(concurring in part and dissenting in part):
¶ 54 I join with the majority in Part I of its analysis, affirming the trial court’s partial summary judgment in favor of Comtrol on the basis of accord and satisfaction. I also join in Parts III and IV of the majority opinion which address unrecoverable change orders and subcontractor lien releases. I respectfully dissent, however, from Part II of the majority’s analysis. I believe that the trial court could not rely on only national averages, but rather, Comtrol was required to prove the actual amount it spent to finish Traco’s subcontract work on the UVSC and Weber State projects. Furthermore, I conclude that the issue of whether average national hourly rates rather than evidence of the actual costs of completing the work could be used to prove Comtrol’s damages is an issue of law. See Bair v. Axiom Design, LLC, 2001 UT 20, ¶ 13, 20 P.3d 388 (“[T]he determination of whether a party has made out a prima facie ease is a question of law which we review for correctness, affording no deference to the trial court’s judgment.”). Consequently, I would not require Traco to marshal the evidence on this point. See Wardley Better Homes & Gardens v. Cannon, 2002 UT 99, ¶ 14, 61 P.3d 1009 (“Challenges to a trial court’s legal determinations ... do not require an appellant to marshal the evidence.”); see also Brigham Young Univ. v. Tremco Consultants, Inc., 2007 UT 17, ¶ 25, 156 P.3d 782.
¶ 55 To prove its damages, Comtrol relied on national cost data on the average wages paid to construction workers (National Construction Averages)1 rather than on evidence of the amount it actually incurred to complete the projects. My first concern is that this approach is contrary to the express contract language governing the rights and obligations of Comtrol and Traco. The default *583provision of each subcontractor agreement states:
In the event that Subcontractor appears likely to be unable to complete its work according to Contractor’s project schedule, or if Subcontractor fails to fully perform its duties under this Subcontract ... then Contractor may (a) withhold payment ...; (b) after giving 48 hours written notice to Subcontractor, eject Subcontractor and take over Subcontractor’s work and terminate Subcontractor’s right to perform under the Subcontract. If Contractor takes over Subcontractor’s work, the Contractor will charge Subcontractor for all costs incurred as a result, including reasonable overhead and profit and including attorney’s fees and other expenses.
(Emphasis added.) The subcontract gives the contractor the right to charge the subcontractor for “all costs incurred as a result [of the subcontractor’s nonperformance], including reasonable overhead” costs. I believe the meaning of “incurred” is unambiguous, limiting the contractor’s damages to what it actually spent to finish the work. Therefore, Comtrol had the burden, as part of its case-in-chief, to provide the trial court with evidence of the actual costs incurred to finish Traco’s work on the USVC and Weber State projects.
¶ 56 My second concern is that, notwithstanding this default provision in the subcontract, Utah case law requires the use of an “actual cost” measure of damages where one party completes the other party’s performance under a construction contract. See Darger v. Nielsen, 605 P.2d 1223, 1225 (Utah 1979) (“The measure of damages for such a breach is the excess of the cost of completing [the other party’s performance], over what defendant would have paid under the contract.” (emphasis added)); Stangl v. Todd, 554 P.2d 1316, 1320 (Utah 1976) (“The contract breaker should pay the cost of construction and completion in accordance with his contract .... ” (emphasis added)). In my view, the National Construction Averages are irrelevant, unless offered to challenge or establish the reasonableness of the wages paid by Comtrol.2 As long as a party “presents] a prima facie case of its damages, and [the other party] d[oes] not present evidence that [those] charges were unreasonable, or that ... [the project could have been] finished at a lower price,” that party is entitled to receive the actual cost of completion. Darger, 605 P.2d at 1225. This court has also explained that the purpose of awarding damages is to fully compensate a party “for actual losses incurred by evaluating any loss ‘suffered by the most direct, practical and accurate method that can be employed.’ ” Price-Orem Inv. Co. v. Rollins, Brown & Gunnell, Inc., 784 P.2d 475, 478 (Utah Ct.App.1989) (emphasis added) (citation omitted) (quoting Even Odds, Inc. v. Nielson, 22 Utah 2d 49, 448 P.2d 709, 711 (1968)).
