Matter of: Labat-Anderson, Inc. File: B-287081; B-287081.2; B-287081.3 Date: April 16, 2001
B-287081,B-287081.2,B-287081.3: Apr 16, 2001
DIGEST Protest that contracting agency improperly eliminated proposal from consideration for award because it failed to comply with solicitation requirement to clearly explain the rationale for including hard-coded entries in electronic version of its priced estimating model is denied where the record shows the agency reasonably found the protester's explanation for these entries insufficient. Aliens seeking benefits under immigration and nationality laws are required to mail their applications and petitions to one of four service centers located across the United States. The initial task order was to be issued on a fixed-price basis. Follow-on task orders were to be issued on fixed-price or time-and-materials bases.
Matter of: Labat-Anderson, Inc. File: B-287081; B-287081.2; B-287081.3 Date: April 16, 2001
Labat-Anderson, Inc. protests the award of a blanket purchase agreement (BPA) to JHM Research and Development, Inc. under request for quotations (RFQ) No. HQ-0-Q-0044, issued by the Immigration and Naturalization Service (INS) to obtain service center operations support services for its direct mail and records management program. Labat contends that INS improperly evaluated its price proposal and eliminated its proposal from consideration based upon the results of that evaluation; improperly evaluated its technical proposal; failed to conduct meaningful discussions with the firm; and performed a flawed best value analysis leading to the selection of JHM's proposal for award.
We deny the protests.
Under INS's direct mail and records management program, aliens seeking benefits under immigration and nationality laws are required to mail their applications and petitions to one of four service centers located across the United States. RFQ Statement of Work (SOW) Para. 2. The services required to support the program include the following task areas: mailroom operations; file assembly; data collection, capture, and scanning; document preparation; fee collection and processing; fileroom operations; word processing; quality control; project management; business process reengineering; and other records management functions. SOW Para. 3. INS currently processes approximately 5 million applications and petitions annually. Id.
Pursuant to this RFQ, issued June 26, 2000, INS planned to award a single BPA to a contractor holding a Federal Supply Schedule (FSS) contract under the General Services Administration's Document Management Services and Products Schedule, Special Item Number (SIN) 51-504, "Records Management Services," and other applicable SINs. INS estimated that the volume of purchases through this agreement would be $344 million over 60 months. RFQ Para. 2.3. The initial task order was to be issued on a fixed-price basis, and follow-on task orders were to be issued on fixed-price or time-and-materials bases. RFQ Para. 1.1.
Offerors were required to submit separate technical and price proposals for evaluation by a technical evaluation committee (TEC) and a business evaluation committee (BEC), respectively. Proposals were to clearly demonstrate the offeror's understanding of the overall and specific requirements of the proposed BPA, and any proposal in which material information requested was not furnished or where indirect or incomplete answers or information were provided might be considered unacceptable. FFP Para. 3.4. Offerors were required to submit both hard and electronic copies of their proposals. RFP Para. 3.4.9.
Award was to be made to the offeror whose proposal represented the best value to the government based on three evaluation factors: technical approach, past performance, and price. The technical approach and past performance factors were equally weighted, and the two combined were significantly more important than price. RFQ Para. 3.3. Of the two equally important technical approach subfactors, estimating model and management, only the former is at issue here.
Offerors were required to submit an estimating model that stated the underlying assumptions and constraints of their method for accomplishing the work under orders issued against the BPA, state the statistical basis for their estimates, and explain how they derived their technical approach from the analysis of the SOW and multiple years of workload data supplied by INS. Offerors were required to explain, illustrate, and explicate their model showing how the workload associated with each of the many forms processed under the program built to a model for the technical approach to operating each service center and the direct mail program as a whole; failure to do so was to be taken as a lack of understanding of the technical requirement and a deficient technical approach. RFQ Para. 188.8.131.52. The RFQ listed specific criteria under which INS planned to evaluate this estimating model to assess the contractor's ability to provide the required services. RFQ Para. 4.2.1. Under one criterion, offerors were required to show how using the INS estimates of forecasted demand and their own allocation and employment of labor assets using available data would result in an appropriately-sized workforce. Under another criterion, offerors were required to explain their methods and rationale for allocating labor among the functional areas (such as filing, mailroom operations, and data entry), given the forecasted demand. Id. INS's "estimates of forecasted demand" were based upon the 1999 historical workload data. Amendment No. 1 at Question and Answer (Q&A) Nos. 22, 29; RFQ Pricing Tables; RFQ attach. 9, reports C, D.
