Heidtman Steel Products, Inc. v. Compuware Corp.

ORDER CARR, District Judge. This is a diversity case in which the plaintiff, Heidtman Steel Products, Inc. (Heidtman), seeks recission of a contract between it and Compuware Corporation (Compuware). Compuware has filed a counterclaim, seeking payment of past due invoices. As originally filed and thereafter amended, the plaintiffs complaint contained several other claims, which were dismissed on defendant’s motion. 1 The parties agreed to a non-jury trial, the plaintiff has rested, and the defendant has moved under Fed. R.Civ.P. 52(e) for judgment on partial findings — i.e., judgment in its favor on the basis that the evidence presented during plaintiffs case shows that the defendant is entitled to a verdict in its favor. For the reasons that follow, defendant’s motion shall be granted. Further proceedings shall relate solely to defendant’s counterclaim. Background Heidtman is a steel processor with headquarters in Toledo, Ohio, and processing plants in Ohio, Michigan, Illinois, Indiana, and Maryland. It purchases coils of steel, which it cuts in slitting machines to fill customers’ orders. The coils vary in several respects, including their chemical composition, physical properties, such as tensile strength, and thickness. Different customers have different needs. Heidtman meets those needs by having in stock, or by being able quicMy to obtain, steel of the precise kind that each customer needs, cutting it to order, and delivering it promptly. In thirty-five years Heidtman has grown from a small operation to the largest company of its kind in the country. The process by which Heidtman determines which steel a customer needs, whether it has steel of that kind in stock or needs to obtain it, and what is to be done with the steel to meet the customer’s order is called “steel application.” The Heidt-man personnel who do steel applications are called Steel Applicators. Efficient steel application is crucial to Heidtman’s business success. Failure to realize that steel is in stock or available in sufficient quantities to fill a customer’s needs can lead to delays in a business that is increasingly time-sensitive. Whenever possible, processing costs are reduced when orders from more than one customer can be filled from the same coil and in a single “set up” of the slitting equipment. Filling more than one order with a single set up reduces movement of coils in and out of inventory, cuts down on the time and manpower costs required to prepare the slitter to cut the steel, and limits scrap. In the 1980s, Heidtman used a computer system, called the “Legacy system,” to assist its business activities. Even with the Legacy system, steel application was largely a manual operation. Though the Legacy system could tell a Steel Applica *871 tor what was in stock, the applicator still had to use pencil and paper to calculate whether there was more than one order for a particular kind of steel and how the order or orders otherwise could be processed. By the early 1990s, Heidtman determined that the Legacy 'system was not adequate for its business. Heidtman also concluded that the Legacy system could not be expanded or upgraded, and that it was necessary to acquire a new system. That led Heidtman to contact Compuware. Compuware designs and installs computer systems to meet its customers’ needs, which, as with Heidtman and its customers, vary significantly from customer to customer. Compuware undertakes to learn what a customer’s requirements are by looking at the customer’s business practices and products, determines appropriate solutions, in terms of computer hardware and software, and acquires or provides the equipment necessary to satisfy the customer’s needs. On some projects, Compuware works with other vendors, such, as in this case, as Oracle Corporation, which is a seller of software. Compuware and Oracle are among the largest companies of their kind in the country. On July 28, 1993, the Heidtman and Compuware signed the first in a series of agreements (a “Technical Services Agreement”) for what came to be known as the “Plus Project.” In that agreement, Com-puware agreed to provide computer programming expertise, coding, and related services. The agreement also included provisions relating to compensation, payment of bills and expenses, assignment of work, liability, modification, termination, limitations on warranties and remedies, and choice of law. Subsequently, Ernst & Young, which initially was to have participated in the project, withdrew. As a result, Compuware’s role changed. These changes were memorialized in a series of agreements, called phase agreements, drafted by Compuware at various stages of the project. The Plus Project came to have four phases, with a fifth being contemplated at the time work terminated on the project at the end of April, 1997. Written agreements were entered into between the parties for phases one, two, and four. No written agreement was signed for phase three. The phase agreements defined the work to be performed during the phases to which they related. In one of my pretrial rulings, I held that the initial services agreement and the phase agreements were to be read together as a single contract and governed by Michigan law. The Phase One Agreement was signed on August 27, 1993, by Salena (Sally) Colby for Compuware and James L. Hill for Heidtman. Among other provisions, the Phase One Agreement describes a “system life cycle,” describes