Penncro Associates, Inc. v. Sprint Spectrum, L.P.

GORSUCH, Circuit Judge. Sprint Spectrum, L.P., does not dispute that it breached its contract with its former bill collector, Penncro Associates, Inc. Still, it offers two reasons why, in its view, the district court’s judgment for Penncro in excess of $17 million should be reversed. First, Sprint argues that the parties’ agreement precludes the sort of damages Penncro seeks. In their contract, the parties agreed to forego “consequential damages,” and Sprint urges us to find that the term, as defined by the parties’s agreement, includes any and all “lost profits”— whether flowing directly or consequentially from Sprint’s breach. Because all of Penncro’s claimed damages are lost profits, Sprint argues the district court’s judgment is fatally flawed. Secondly, and alternatively, Sprint contends that Penncro’s damages should be calculated on the basis of the work it was ready and able to perform, rather than on the basis of a fixed monthly fee, as the district court found. For its part, Penncro cross-appeals, arguing that it is entitled to an additional $6.5 million in damages. Penn-cro submits that the district court erred when it found that the company was able, by taking on new work after Sprint’s breach, to avoid losses in this amount. We affirm the district court on all three questions presented to us. While parties to a contract may define their terms as they please — a duck may be a goose — we see no evidence that Sprint’s and Penn-cro’s definition of the term consequential damages was designed to embrace (and thus foreclose the award of) profits lost as a direct result of Sprint’s breach. Likewise, the plain and unambiguous language of the parties’ agreement obliged Penncro to provide Sprint with a fixed amount of available labor capacity, and required Sprint to pay for that capacity, whether utilized or not. Finally, we see no clear error in the district court’s finding that Penncro managed to avoid a modicum of the losses that Sprint’s breach imposed. I Originally, Sprint, a national telecommunications company, handled for itself the not-inconsiderable task of trying to collect from its cell phone customers behind on their monthly bills. Beginning in April 2002, however, it decided to “outsource,” contracting first with Penncro and subsequently with two additional vendors to assume the job. Under the terms of the parties’ agreements, customers with overdue Sprint accounts trying to make outgoing calls were automatically routed to centers run by one of the three vendors, based on which one had the shortest estimated wait time. Penncro’s employees introduced themselves as Sprint’s agents, informed callers that their accounts were past due, and attempted to collect monies owed to Sprint — a service known as first-party inbound collections work. 1 A The nature of the parties’ agreement was spelled out in four interrelated docu *1153 ments: (1) a Master Services Agreement (“MSA”); (2) a Contract Order; (3) an attachment to the Contract Order (“Attachment A”); and (4) an Addendum to Attachment A. The MSA contained certain generic terms and conditions Sprint employed with all vendors. It obligated neither party to perform and expressly indicated that the scope and specific terms of the services provided would be governed by contract orders. MSA § 2.2. The Contract Order was just such a document, involving only Sprint and Penn-cro, and detailing the particular services, staffing levels, and compensation rates attending to the parties’ relationship. Under the terms of the Contract Order, Penncro agreed, among other things, to “maintain staffing levels” sufficient to provide Sprint with “80,625 productive hours” per month. Contract Order § C. A productive hour was defined as time spent by a fully trained Penncro employee handling calls, waiting for calls, training, or waiting due to system downtime. Id. § B. This amounted, more or less, to an agreement to maintain approximately 500 full-time call center employees at Sprint’s disposal. 2 In exchange, Sprint agreed “to pay for 80,625 productive hours per month” at a rate of $22 per hour (less for training hours, more for overtime hours). Id. §§ B, C. 3 The parties’ agreement anticipated a three-year commitment at these levels of service, but also anticipated that the number of “productive hours” could vary according to certain terms specified in Attachment A. In Attachment A, the parties set forth various performance metrics on which Penncro and other vendors were evaluated. Poor performance for three consecutive months could result in a reduction of “the number of productive hours requested by ... 