Offerors were required to submit price proposals that included three sections relevant to this protest. First, offerors were to provide an explanation of pricing that described all assumptions made and constraints affecting price, as well as the offeror's price proposal methodology. Second, offerors were to complete electronic pricing tables in accordance with specific instructions. These pricing tables, which were to serve as the mechanism by which INS ordered services under the BPA, were built upon INS's estimates of forecasted demand. Third, offerors were to provide a priced version of the estimating model developed in their technical proposals. The models submitted for both the technical and price proposals were required to be the same except that the model submitted for the technical proposal could not contain any labor rates or prices. Offerors were required to "electronically link" the priced estimating model to the pricing tables in a manner that would allow INS to see how their proposed price was determined. RFQ Para. 184.108.40.206.
Price was to be evaluated for price reasonableness and cost realism, as well as total evaluated price. INS was to conduct its price analysis using one or more of the techniques specified in Federal Acquisition Regulation (FAR) Sec. 15.404-1(b), and its cost realism analysis in accordance with FAR Sec. 15.404-1(d). INS might also reject any proposal that was unreasonable or materially unbalanced as to prices for basic and follow-on year quantities. The RFQ defined an unbalanced proposal as one that incorporated prices significantly less than cost for some items and/or prices that are significantly overstated for other items. RFQ Para. 4.2.3.
INS received proposals from four offerors by the August 7 closing date, including those from JHM and Labat, the incumbent. One offeror was removed from consideration for reasons not relevant here, and the BEC and TEC conducted clarifications and completed their evaluations of initial proposals on August 29. The source selection advisory council (SSAC) completed its review of the BEC and TEC reports and concluded that discussions were necessary. INS established a competitive range comprised of the three remaining proposals and proceeded with discussions. Offerors submitted final proposal revisions (FPR) on October 3. The BEC and TEC evaluated FPRs and eliminated the third offeror's proposal from the competitive range for reasons not relevant here. On October 19, the SSAC concluded that additional information would be needed from Labat and JHM before an award recommendation could be made.
On October 26, INS issued further clarifications and instructions to JHM and Labat along with amendment No. 6, and asked for FPRs by November 2. As discussed below, amendment No. 6 added language to RFQ Para. 4.2.3. requiring the electronic version of the priced estimating model to contain all formulae and links between spreadsheets so the BEC could analyze the electronic version of the model to determine how prices were derived. If the electronic model contained "hard-coded" entries -- those which lacked formulae or links to another spreadsheet -- the offeror was required to clearly explain the rationale for their inclusion. If an offeror failed to do so, the BEC would find the proposal non-compliant with the RFQ instructions and would recommend that it be eliminated from further consideration.
The BEC found Labat's FPR deficient because it failed to provide a rationale for several hard-coded entries in its priced estimating model as required, and it was unable to determine that Labat's total evaluated price was reasonable and realistic. The BEC recommended that Labat's proposal be eliminated from further consideration as a result, and that award be made to JHM, whose proposal was found to be reasonable and realistic. The TEC found Labat's technical approach deficient with respect to the estimating model subfactor because its proposal did not meet the RFQ's minimum requirements as a result of its use of incorrect workload quantities in several task areas and as a result of various ambiguities and omissions.
The BEC and TEC provided the SSAC with the following evaluation results:
Technical Acceptable Marginal Approach
* Estimating (Acceptable) (Marginal) Model
* Management (Good) (Good)
Past Performance Good Good
Price $325,995,032 $317,347,959
The SSAC unanimously accepted these findings. Based on the BEC's conclusion that Labat failed to explain hard-coded entries in its priced estimating model, the SSAC recommended that Labat's proposal be excluded from the competition.
The SSAC nonetheless considered whether JHM's proposal should be selected despite the fact that its price was 2.8 percent higher than Labat's, and conducted a best value analysis. To do so, the SSAC combined the equally important technical and past performance ratings of each offeror and arrived at an overall rating of "Good" for JHM's proposal and "Marginal" for Labat's proposal. The SSAC also conducted an impact analysis in which it examined the significant differences, or discriminators, between the offerors based on a comparative analysis of their strengths, weaknesses, and risks. The SSAC characterized these discriminators as either features or risks, and determined that JHM's proposal presented seven features, with no risks, and Labat's proposal presented seven features and two risks, as noted above. The SSAC found that the proposals' features were similar and that the differences were attributable to Labat's incumbency.
The SSAC concluded that JHM's overall rating of "Good" meant that it presented a technically superior proposal at a negligible difference in price. The SSAC stated that even if it had rated JHM's proposal as "Acceptable" overall, the negligible difference in price would still have resulted in its being a technically superior proposal. While it contained fewer technical features than did Labat's proposal, it presented no major risks as compared with the two major risks in Labat's proposal. The SSAC recommended award to JHM. The SSA concurred and award was made to JHM on January 3, 2001.