the principal “deliver-ables,” defines the parties’ responsibilities, and prescribes a time and materials method of compensation. Heidtman agreed to appoint a steering committee and designate a Project Director. The system life cycle, as described in the Phase One Agreement, was to consist of: • project initiation • requirements definition (i.e., determination of Heidtman’s hardware and software needs in light of its business operations and practices) • logical design (i.e., the functional design of the system’s operations) • physical design (i.e., the technical design of the system) • construction (i.e., coding of the programs and installation) *872 • training and implementation (i.e., training and “rollout” of the system in final form) • maintenance and enhancement According to the Phase One Agreement, the “Information Requirements Definition” process was to take the project through the second step, requirements definition. I find, in light of the evidence at trial, that the parties contemplated that by completion of this step, Compuware would have become sufficiently familiar with Heidt-man’s business operations and practices and computer needs to be able to participate effectively and efficiently in the next steps, which were to begin with design of the system, continue with construction, and be completed with training and implementation. This finding is supported by the description in the Phase One Agreement of the “deliverables” to be provided during the project’s initial stage. Those deliverables (i.e., the products) were to be accomplished through a series of “Joint Application Definition” (JAD) workshops. These were to be sessions at which Compuware learned what it needed to know about Heidtman to be able to proceed with functional and technical design of the system. Comprehensive understanding of the user’s business, and how that business can be improved by computerization — the objectives of the JAD workshops — are crucial in any computer project. As the project moves further into the design and coding stages, discovery of new user requirements or redefinition of already identified requirements can significantly and adversely affect software design. Even a minor change can require extensive reworking and redesign of many inter-related components. Such reworking increases the risk that the system will have bugs, which in turn, will need additional work to correct. Having to re-do work that seemed to have been completed is one factor that can contribute to “scope creep” (i.e., potentially unconstrained expansion of a project beyond its initially anticipated budget and technical boundaries). The Phase One Agreement, as noted, also defined the parties’ responsibilities. A portion of thfe Phase One Agreement stated that “Compuware contributes experience and expertise to collecting, organizing and analyzing facts concerning the business organization and its need for systems.” In addition, that provision emphasized Heidtman’s obligations to “provide the facts and participate in every phase of the analysis,” appoint a management level steering committee, and assign a Project Director, who was to “devote essentially all of their [sic] time to the project.” Among other tasks, the Heidtman project director was to work with the Compuware project manager, “represent the project to the Steering Committee, and act as the liaison between the steering committee and the rest of the project team.” The Heidtman project director was James L. Hill. Hill had worked previously for a large accounting firm and a computer consulting company owned by Heidtman. At the time of his assignment as Heidt-man’s project director on the Plus Project, Hill was Heidtman’s director of management information (i.e., computer) services. Although plaintiff did not call Hill as a witness, the evidence during its case projected a clear picture of his role in the Plus Project and responsibility for its fate. I find that Hill was an active participant in the project, made several important decisions, including several that were to have an adverse impact on the project, was thoroughly knowledgeable about its course, and was in every respect Heidt-man’s agent in its relationship with Com-puware and the other vendors on the project. While there may be much to fault *873 about Compuware’s management (at least when viewed, as it has been so far, solely through the prism of plaintiffs evidence), the record is also clear that Hill played a significant role in the events leading to this lawsuit. Furthermore, as also noted, the Phase One Agreement provided that Heidtman was to compensate Compuware on a time and materials (as opposed to a flat fee or fixed cost) basis. “It is intended,” the Phase One Agreement stated, “that the project will be managed toward the budget as established.” The estimated budget for developing Heidtman’s information requirements definition was $384,000, which was “believed to be accurate within a +/- 10% variance range.” The evidence at trial shows that the principal goals envisioned in the Phase One Agreement were not accomplished. Of most crucial importance, the JAD workshops failed to disclose Heidtman’s requirements as fully and completely as needed to enable the design phase to be effective and efficient. For almost the entire duration of the project, Compuware was learning about new requirements, which meant that a lot of the work done during the design phase of the project had to be redone. This, combined