¶ 57 Likewise, I find Comtrol’s reliance on Kilpatrick v. Wiley, Rein & Fielding, 2001 UT 107, 37 P.3d 1130, and Terry v. Panek, 631 P.2d 896 (Utah 1981), unpersuasive. Kilpatrick dealt with a new business’s lost profits. See 2001 UT 107, ¶¶ 69-70, 76-77, 37 P.3d 1130. The Kilpatrick court allowed evidence of a “hypothetical stream of profits,” id. ¶ 70, because an actual record was not available. See id. ¶ 76 (“While start-up businesses, such as [this one], lack an actual record of past earnings, which decreases ‘the certainty with which one could predict future profits[,] ... that fact should not automatically preclude new businesses from recovering lost profits_’ ” (emphasis added) (second alteration and omissions in original) (quoting Cook Assocs., Inc. v. Warwick, 664 P.2d 1161, 1166 (Utah 1983))). Similarly, the damages in Terry related to the estimated value of a nonexistent well that was promised by the sellers in a real estate transaction. See 631 P.2d at 897 & nn. 1-2. The Terry court allowed the buyer to testify as to the seller’s previous rough estimates of the value of the well because only such “sparse” evi*584dence was available. Id. at 898.3
¶ 58 A plaintiff “need only [prove damages] with reasonable certainty rather than with absolute precision.” Price-Orem Inv. Co., 784 P.2d at 478. However, in a case such as this, where the costs have already been incurred, Comtrol should have provided the best records available to prove the actual costs it incurred in completing Traco’s work. See, e.g., Mahmood v. Ross, 1999 UT 104, ¶ 19, 990 P.2d 933 (“[L]egal damages serve the important purpose of compensating an injured party for actual injury sustained, so that [it] may be restored, as nearly as possible, to the position [it] was in prior to the injury.” (internal quotation marks omitted) (emphasis added)); Highland Constr. Co. v. Union Pac. R.R., 683 P.2d 1042, 1045 (Utah 1984) (“[D]amage[s] must be established by substantial evidence and not by conjecture .... [and] must be traceable to the wrongs complained of.” (citations omitted) (emphasis added)). Therefore, I would hold that the trial court erred by relying on evidence of 2001 National Construction Averages to calculate Comtrol’s damages.
¶ 59 Based on the foregoing, I would remand for further proceedings on the issue of damages to determine whether Comtrol has met its burden, with testimony and documents related specifically to these projects,4 to prove its damages with “reasonable certainty rather than with absolute precision.” Price-Orem Inv. Co., 784 P.2d at 478. Until the trial court makes further findings on damages, unrelated to the National Construction Averages, the determination of the prevailing party is uncertain. Therefore, I also dissent from the majority’s decision as it relates to the award of attorney fees.
. Comtrol introduced, over Traco's objection, the R.S. Means Company’s Building Construction Cost Data. See Building Construction Cost Data: 2001 Western Edition (Phillip R. Waier ed., 14th ed.2001).
. The trial court attempted to determine “a reasonable fair market value amount to complete Traco’s work,” finding “that Comtrol's claimed back charges to complete Traco's work ... [were] overstated” and "excessive.” Although the trial court correctly considered the reasonableness of Comtrol’s back charges, it should have started its inquiry with the costs that Com-trol actually incurred rather than with the National Construction Averages.
. The decisions from other jurisdictions relied upon by Comtrol are also distinguishable. See ABT Bldg. Prods. Corp. v. National Union Fire Ins. Co., 472 F.3d 99, 109 & n. 13 (4th Cir.2006) (noting that the class action settlement agreement expressly adopted a formula using National Construction Averages to calculate damages); Nationwide Mut. Fire Ins. Co. v. Tomlin, 181 Ga.App. 413, 352 S.E.2d 612, 616-17 (1986) (determining evidence that included National Construction Averages was sufficient to distinguish between recoverable and non-recoverable damages caused by a drop in house’s foundation); Carlisle Corp. v. Medical City Dallas, Ltd., 196 S.W.3d 855, 865, 867 (Tex.App.2006) (noting that where evidence included actual costs incurred to replace roof, National Construction Averages were properly used to establish that costs were reasonable), cert. granted, No. 06-0660, 2007 Tex. LEXIS 407 (May 4, 2007).
. Traco challenges the trial court's order quashing its subpoenas, served on Comtrol during trial, which sought records of the actual hourly wages paid by Comtrol. Because I would hold that it was Comtrol's obligation to introduce those records, I would not reach this issue.