Labat filed these protests after its debriefing. Labat contends that INS improperly evaluated its price proposal and eliminated its proposal from consideration based upon its failure to comply with RFQ Para. 4.2.3; improperly evaluated its technical proposal; failed to conduct meaningful discussions with the firm; and performed a flawed best value analysis leading to the selection of JHM's proposal for award.
As a preliminary matter, the RFQ stated that INS intended to issue a BPA against the vendor's GSA FSS contract. Accordingly, the provisions of FAR Subpart 8.4 apply here. Those provisions anticipate that agencies will review vendors' federal supply schedules and place an order directly with the schedule contractor that can provide the supplies or services that represent the best value and meet the government's needs. FAR Sec. 8.404(b)(2); Digital Sys. Group, Inc., B-286931, B-286931.2, Mar. 7, 2001, 2001 CPD Para. __ at 6. Where, as here, the agency intends to use the vendors' responses as the basis of a detailed technical evaluation and price/technical tradeoff, it may elect, as INS did here, to use an approach that is like a competition in a negotiated procurement. Where an agency takes such an approach, and a protest is filed, we will review the agency's actions to ensure that the evaluation was reasonable and consistent with the terms of the solicitation. /1/ COMARK Fed. Sys.,
B-278343, B-278343.2, Jan. 20, 1998, 98-1 CPD Para. 34 at 4-5. This RFQ specifically stated that the source evaluation was to be conducted and selection made in accordance with the guidelines set forth in the FAR, and set forth specific procedures for the evaluation of proposals. Accordingly, while the provisions of FAR Part 15, which govern contracting by negotiation, do not directly apply, Computer Prods., Inc., B-284702, May 24, 2000, 2000 CPD Para. 95 at 4, we analyze Labat's contentions by the standards applied to negotiated procurements. Digital Sys. Group, Inc., supra.
The estimating model is at the center of these protests. In addition to serving as a tool to help INS assess an offeror's ability to provide the required services, the model was to serve as a management tool during performance. As Labat acknowledged in its proposal, "[DELETED]." Labat Technical Proposal at 43. INS also intended to use the model as a pricing tool. If GSA modified the contractor's schedule rates, INS planned to use the offeror's proposed discounts to adjust the rates in the model and calculate the adjusted pricing in the pricing tables. Amendment No. 1 at Q&A No. 36.
The principal issue for our review is whether INS improperly eliminated Labat's proposal from consideration for award based on its finding that it did not comply with the requirements of RFQ Para. 4.2.3 by failing to provide a clear rationale for hard-coded entries in its priced estimating model.
Along with a narrative explanation of its pricing methodology, the RFQ required each offeror to submit a set of interrelated items in its price proposal. The first item was a completed set of pricing tables. The pricing tables, which were to serve as the mechanism by which INS ordered services, were given to offerors in the form of electronic spreadsheets in Microsoft Excel format. INS inserted its estimates of forecasted demand in the tables and offerors were to complete the tables by inserting their pricing information. The second item was a priced version of the estimating model developed for the technical proposal. The electronic version of the model was to include spreadsheets in Microsoft Excel format. Offerors were required to "electronically link" their priced estimating models to the pricing tables in a manner that would allow INS to see how they determined their proposed prices. RFQ Para. 220.127.116.11.
The primary document used to store and manipulate data in Excel is a worksheet, which consists of cells organized into columns and rows. A worksheet is always part of a workbook, and several worksheets can appear in one workbook. Julia Kelly, Using Microsoft Excel 97 at 570 (3rd ed. 1998). An "electronic link" is a data connection between a dependent worksheet (the worksheet that uses the data) and a source worksheet (the worksheet in which the original data resides). The dependent worksheet is updated whenever the data changes in the source worksheet. Id. at 565. Since the electronic versions of the pricing tables and priced estimating models were in Excel format, "electronically linking" the pricing tables to the priced estimating model would enable INS to trace each proposed price in the pricing tables to the price components found in the priced estimating model.
Amendment No. 6 added the following language to RFQ Para. 4.2.3:
[Offerors] are reminded that the electronic version of the estimating model must contain all formulae and links between spreadsheets. The BEC will analyze the electronic version of the estimating model to determine how the prices were derived. If a particular row or column within a spreadsheet or a particular spreadsheet within a workbook contains hard coded entries (i.e., lacks formulae or links to another spreadsheet), the Offeror must clearly explain the rationale for the inclusion of such hard coded entries. If an Offeror fails to provide an electronic version of the proposal that contains all of the formulae and links between the spreadsheets and tables presented in its proposal, or provides an estimating model that contains hard coded entries without any rationale for or explanation of the hard coded entries, the BEC will determine the Offeror's proposal to be non-compliant with the [RFQ] instructions and recommend to the Contracting Officer that the Offeror's proposal be eliminated from further consideration in the evaluation.