with an apparent lack of technical expertise on the part of some of the persons assigned by Compuware to the project and other circumstances, created massive scope creep. Other problems affected the project from its earliest stages. Among these was resistance on the part of Heidtman personnel to the prospect of changing business practices and processes. This accounted for some, though hardly all, of the problems with requirements definition. To cite a single, but representative example: Heidtman’s employees were opposed to having to work with a computer screen that looked different in major respects to the screen that they had been accustomed to viewing with the Legacy system. In addition, different employees would often express the need for different requirements at different times during the project’s course. User understanding and acceptance of change are important aspects of a project. No provider of computer services should expect that a customer and its employees will abandon customary methods of running their business in toto. But, on the other hand, the customer should be open to manageable changes in its practices that will facilitate using simpler, rather than more complicated, computer solutions to the problems sought to be solved with the new system. The record fails to show a reasonable degree of accommodation to change on Heidtman’s part, which compounded Compuware’s failure to ascertain, or refusal to proceed until ascertaining, Heidtman’s requirements comprehensively and with finality before the design phase began. The effects of these circumstances were amplified by the failure of the Steering Committee to participate actively in the project. Indeed, by December, 1993, the Committee had ceased meeting. As a result, Hill, who was to have been the interface between the project and the Committee (which, in turn, was to speak on behalf of Heidtman) became Heidtman’s primary, if not sole spokesman. This was so, even though, for a period of time, Hill was assigned to the project on only a part-time basis. Hill had made clear from the outset in conversations with Sally Colby, Compu-ware’s senior project manager (until she was removed at Hill’s request in mid-summer, 1995), that he wanted Compuware to bring in someone who would allow Heidt-man to run the project the way it wanted it run. According to Colby, Hill wanted, *874 given what he considered to be his “system expertise,” simply some support for his overall responsibility for managing the project — someone to do day-to-day activities. Throughout the project’s span, Hill was Heidtman’s most active participant. As such, the record discloses, he had primary, if not exclusive input on Heidtman’s behalf into day-to-day operational issues and more crucial, and more far-reaching, decisions, such as personnel assignments, methods and mode of reporting to Heidt-man on project developments, and project direction. In addition, the budget for the information requirements development phase was exceeded, beginning a trend that would accelerate in the months and years to come. A budget adopted in April, 1994, was exceeded immediately; a revised budget adopted in September, 1994, was exceeded in January, 1995; the budget adopted in April, 1995, was exceeded by August, 1995, by which time more than $5,000,000 had been spent on the Plus Project. Expenditures already had exceeded the final budgets (of about $6,000,000) which were adopted in November, 1995, and April, 1996. By the time the project ended, approximately $12,000,000 had been spent — more than double the last budgeted amount. In addition, Compuware had billed Heidtman for approximately $2,700,000 for additional services. Although the principal goals of the first phase had not been met, because Heidt-man’s requirements had not been ascertained comprehensively, the parties signed the Phase Two Agreement on February 3, 1994. The deliverables during phase two were: • Recommended database and application vendor (i.e., select the software vendor) • Initial system architecture and configuration (i.e., functional design) • Detailed application package analysis (i.e., determine extent to which “vanilla” software could meet Heidtman’s requirements) • User defined package modifications, enhancements and additions (i.e., determine extent to which software had to be customized to meet Heidtman’s needs) • Updated project plan estimates through implementation As in the first phase, Heidtman was to pay Compuware on a time and materials basis. Phase two was to last three months, from January 17 to April 15, 1994. Its estimated budget was $337,000. Attached to the exhibit containing the Phase Two Agreement is a project meeting agenda. That document includes a total estimated total budget for 1994 of $2,267,000. Six hundred thousand dollars of that total was allocated to the purchase of software. The estimate of the cost for phase two, $337,000, was based on several assumptions. Most importantly, at least with regard to this lawsuit, was the assumption that “[ijnformation obtained in Phase I regarding [Heidtman’s] business processes and information flow is complete and accurate.” This assumption was unfounded. In addition, as the project was about to embark on this phase, Hill became assigned part-time to the project as Heidt-man’s project director. He, along with Robert Bryson, Compuware’s project director (also part-time), were to report to an executive (i.e., steering) committee. Sally Colby was to report to