially in the preparation of the statement of work used in the predecessor contract for these food services at Fort Knox, the record shows that this individual retired from the government afterwards, and there is no evidence in the record that he had any involvement with the instant procurement. In addition, the record shows that the statement of work has changed since the earlier contract, and SSEB members testified that they prepared the present statement of work in the Spring of 1993, after the departure of Mitchco's project manager.
Based on the record before us, we also conclude that none of the other individuals mentioned by Mitchco had access to the IGE, or other procurement sensitive information. For example, while Mitchco's office manager was the Army secretary who typed portions of the solicitation and statement of work, this material was later provided to all offerors. There has been no showing that this individual participated in any meaningful way in the development or drafting of these materials, or that she typed the government estimate. Finally, the owner of Mitchco retired from the Army 10 years ago, and, the agency advises, was never a procurement official. Thus, we conclude that the Army reasonably determined that there is no basis for disqualifying Mitchco from the competition.
CONCLUSION AND RECOMMENDATION
For the reasons stated above, we find that the Army's evaluation here failed to provide a meaningful assessment of the relative strengths and weaknesses of the competing proposals. Our review of the evaluation, together with the RFP's request for unique and innovative proposals and its stated preference for the offer most advantageous to the government, leads us to recommend that the Army revise its evaluation scheme and reevaluate the proposals, holding discussions and obtaining revised proposals if necessary. The agency should then make award to the offeror whose proposal best meets the requirements set forth in the RFP. We also find that the protester is entitled to recover its costs of filing and pursuing these protests, including reasonable attorneys' fees. 4 C.F.R. Sec. 21.6(d) (1993). KCA should submit its detailed and certified claim for such costs to the agency within 60 days of receipt of this decision.
The protest is sustained.
1. In some cases the RFP required separate operational plans for the same building where the building has different operational needs on weekdays than it does on weekends or holidays. Other buildings simply required operation 7 days per week.
2. The table, at page 34a of the 55P, ends with a percentage deviation of 51.00 percent, which receives a score of 0.
3. "First cooks" have overall supervisory authority.
4. During the hearing held in connection with this protest the evaluators who prepared the government staffing estimate testified that the estimate was a minimum acceptable level of staffing. As KCA points out, the agency then unreasonably evaluated proposals as if excess staffing to meet contingencies would jeopardize a contractor's ability to accomplish the statement of work.
5. The protester also alleged that a source selection official had a close family member who worked for a firm affiliated with or controlled by Mitchco. The contracting officer, who was the source selection official, has stated that she has no family members who work for such a firm, and the protester has provided no more details regarding the allegation.