As this language suggests, "hard-coded" entries are numbers or values that are manually keyed into an electronic spreadsheet, as opposed to derived from links to other cells in the electronic spreadsheet or based on a formula using data from other cells. Supplemental Protest, Feb. 5, 2001, at 2.
Simultaneous with the issuance of amendment No. 6, INS advised Labat that its pricing tables and their values could not be derived from its priced estimating model because links and formulae either had been hard-entered, were missing, or were obviously uncorrected and/or incorrect in their final form. As an example, INS stated that the administrative overhead allocation percentages for fixed-price tasks were traceable only to hard-entered numbers that were not linked by formulae in such a manner as to be derived from the backup data provided.
In its FPR, Labat stated that it had resubmitted the entire estimating model and any links or formulae had been inserted. In evaluating Labat's FPR, the BEC found that the firm did not provide sufficient rationale or support for the hard-coded entries it used to determine its proposed administrative overhead, as required by RFQ Para. 4.2.3. The BEC's concerns were broken down into four categories. The BEC stated that while, on the surface, Labat's proposed prices appeared to be reasonable and consistent with those in the independent government cost estimate, the firm's failure to explain its hard-coded entries prevented the BEC from determining that the proposal was reasonable and realistic. Referring to INS's plans to use the model as a pricing tool, the BEC concluded that Labat's model was not capable of generating a fair and reasonable price. Final BEC Report at 10.
Labat does not dispute that its priced estimating model contained these hard-coded entries, but contends that it provided a sufficient rationale for their inclusion. Labat alternatively contends that even if the rationale it provided was insufficient, the omissions are immaterial and the BEC should have been able to determine that its proposed price was reasonable notwithstanding the inclusion of these hard-coded entries.
Again, the RFQ stated that price proposals were to be evaluated for price reasonableness and cost realism, as well as total evaluated price. The analysis of a proposal for price reasonableness involves the evaluation of a proposed price without evaluating its separate cost elements and proposed profit. FAR Sec. 15.404-1(b)(1). Here, the BEC used an array of permitted price analysis techniques, including an analysis of offerors' pricing information, since the offerors' use of alternative work methods made a mere comparison of their pricing with each other or with the government estimate insufficient to determine price reasonableness. While the solicitation uses the term "cost realism," a cost realism analysis is not used where a solicitation contemplates the award of a fixed-price contract. We assume that the agency meant a "price realism" analysis, which may be used in a solicitation for the award of a fixed-price contract to assess the risk in an offeror's approach. In this regard, the RFQ advised that the BEC planned to assess the price of doing business with each offeror and that "cost realism" was to be used as an aid in determining the offeror's understanding of the solicitation's requirements. Proposals that were unrealistic in terms of technical commitment were to be deemed to show an inherent lack of technical competency or failure to comprehend the complexity and risk of the BPA requirements, which might be grounds for rejection of the proposal. RFQ Para. 4.2.3.
Where, as here, a BPA contemplating fixed-price task orders is to be awarded, the "realism" of offerors' proposed prices is not ordinarily considered, since a fixed-price contracting vehicle places the risk and responsibility for contract costs and ensuing profit or loss on the contractor. See Human Resources Sys., Inc.; Health Staffers, Inc., B-262254.3, et al., Dec. 21, 1995, 96-1 CPD Para. 35 at 5. However, an agency may, at its discretion, provide for the use of a price realism analysis in a solicitation for the award of a fixed-price contracting vehicle to assess the risk in an offeror's approach. See Hydraulics Int'l, Inc., B-284684, B-284684.2, May 24, 2000, 2000 CPD Para. 149 at 14. The nature and extent of an agency's price realism analysis are matters within the agency's discretion, Cardinal Scientific, Inc., B-270309, Feb. 12, 1996, 96-1 CPD Para. 70 at 4, and our review of an agency's price evaluation is limited to determining whether it was reasonable, consistent with the solicitation's evaluation criteria, and compliant with the FAR. Hydraulics, Int'l, Inc., supra.
Our review of the record shows that INS reasonably concluded that Labat failed to "clearly explain the rationale" for its hard-coded entries as required by the solicitation, and that INS reasonably concluded that this failure prevented it from determining whether Labat's price proposal was reasonable and realistic.
The BEC's first concern was associated with Labat's proposed "administrative overhead" weighted labor rates. /2/ Taking advantage of the flexibility afforded by the RFQ, Labat chose to price all labor hours for breaks, computer reports, Freedom of Information Act request responses, quality control, reports and statistics, system downtime, and training meetings in its administrative overhead pool. For each service center, for each year of the contract, Labat included a worksheet in its model that calculated an administrative overhead weighted labor rate. One component of the rate was Labat's allocation, by percentage, of the total number of administrative overhead hours among the various indirect labor functions. For example, 10 percent of the hours might be allocated to breaks, 15 percent to file maintenance, 20 percent to quality control, and so on. All of these allocation percentages were hard-coded entries. The resulting rate was used elsewhere in the model to arrive at total prices for various task areas. The explanation provided for these hard-coded entries is in the form of a note that states, "[t]hese values are programmed into the model based on management knowledge of each site." See, e.g., Final Price Proposal Priced Evaluation Model, Base Year California Service Center, at 1, 3.