Hill and Bryson. Oracle was selected as the software vendor in the Spring of 1994. Though recom *875 mended and endorsed by Compuware, selection of Oracle was made by Hill. Phase three appears to have started around May, 1994, after Kathleen Gahan, an Oracle employee charged with the system’s functional design, began working on the project. At Hill’s direction, there was no written Phase Three Agreement. There were, likewise, no specification of personnel assignments and responsibilities, definition of deliverables, or either short- or long-term budget estimates. Phase three lasted until late 1994 or early 1995. Problems with the Plus Project that were apparent during this period were flagged in an internal audit prepared for Compuware by Joan Colwell, Director of Project Services, in the Fall of 1994. Hill earlier had rejected Colwell’s suggestion that an external audit be prepared. Colwell’s memo expressed concerns about the failure adequately to articulate Heidtman’s business needs, the lack of user involvement, high turnover of Compu-ware’s staff, inadequate risk management and related issues. Although, during the course of preparing her audit, Colwell had sought to interview Hill, he had not found time to meet with her. Hill’s reaction to receiving a copy of the audit was even more noteworthy than the memo itself — he threw it in a wastebasket. This memo, despite Hill’s inappropriate response, placed Heidtman, through him, on relatively early notice that the Plus Project had some serious flaws that required corrective action if the project were not to risk failure. Though plaintiff contends that Colby should have seen to it that a copy of this audit was given to Mark Ridenour, Heidt-man’s Chief Financial Officer, and member of the putative Steering Committee, I disagree. It was not her responsibility to do Hill’s job, which, in part, was to keep Heidtman’s management fully informed about the project. Colby fulfilled whatever obligation she may have had to ensure that Heidtman, following the Colwell audit, was kept abreast of the status of the project and its problems by submitting weekly status reports beginning in early 1995. Gahan’s work on the project lasted until January, 1996, when functional design was completed. Her testimony confirms Hill’s role in running the project from the Heidt-man side. Gahan was assigned “gap analysis,” which determines what the software as delivered can and cannot do, and what software customization is needed to meet the customer’s needs. Gahan was troubled by several conditions, including: inadequate requirements gathering earlier in the project; lack of cooperation on the part of some Heidtman users; and reluctance on the part of many Heidtman users to accept changes in the design of screens from the Legacy system. She likewise expressed concerns about the competence of persons assigned to work with the project by Compuware. By the time Gahan left the project in January, 1996, her reservations about the competence of the Compuware personnel had not been alleviated, and she had concerns about whether the project could be completed successfully- By January, 1995, Heidtman, through Ridenour and Rick Tarquinio, Vice President for Purchasing and Quality Control, was expressing concerns about rising costs, having been informed that the additional cost of completing the work, on which about $2,000,000 had already been spent, would be about another million dollars. The parties signed the Phase Four Agreement on February 20, 1995. That agreement indicates that work on phase four had begun on November 1, 1994, and was scheduled to be completed by October *876 15, 1995. The deliverables included implementation of: • Oracle financial applications (i.e., conversion to Oracle financial program) • Heidtman customer service requests (i.e., requests for changes in the software to conform to Heidtman’s needs) in its production environment • Oracle manufacturing applications into the production environment • Interfaces (i.e., between other computer programs and the Oracle software) into the production environment In other words, the Phase Four Agreement anticipated that the entire project would have been completed, installed, and operational by October 15,1995. The Phase Four Agreement also specified that “Compuware is responsible for the overall project management of the project, in conjunction with you [i.e., Hill], and Compuware personnel on the project. Compuware is not responsible for rates or performance of personnel from [other vendors].” The project budget for this phase was estimated at $1,974,000. From a memorandum sent by Colby to her supervisors at Compuware about a meeting with Ridenour and Hill on June 6, 1995, it was clear by that date Compuware would not meet either the completion date or estimated budget, as revised a few weeks earlier. Prior to the meeting, according to the Colby memo, Hill had wanted to “pull the pin now” on the project. Colby encouraged Hill to have Heidtman’s executive officers make that decision, though he also understood that he could set forth his views at the June 6th meeting. He did not do so. In her memo Colby stated that it would cost twice as much as estimated to finish the project, but that Compuware was unable to provide a detailed breakdown at the June 6th meeting. A summary was, however, given to Ridenour: it indicated that the total project cost would be nearly $5,000,000, or slightly more than $3,000,000 over the initial