The BEC determined that this was not a clear explanation of Labat's rationale for including these hard-coded entries. The BEC's analysis showed that Labat's allocations of administrative overhead were inconsistent with the historical workload data, often by substantial margins. There was no explanation for these discrepancies, or for related discrepancies in the fileroom task area between the historic ratios of administrative overhead to direct labor and Labat's ratios. Final BEC Report, Tab D, at 5-13. In addition to undermining the reasonableness of the rate, these unexplained discrepancies raised questions regarding Labat's understanding of and approach to administrative overhead and its components. Consistent with its charge to evaluate proposals for realism, the BEC assessed not only Labat's allocations but its approach to allocating administrative overhead among the various components since several were work activities required by the SOW. Based upon the discrepancies identified above, and in the absence of any explanation for the discrepancies save the reference to "management knowledge," the BEC concluded that the firm proposed to significantly reduce the level of services for some components and increase the level of services for others. BEC Chair Statement at 8. The BEC also concluded that it could not use Labat's model as a post-award tool for determining price adjustments as envisioned. Final BEC Report at 10.
In light of the RFQ's requirements, the BEC reasonably concluded that Labat provided an insufficient explanation of these hard-coded entries. In addition to the general admonition that indirect or incomplete answers or information might be considered unacceptable, the RFQ specifically required the priced estimating model to contain a "clear explanation" of the rationale for the inclusion of hard-coded entries. RFQ Para. 4.2.3. This requirement must be read in the overall context of this paragraph, which informs offerors that the BEC planned to use the information in the price proposal to validate the model, evaluate for reasonableness and realism, and determine how prices were derived. Under the circumstances, Labat should have known that its mere reliance on "management knowledge" to explain these allocations would be insufficient. Labat's disagreement notwithstanding, a comparison of the historic allocations with those proposed by Labat reasonably raised the concern that Labat was proposing to change the level of support for certain components, and there was no explanation for the discrepancies to allay this concern.
In a related finding apart from this deficiency, the BEC concluded that Labat's proposal contained unbalanced pricing in the administrative overhead pool because it significantly understated hours for legally-required breaks and proposed significantly higher hours than historical for other items, and unbalanced pricing for the fileroom task area because the total administrative overhead pool proposed was not sufficient to cover legally-required breaks and other items. Whether or not these breaks were "legally required," the record confirms that the unexplained discrepancies in Labat's proposal reasonably raised this concern. Labat's argument that BEC should have combined the hours it allocated for system downtime along with those allocated for breaks is unpersuasive and unsupported by its proposal. /3/
The BEC's second concern was associated with Labat's allocation of administrative overhead to various task areas. For each service center, Labat provided a worksheet in which it allocated the percentage of total administrative overhead hours across various task areas. These allocation percentages were hard-coded entries with the same note of explanation referenced above, which the BEC found to be insufficient. Since the entries were carried out to more than 10 decimal places, the BEC also thought that Labat used formulae to calculate these allocation percentages but did not include or had deleted them. The underlying data shows that the allocations were different for each service center and varied widely across some task areas.
Labat contends that it was "entitled" to hard-code values for the allocation of administrative overhead among the task areas on the basis of its management judgment. We do not read INS's evaluation documents or pleadings to suggest otherwise but, under the RFQ's terms, offerors who chose to manifest their management judgment as hard-coded entries were required to clearly explain the rationale for such entries in a manner sufficient to allow the BEC to evaluate proposals for reasonableness and realism. Labat failed to meet this requirement. Labat also objects to the BEC's assumption that it omitted or deleted underlying formulae used to arrive at these allocations. Labat asserts that the allocations were "ballpark estimates" entered into the spreadsheets as round numbers, and that subsequent minor modifications to the spreadsheets caused the round percentages to stretch out to 10 decimal places. We do not dispute Labat's account, which was not included in its proposal, but believe the BEC's assumption was reasonable given the presence of these unexplained figures.