estimates. Ri-denour was described as “anxious to see the bottom line,” and wanting to have a fixed price understanding — which Compu-ware had indicated would not be acceptable. Attached to the Colby memo to her supervisors at Compuware was a memorandum that had been given to Ridenour at the June 6th meeting. That memorandum details the problems that had arisen during the project. Among these were difficulties in ascertaining Heidtman’s requirements, user opposition to differences in screen appearance, reworking of designs, insufficiently rigorous standards, and absence of a central design repository. The attachment also delineates risks facing the project, including the insufficient user involvement in design, uncertainty about managerial responsibility, absence of project milestones for evaluating progress, and lack of change control procedures. At about this time, Heidtman, Ridenour testified, gave some consideration, though, he testified, “not seriously,” to cancelling the project. He acknowledged that he “may have” threatened to cancel the project in June, 1995, and that he is not one to make threats lightly. These considerations notwithstanding, Ridenour approved successive budgets in April, 1995, and May, 1996. There is nothing in the record showing that Ridenour thereafter received (or asked for) further budgetary estimates, or that he was asked to approve, or approved revised budgets. He acknowledged that it would have been Hill’s responsibility to tend to such matters and keep him informed. The record does not show that Hill did so, though costs were increasing dramatically during the Summer and Fall of 1996 and early *877 Winter of 1997. I find that during that period there was no fixed budget, or understanding about what the budget was, between the parties. To the extent that Ridenour and Heidt-man contend that the May, 1996, budget remained operable, I reject that contention. By June, the estimated project implementation date already had been revised. In that circumstance, it would have been unreasonable, unrealistic, and, most importantly, most unlikely that a person in Ridenour’s position of Chief Financial Officer would have believed that the extension would not require additional work, or would not result in costs in excess of the budget. Between May, 1996, and February, 1997, Ridenour did not meet with Compuware representatives to discuss the project, its costs, or the shifting completion date and cost estimates. Ridenour did, however, have conversations with various Heidtman employees during late 1996 and early 1997, in which he learned that there were serious problems with the system, including the fact that programs not only were not complete, but that they were not close to completion, and that it was highly doubtful that they could be completed within a reasonable period of time. Ridenour acknowledged that during the Fall of 1996, Hill had suggested the possibility of canceling the project. In early February, 1997, Ridenour learned that the total cost of the project from its inception to completion was likely to be $9,269,000. The May, 1995, budget had been for approximately $4,450,000. Ridenour also was told that actual expenditures had been slightly less that $6,000,000; he was to compute them to have been as much as $8,000,000. There ensued a series of conversations with Compuware, from which Ridenour did not believe that he had gained a complete understanding of the situation. In mid-March, 1997, Ridenour fired Hill and appointed a new oversight committee. By that point, Ridenour had decided to become actively involved in attempting to learn where they were, why they had paid so much, and where project stood. In his view, he was “staring at at least $8,000,000,” which was “way beyond approved expenditures — and still sitting without a project completion or a system that is implemented and working.” In the meantime, about $2,700,000 in unpaid invoices from Compuware had accumulated since the preceding November. Despite conversations and meetings with Compuware representatives, Ridenour never felt that he understood how the cost overrun had grown to be so large. On February 15, 1997, Compuware stated that the estimated completion date would be July 4,1997. By this time, John Bates, Heidtman’s owner and CEO, also had become more aware of the problems with the Plus Project. Attending a demonstration/training session in early April, 1997, Bates, an experienced Steel Applicator, tried to operate the system. He concluded that performing the steel application function would be more cumbersome and time-consuming with the new system than with the Legacy system it was meant to replace. Before attending that session and trying the new system, Bates had understood that it was ready to be “launched.” Believing the system to be unworkable and unacceptable, Bates called Ridenour and told him the he did not want him “to pay another dime,” noting that the project was several years late. In response, Ri-denour said, “you don’t know how bad it is.” Had the demonstration been acceptable, Bates would have paid the outstanding invoices. *878 On April 2, 1997, Ridenour sent a letter to Geri Carlson, who had replaced Colby as Compuware’s project manager, stating that Heidtman was unable to approve the time sheets submitted in support of outstanding invoices. A meeting on April 11, 1997, with Compuware representatives did not resolve the time sheet issue. Though Bates still was reluctant to pay anything