Labat's assertion that JHM also relied upon its management judgment to allocate indirect labor is misplaced. The record shows that JHM's reliance on its experience and judgment was supplemented by an analysis and review of the historical trends for all elements of indirect labor at all service centers. More important, JHM advised that it used the RFQ-provided data for [DELETED] to develop its prices, and the BEC did not identify any inconsistencies between that easily-identified set of data and JHM's pricing or approach. In contrast, Labat advised that it used the RFQ-provided data from a much longer span of time, but that it had made certain unexplained adjustments to that data. Since these adjustments were unexplained, the set of data used by Labat could not be easily identified, which prevented the BEC from understanding Labat's methodology. As a result, INS did not evaluate the proposals on a disparate basis as Labat contends, but properly evaluated them based upon their different approaches to meeting the requirements. /4/ See Engineered Air Sys., Inc.; Hunter Mfg. Co., B-283011 et al., Sept. 21, 1999, 99-2 CPD Para. 63 at 6-7.
The BEC's third concern was associated with Labat's allocation of data entry hours and data entry overhead between the data entry and fee collections task areas. The RFQ's historical data did not differentiate hours between fee processing and data entry task areas except for one function. The RFQ, however, required that the hours that had been included in data entry be divided between two new pricing categories, "fee collection and processing" and "document preparation." Labat's model included a worksheet for each service center in which it allocated both data entry hours and data entry overhead between these two task areas. These allocations were hard-coded entries, and differed from service center to service center. In addition to providing the same note of explanation as discussed above, Labat's proposal included a figure that set forth its method for allocating the fee processing hours for the California service center. Labat also explained that its service center managers developed estimates of per form labor hours for fee collection and then extracted those hours from the data collection task area, and that it performed a time-based study that analyzed fee-bearing forms to determine the time it takes to process the fee.
The BEC found that Labat did not clearly explain its rationale for the inclusion of these hard-coded entries. The BEC acknowledged that the figure referenced above provided the firm's method for allocating fee processing hours for the California service center, but noted that this data was not provided for any of the other three service centers. This, and the fact that Labat's allocations were carried out to multiple decimal places, raised the question whether Labat used an unstated methodology for making these allocations at the other service centers.
Labat asserts that its explanation above regarding the actions taken by its service center managers was sufficient to explain its hard-coded entries. As INS points out, however, Labat did not provide any information regarding the basis for the estimates it developed, or its time-based study, and the BEC could not validate the allocations or determine their realism. As for Labat's claim that it clearly meant for the data in the figure for the California service center to apply to all of the service centers, its proposal does not evidence this intention. The fact that the figure specifically referenced one service center and not just "service centers" undermines the claim.
The BEC's fourth concern was associated with Labat's overall failure to clearly explain its rationale for adjustments made to the historical data provided with the RFQ. The BEC understood that Labat used the RFQ-provided data for a multiple-year period of time as the basis for its model, but the firm stated that it entered this data, along with its assumptions and constraints, into the model, and that, "[a]fter consideration was given to efficiencies, facility, and many other planning scenarios for an efficient work process, the model was developed." Final Business Proposal at 8. The BEC concluded that Labat did not provide or identify how the assumptions, constraints, efficiencies, facility, and many other planning scenarios affected or changed the RFQ-provided historical data. We have reviewed the proposal citations provided by Labat in support of its argument that sufficient explanation was provided, but cannot find any further insight into Labat's adjustments to the historical data. This failure permeates the BEC's findings with respect to this deficiency, and Labat has given us no basis to find the BEC's conclusion unreasonable.
In sum, Labat has not shown that INS unreasonably found that its proposal failed to comply with the requirements of RFQ Para. 4.2.3, or that INS unreasonably found that it was unable to validate the model or ascertain that its pricing was reasonable and realistic. Labat has also not persuaded us that this deficiency as a whole was immaterial. The firm's arguments that shifts in its allocation percentages would have little effect on its rates, and that its allocations have not been shown to be unreasonable, do not address the fact that the BEC was concerned with both the reasonableness of Labat's pricing and its own inability to ascertain that Labat's understanding of and approach to administrative overhead and its components were realistic. In this regard, since what the TEC referred to as the "indirect requirements," such as those associated with administrative and support tasks, historically amount to about one-third of the total labor hours, see Final TEC Report at 64, Labat's failures here, when taken as a whole, have the potential to significantly affect pricing and performance. Under the express terms of the RFQ, the BEC had the discretion to recommend that Labat's proposal be eliminated from consideration for award under these circumstances. Labat has given us no reason to question the BEC's exercise of that discretion, and no reason to question the SSA's acceptance of that recommendation.
The language of the RFQ and at least one discussion item were sufficient to put Labat on notice of the requirements at issue here. Since the deficiency first arose in Labat's FPR, the agency was not required to reopen discussions or to allow it further opportunity to revise its proposal. See Earle Palmer Brown Cos., Inc., B-243544, B-243544.2, Aug. 7, 1991, 91-2 CPD Para. 134 at 11. If Labat believed that amendment No. 6 made a radical change to the requirements to which it had insufficient time to respond, a protest raising that issue, to be timely, had to be filed not later than the closing date for receipt of FPRs. 4 C.F.R. Sec. 21.2(a)(1) (2000); Intermagnetics Gen. Corp., B-255741.2, B-255741.3, May 10, 1994, 94-1 CPD Para. 302 at 6.
While we conclude that INS reasonably eliminated Labat's proposal from consideration, our review of the issues raised by the technical evaluation, discussed briefly below, shows that there is no basis to conclude that INS's evaluation of Labat's technical proposal was unreasonable and, hence, no basis to conclude that the source selection decision would have been affected had Labat's proposal been retained.
The TEC found that Labat's FPR introduced a new deficiency associated with the firm's failure to provide an estimating model that described the discrete labor in each major process for each form type and task area. Specifically, Labat's FPR failed to propose labor hours per form for two form types at one service center; failed to use the INS-provided workload estimates for the mailroom task area; failed to use the RFQ quantities for certain service center combinations; and failed to use the RFQ quantities for fee-bearing forms by service center at any service center. The TEC rated Labat's proposal marginal for the estimating model subfactor and marginal for the technical approach factor overall. /5/
In its January 10 written debriefing, INS advised Labat of extremely detailed reasons for each of the four areas underpinning this deficiency. Despite its knowledge of these specific criticisms, Labat's January 12 protest essentially consisted of a disagreement with the agency's conclusions regarding the technical evaluation and general objections, and Labat's January 22 protest specifically challenged only two of these areas. Given the extensive detail available to Labat at the time it filed its protests, its broad allegations as to all but two areas were insufficiently specific to constitute a valid basis of protest. Global Eng'g & Constr. Joint Venture,
B-275999.4, B-275999.5, Oct. 6, 1997, 97-2 CPD Para. 125 at 4; Cornet, Inc.; Datacomm Management Servs., Inc., B-270330, B-270330.2, Feb. 28, 1996, 96-1 CPD Para. 189 at 4. An offeror who receives specific information in its debriefing but ignores it when drafting its initial protest does so at its peril. Global Eng'g & Constr. Joint Venture, supra, at 4 n.2. Labat's specific challenges to these criticisms, raised for the first time in its comments, are untimely. Under our Bid Protest Regulations, protests based on other than solicitation improprieties must be filed no later than 10 days after the protester knew or should have known their bases. 4 C.F.R. Sec. 21.2(a)(2). These regulations do not contemplate the piecemeal presentation or development of protest issues; where a protester raises a broad ground of protest in its initial submissions but fails to provide detail within its knowledge until later, so that a further response from the agency would be needed for a fair review of the matter, these late issues will not be considered. Litton Sys., Inc., Data Sys. Div., B-262099, Oct. 11, 1995, 95-2 CPD Para. 215 at 2-3.
Given their similarities, our discussion of the two areas of concern that were protested in a timely fashion will focus on the mailroom task area as an example.
Again, the RFQ listed several criteria under which the TEC was required to evaluate each offeror's technical approach as explicated in the estimating model. Under one criterion, offerors were required to "provide a model that describes the discrete labor in each major process, e.g. mailroom, for each form type and task area. The INS provides the estimated volume of forms for evaluation purposes only."
RFQ Para. 4.2.1. Under another criterion, offerors were required to explain their methods and rationale for allocating labor among the functional areas, such as mailroom operations, given the forecasted demand. Id. As noted above, INS's estimated volumes were defined as the quantities entered into the RFQ's pricing tables; these estimates of forecasted demand were based on 1999 data.
Labat's technical proposal did not use the estimates of forecasted demand for the mailroom task area. When the TEC compared the estimates of forecasted demand with Labat's estimates, it found that Labat understated the quantity of mail -- and, hence, the necessary labor hours -- in the mailroom task area at every service center. The aggregate error was more than 7 million pieces of mail in the base year, or 25 percent of the 30 million pieces specified in the RFQ workload estimate. This resulted in a substantial staffing understatement.
Labat has argued that the lower volume of mail projected by its model was the result of its decision to transfer certain tasks, [DELETED], from the mailroom task area to the [DELETED] task area. [DELETED]; the volume of mail was not understated but just moved to another task area.
The TEC Chair states that the TEC realized Labat proposed to move some of the [DELETED] effort that was historically accounted for in the mailroom to the [DELETED], but that its proposal did not explain the methodology it used to do so in such a way that the TEC could ascertain the basis for the difference in mail volume. More generally, the TEC understood that Labat based its projections on multiple years of workload data to which it had made certain adjustments, but the proposal did not contain an explanation of these adjustments that would enable the TEC to assess whether Labat's estimates of mail volume would allow for sufficient staffing. Since it was the TEC's responsibility to evaluate whether the offeror had allocated sufficient staff to meet the estimates of forecasted demand as defined in the RFQ, and since Labat's approach was based upon differing and unexplained understated estimates, the TEC was forced to calculate on its own the impact of this understatement on the offer by using Labat's underlying information against the RFQ quantities with the resulting staffing shortfall. The TEC could not, as Labat asserts, simply subtract the 1999 quantity for [DELETED] from the mail volume because Labat did not base its projections on the 1999 data. Under the circumstances, we cannot find the TEC's evaluation improper.