further to Compuware, he acceded to Ridenour’s suggestion that Heidtman offer to pay $1,000,000 to have the project completed. Though the record contains inconsistent testimony about the purpose of the $1,000,000 offer, I find that Riden-our and Heidtman intended the payment to be for both the work remaining to be done and the $2,700,000 in unpaid invoices. In other words, Heidtman was proposing to make one last payment for the total project. In light of Bates’ instruction not to pay “another dime,” Bates’ reluctance to extend even the $1,000,000 offer, and Ri-denour’s question, relative to the April 11, 1997 meeting, “where’s this animal going to end,” it is more likely than not that Bates and Ridenour viewed that offer as covering both the unpaid invoices and the cost of getting the job, at long last, done. That offer was made at a meeting with several Compuware representatives on April 30, 1997, and was confirmed on May 1, 1997 in a letter from Ridenour to Kenneth Baldwin, Compuware’s Director of Professional Services. Shortly after the meeting, Compuware instructed Heidtman personnel to leave Compuware’s premises in Farmington Hills, Michigan, and it terminated the landline service between Heidtman and Compuware. In addition, Compuware’s president, Joseph Nathan, sent a letter on May 1, 1997, to Ridenour rejecting the $1,000,000 offer and stating that “the issue of payment needs to be resolved before Compuware can move forward.” In a letter on May 7, 1997, to Nathan, Ridenour reformulated Heidtman’s offer, stating that it was for completing the project, with further discussions needed if Compuware wanted payment on its outstanding invoices before proceeding further. On May 8, 1997, Compuware’s General Counsel, Thomas Costello sent a letter to Bates, confirming Compuware’s rejection of Heidtman’s, $1,000,000 offer, demanding payment of the unpaid invoices, and stating that Heidtman had terminated the project in light of its failure to pay invoices in a timely manner. After the project ended, Heidtman retained Ernst & Young to evaluate the situation. Sometime during the Summer of 1997, Ernst & Young reported orally that, in its view, nothing that Compuware had provided to Heidtman was of any use, or even could be salvaged, except for the financial applications, which were valued at about $232,000. About four months later, in early December, 1997, Ernst & Young confirmed its evaluation in a written report. Shortly after receiving this report, Heidtman filed an amended complaint (this suit having originally been filed in May, 1997, almost immediately after the project ended), adding its recission claim. Through its recission claim, Heidtman wants to recover about $12,000,000 which it spent on the Plus Project. Aside from the financial application, Heidtman is not using any portion of the Plus Project system. Indeed, Heidtman continues to rely on the outmoded Legacy system, which becomes increasingly difficult due to the unavailability of support for a system that is, in computer terms, antiquated. Discussion Under Michigan law, the party seeking rescission of a contract must prove three elements: *879 (1) a seasonable assertion of the rescission right; (2) tender of the consideration and benefits received; and, (3) demand for repayment of any price paid. See Mesh v. Citrin, 299 Mich. 527, 300 N.W. 870, 872 (1941). The party seeking rescission must first return the other party to the pre-contract status quo, McIntosh v. Fixel, 297 Mich. 331, 297 N.W. 512, 518 (1941), and rescission is not available to a party who has failed to make payments required by a contract and is thus in default. Hafner v. A.J. Stuart Land Co., 246 Mich. 465, 224 N.W. 630, 631 (1929). Two Men and a Truck/Int’l Inc. v. Two Men and a Truck/Kalamazoo, Inc., 949 F.Supp. 500, 507 (W.D.Mich.1996) (citing Dynamic Enterprises, Inc. v. Fitness World of Jackson, 32 B.R. 509, 522 (Bankr. M.D.Tenn.1983), abrogated on other grounds, In re SMEC, Inc., 160 B.R. 86 (M.D.Tenn.1993)); see also Gloeser v. Moore, 283 Mich. 425, 278 N.W. 72, 74 (1938) (to recover in recission, plaintiff “must seasonably assert such rescission, tender back what he has received, and demand repayment of the purchase price.”). 1. Assertion of the Right to Recission was Seasonable Heidtman did not assert a demand for recission until it filed its first amended complaint on December 9, 1997. Heidtman had, however, known since mid-summer, 1997, at the latest, that it had purchased a computer system that was, for its purposes and in its view, worthless. Though Ernst & Young filed no formal report until shortly before Heidtman filed its complaint, Ernst & Young had told Heidtman orally about four or five months earlier that, aside from the financial applications, Heidtman had received nothing useful after spending $12,000,000. Heidtman, moreover, had contemplated cancelling the contract as early as June, 1995, about the time of the meeting, memorialized in Colby’s June 6, 1995, memo, between her, Ridenour, Hill, and Carlson. In addition, Hill had suggested cancelling the project in the Fall of 1996. Compuware argues that for much of the life of the project Heidtman, either through Hill, in light of his extensive involvement with the project, or more directly through meetings in which Ridenour participated and memoranda which he received, was aware of the manifold and growing problems with the project. This contention is true, but it ignores the fact that Compuware and the other vendors continued to work on the project, and, more importantly, communicated in word and action the expectation that it would be completed successfully. Despite Heidtman’s occasional thoughts about cancelling the project and the extensive awareness of its continuing and expanding problems, I cannot find, on the basis of the evidence presently before me, that its demand for recission was untimely. “[WJhat is a reasonable time,” the court stated in Grabendike v. Adix, 335 Mich. 128, 55 N.W.2d 761, 767 (1952), “depends upon the circumstances of each particular case.” The facts in this case are similar to those in Fargo Mach. & Tool Co. v. Kearney & Trecker Corp., 428 F.Supp. 364 (E.D.Mich.1977), involving problems with automated industrial equipment used for machining metal. The court stated: [C]ase law uniformly recognizes that complicated machinery and systems may have relatively long ironing out periods during which the buyer in good faith continuously attempts to have the seller cure or repair complicated problems. Accordingly, courts have held that revocation in such circumstances is not un *880 reasonably delayed where buyer promptly notifies seller of the defects, attempts at cure are ongoing, and buyer does not formally notify of revocation until it is apparent that seller cannot perform repairs. See Dopieralla v. Arkansas-Louisiana Gas, 255 Ark. 150, 499 S.W.2d 610 (1973) (forty-month delay after installation was not unreasonable for heating/air conditioning system); Fablok Mills v. Cocker Machine & Foundry Co. [125 N.J.Super. 251, 310 A.2d 491 (1973)], supra, (two-year delay was not unreasonable for knitting machines); Regents, The University of Colorado v. Pacific Pump, 35 Colo.App. 36, 528 P.2d 941 (1974) (three-year delay was not unreasonable for a power plant). At bar, assuming that Fargo’s institution of suit constituted notice of revocation, the time that had elapsed since installation was approximately eighteen months. Repairs had been “continuously” attempted in the sense that Kearney & Trecker’s last service call was made less than three weeks before suit was filed, although repairs on the three principal complaints had not been attempted for some time, and the majority of calls were for unrelated problems which were corrected. 428 F.Supp. at 379. 2 This case involved a highly complicated computer system. Compuware never told Heidtman that it would be unable to provide a working system that would meet Heidtman’s needs. Those reassurances, particularly in the context of something as complex as a computer system, much of which had to be customized to meet Heidt-man’s needs, make Heidtman’s failure to have given notice of an intent to rescind prior to the. contract’s termination reasonable. For the next few months, Heidtman was attempting to ascertain whether the system could be functional. Though it learned by mid-August that it could not work, I do not find, on the basis of the record now before me, that its delay of less than four months before making a demand for recission bars its claim. Compuware, which had ceased working on the project long before then, and which had been sued shortly after it had stopped work, lost nothing as a result of the delay. To be sure, delay can bar Heidtman’s claim even without a showing of prejudice. Banque Arabe Et Internationale D’Investissement v. Maryland Nat. Bank, 850 F.Supp. 1199, 1211 n. 6 (S.D.N.Y.1994). That does not mean that the absence of prejudice, such as exists here, should not be taken into account when determining whether notice was seasonable. The question of prejudice is simply one factor to consider, along with all the other circumstances. Compuware claims that Heidtman assumed the risk of project failure, and that, due to such assumption, Heidtman is not entitled to rescission. See McCredie v. Buxton, 31 Mich. 383, 1875 WL 6392, *4 (1875). This doctrine is not applicable where, as here, the risks that were assumed were accepted in the expectation, which arose in substantial part from Compuware’s reassurances, that the project would be successful. I conclude, accordingly, that notice of the demand for recission was not unrea *881 sonably delayed, and was seasonable. Heidtman’s claim for recission cannot be barred on that basis. 2. Heidtman Failed to Fulfill its Obligations Under the Contract “Rescission is not available,” as the court stated in Hafner v. A.J. Stuart Land Co., 246 Mich. 465, 224 N.W. 630, 631 (1929), “to a party who has failed to make payments required by a contract and is thus in default.” I have previously ruled that the defendant has the burden of proof on the issue of Heidtman’s performance, or failure thereof, of its obligations under the contract. (Doc. 261). Under Rule 52(c), judgment can, however, be entered in favor the defendant on an issue as to which it has the burden of proof where admissions by the plaintiffs witnesses show that the defendant is entitled to such judgment. Nobelpharma AB v. Implant Innovations, Inc., 141 F.3d 1059, 1065 (Fed.Cir.1998). Those admissions are present in this case. The contracts were payable on a time and materials basis. Starting in November, 1996, Heidtman stopped paying invoices, which continued to be submitted and were due for work that had been performed with Heidtman’s knowledge and concurrence. In April, 1997, when Heidtman’s owner and C.E.O., John Bates, tried to use the system, he had been unable to do so to his satisfaction. Indeed, he had concluded that it was more