Labat contends that, during discussions, INS "told" it to base its mailroom estimates on the multiple years of data referenced in its proposal. The record does not support Labat's contention.
In an October 25 discussions letter sent by the BEC, Labat was advised that its price for the mailroom task appeared to be unbalanced. The BEC said it believed that the error resulted from Labat's dividing the total cost for the mailroom by a greater quantity of pieces than Labat used to generate the hours. In this regard, while the narrative in Labat's proposal stated that the firm used the historical data from the period [DELETED] as the basis for its labor estimates per form and unit pricing, the firm's estimating model used the 1999 data to develop its unit pricing. The BEC advised, "The Offeror should have used the number of pieces for the period [DELETED] specified as the Labat basis of estimate [in its proposal]. The per piece price should have been calculated first with the appropriately related number of pieces for the period forming the basis of estimate, then multiplied by the RFQ quantity for each center for the base and option periods." Discussions Letter, Oct. 25, 2000, at 1.
This discussions item did not "tell" Labat to use its estimates based upon [DELETED] workload data to develop its model, but simply acknowledged that Labat was using that data to arrive at its estimates and cautioned that this decision created an inconsistency. INS further advised Labat as to how it could use its estimates and solve the problem by basing its per piece price on the estimates it had calculated using the [DELETED] data, but then multiplying the resulting per piece price by the RFQ quantities for purposes of evaluation. This advice regarding Labat's price proposal, not its technical proposal, was consistent with the RFQ and should have served to notify the firm that, if it planned to base its estimates on a different set of data than that reflected in the RFQ's quantities, it needed to provide a way for the agency to make connections between the two.
In this regard, we do not read INS's evaluation documents or pleadings to prohibit Labat's use of this larger set of workload data to arrive at its estimates. The record shows the agency's willingness to trace the connections between Labat's estimates and those in the solicitation; Labat's proposal was not downgraded because it did not use the RFQ's estimates per se but because it did not provide a sufficient connection between the RFQ's estimates and its own. Since the RFQ specifically required the TEC to use the 1999 data for purposes of evaluation and Labat did not provide the agency with sufficient information to enable it to do so, we cannot find the TEC's evaluation unreasonable.
Since we cannot conclude that the price or technical evaluations were unreasonable, we have no basis to question the SSAC's best value analysis springing from those evaluations. The SSAC's comparative analysis of the offerors' discriminators showed that JHM's proposal presented seven features and no risks as opposed to Labat's seven features and two risks, and that the proposals' features were similar. The SSAC concluded that whether the overall rating for JHM's technical proposal was "Good" or merely "Acceptable," it offered the technically superior proposal at a negligible difference in price, with no major risks. Labat has provided us with no basis to find this best value analysis, and resulting award decision, unreasonable. See Keane Fed. Sys., Inc., B-280595, Oct. 23, 1998, 98-2 CPD Para. 132 at 15-16.
The protests are denied.
Anthony H. Gamboa General Counsel
1. JHM has asked that we reconsider our position that we have jurisdiction over protests where the solicitation anticipates the issuance of task orders placed under GSA schedule contracts. Severn Cos., Inc., B-275717, Apr. 28, 1997, 97-1 CPD Para. 181 at 2 n.1. In light of JHM's concerns, we have reviewed our rationale for assuming jurisdiction over such protests, as set forth in Severn Cos., Inc., and find no basis to change our position.
2. Labat calculated two administrative overhead weighted labor rates, [DELETED]. Both rates contain hard-coded entries. Our discussion is limited to the former rate, since Labat acknowledges that it mistakenly failed to use the latter rate in calculating its prices. This failure was apparently not noticed by the BEC, but it raises additional questions about Labat's pricing.
3. The source selection did not turn on this issue, and INS did not reach the point of performing a risk analysis to determine whether award on the basis of an apparently unbalanced offer would result in paying unreasonably high prices or would otherwise present an unacceptable level of risk to the government. FAR Sec. 15.404-1(g)(2).
4. Labat makes a similar argument regarding JHM's allocations of data entry hours, which we reject for similar reasons.
5. In addition to this deficiency, the TEC noted two new weaknesses in Labat's FPR, both of which were associated with inconsistencies between the estimating model and technical approach. While these inconsistencies are, in fact, present, the SSAC did not view them as major, SSAC Report at 14, and we will not discuss them further.