cumbersome and less efficient than its antiquated predecessor, the Legacy system. Shortly thereafter, Bates had instructed Ridenour, not to “pay another dime.” This led to Ridenour’s offer, to which Bates had given reluctant assent, of a fixed and final payment of $1,000,000. That amount was intended and offered as payment on the outstanding invoices and for completion of the work. By making that offer, and refusing to pay the past due invoices, Heidtman was in default of its obligations under the contract. Heidtman did not have the option, without Compuware’s assent, unilaterally to convert the contract to a fixed price agreement. That is, however, what Heidt-man was attempting. Heidtman contends that the evidence shows that there was a “billing dispute” and that invoices could not be reconciled. There may well have been substantial disagreement between the parties about the amounts owed. But Heidtman never has contended that it owed nothing, or even that what it owed was insubstantial. Instead, it asserts that it could not get a comprehensive and comprehensible accounting for all the charges it was being asked to pay. This inability to resolve all uncertainties, Heidtman appears to contend, entitled it to refrain from paying anything at all on the outstanding balance. Heidtman did not make an offer to pay those amounts as to which there was no disagreement, and concurrently to continue to pay for future work on the project. Had it done so, the circumstances arguably might have been different. But that was not Heidtman’s offer. Instead, it proposed a fixed payment to cover past due and forthcoming invoices. 3 It was not entitled *882 to make such demand, and its continuing failure to pay the invoices, or, at the very least, those invoices as to which there was no dispute,. violated its contractual obligations. Heidtman also contends that it was entitled to withhold payments because Compu-ware had exceeded the project’s budget. This argument presumes that there was a budget. The record shows, however, that after midsummer, 1996, the project had no budget. Contrary to this argument, the record shows that by midsummer, 1996, there was no budget for the project. The contention that Heidtman’s failure to pay past due invoices was justified because Compuware exceeded budgetary limits is without merit. I find, accordingly, that Heidtman was in default. That default was material, rather than inconsequential. While Heidt-man may not have agreed that it owed everything for which it had been billed, it did not take the position that it would pay what was not subject to dispute. Instead, by making its demand that Compuware accept one final, flat $1,000,000 payment, it was unilaterally renouncing its responsibility for paying, and asserting an intent not to pay, the outstanding invoices. Nothing in the contract or the law of recission entitled Heidtman to take that position. Having done so, it is not entitled to rescind and recover what it paid — even though its expenditure of $12,000,000, and continuing exposure on invoices of upwards of $3,000,000 have bought it nothing but a computer system that is nonfunctional and a lawsuit that is unsuccessful. Conclusion In light of the foregoing, I conclude that Heidtman is not entitled to recission the contracts or recovery against Compuware. It is, therefore, ORDERED THAT judgment be, and it hereby is entered pursuant to Fed.R.Civ.P. 52 for Compuware and against Heidtman on Heidtman’s claim for recission. The Clerk shall set this case forthwith for a scheduling conference to set a date for further proceedings on Compuware’s counterclaim. So ordered. Footnotes [1] In pretrial rulings, judgment was entered in favor of Compuware on Heidtman’s claims under the Uniform Commercial Code, and for negligence, professional malpractice, breach of implied duty of good faith, and promissory estoppel (Doc. 131), breach of implied and express warranties and certain other remedies (Doc. 143), and fraud. (Doc. 199). As a result of these rulings, the only claim remaining for trial was plaintiff's claim for recission. [2] In Uganski v. Little Giant Crane & Shovel, Inc., 35 Mich.App. 88, 192 N.W.2d 580, 589 (1971), which likewise involved “a highly complex piece of heavy equipment,” the court noted that “notice of election to rescind, ... is the same as revocation of acceptance under the Commercial Code.” It appears that principles from cases involving sales of goods are thus applicable, where such appears to be appropriate, to service contracts, such as the contract in this case. [3] Though not supported by the greater weight of the evidence, it is arguable that Heidtman, in light of the testimony of one of Heidtman’s managers, Cal Roth, believed that it was simply making an opening gambit directed toward compromising past due invoices and limiting future expenditures. If so, Heidtman did not make its understanding of what it was doing clear to Compuware. Compuware, for its part, appears to have construed the Heidt-man offer as a demand to convert the contract to a fixed price agreement to finish the project, while also being expected to forego recovery on the invoices — all for a flat *882 $1,000,000. This was a reasonable perception, even if (and contrary to my finding that that was the intent of the Heidtman offer), Compuware was mistaken to the extent that Heidtman anticipated further negotiation on the outstanding bills. Heidtman is responsible for any ambiguity about its intentions.