20%.” Attachment A at 3. Six months of consecutive poor performance entitled Sprint to terminate the contract for cause. Id.; Contract Order § D; MSA § 5.3. Sprint emailed the Addendum to Attachment A to Penncro in September 2002, several months into the parties’ relationship. A cover email outlined a number of changes to the parties’ incentive program effected by the Addendum, and the document was included as an email attachment. One change not outlined in the cover email, but made in the Addendum, was the removal of the word “guaranteed” before the reference to the number of productive hours outlined in Section C of the Contract Order — that is, the number of call center hours that Penncro promised to supply and for which Sprint promised to pay. 4 *1154 B The presence or removal of the word “guaranteed” does not seem to have had much bearing on the parties’ performance, which was fraught with difficulty. Penn-cro endured considerable staffing problems and was for many months unable to retain sufficient employees to provide the number of productive hours required by the parties’ contract. The hours Penncro did provide, moreover, were of sufficiently poor quality to rank Penncro last among Sprint’s three vendors in several performance-categories for a number of months. Sprint, meanwhile, did not experience the call volume it had anticipated and so never called on Penncro to provide the contracted-for number of productive hours. From the beginning, Sprint and Penncro discussed the actual number of hours that Sprint needed and Penncro could provide, agreeing, for the time being, to have Penn-cro bill and Sprint pay only for the hours that Penncro actually supplied. At no time did the total hours (supplied, billed, or paid for) reach the 80,625 hours for which the parties had contracted. According to Penncro’s CEO, this arrangement was acceptable — neither party complained or took action under the contract — because both parties had difficulties performing, the contract was in its early stages, and the parties fully expected to meet their obligations over the contract’s three-year term. In September 2002, Sprint emailed Penncro to announce a unilateral reduction in the number of FTEs (and therefore productive hours) due to “lower than expected call volume.” Penncro voiced no objection. As call volumes and the hours necessary to handle them waned, Sprint gave notice of its intent to terminate Penncro’s first-party inbound collections contract by letter dated January 17, 2003. The parties then entered a four-month ramp-down period during which the number of FTEs requested of Penncro was incrementally reduced. The reason Sprint gave for its action was that Penncro was in third place or below in six agreed performance measures for the six months from July 2002 through December 2002 — an event entitling Sprint to terminate the parties’ relationship for cause under Section D of the Contract Order. As the parties were ramping down, and in spite of their prior performance problems, Sprint and Penncro entered into a new contract order for a different but related service — third-party outbound collections work (see supra note 1). Sprint asserted that it would not have given this new work to Penncro if Penncro was still performing under the parties’ initial, first-party inbound contract, as Sprint did not think Penncro capable of handling both tasks. Penncro also entered into two other contracts for third-party outbound collections work during the same time-period, one with AT & T and another with a utility company, American Water. C In November 2004, Penncro sued Sprint for breach of the parties’ first-party inbound collections contract in federal district court, invoking the court’s diversity jurisdiction. Penncro claimed that Sprint was liable for breach of contract as a matter of law because Sprint’s stated reason for termination was erroneous: Penncro was not in last place for the full six consecutive months necessary to justify termination under the parties’ contract. The district court agreed and entered summary judgment for Penncro on the question of liability. For its part, Sprint itself chose to pitch its battle primarily on the field of damages and a three-day trial on damages followed, after which the district court issued a detailed 45-page ruling. During trial, Sprint *1155 pointed to Section 13 of the MSA, which rules out the award of consequential damages and proceeds to define the term as “includ[ing], but ... not limited to, lost profits, lost revenues and lost business opportunities.” This language, Sprint submitted, prohibited either party from obtaining recovery of any and all lost profits. The district court disagreed, holding that the MSA forbids only “consequential” or “indirect” damages that are “beyond direct economic loss or ordinary loss of bargain damages,” thereby doing nothing to rule out of bounds lost profits suffered as a direct consequence of Sprint’s breach. As a fallback argument, Sprint contended that any award of lost profits should be limited to the hours that Penncro’s employees actually worked or could have worked. Again the district court disagreed, holding that, in spite of the difficulties that both sides had in living up to their obligations under the contract, Sprint’s agreement constituted an unambiguous and “unqualified promise to pay for 80,625 productive hours per month without regard to whether it actually called upon Penncro to provide those hours.” Based on these holdings, the district court awarded Penncro $17,136,612 in expectation damages: $53,109,386 in lost contractual revenues (lost profits), minus $28,307,302 in costs avoided by not having to perform and $7,665,472 in losses avoided due to the breach. Penncro contested many of these reductions, especially those concerning avoided losses. The district court found that, solely because of Sprint’s termination which freed up call-center capacity, Penncro was able to mitigate its losses by taking on new work for Sprint (the third-party outbound collections contract), AT & T, and American Water. Penncro disputed the court’s assessment, arguing that it easily could have handled the new work as well as Sprint’s existing first-party inbound collection work. Both parties appeal. We turn first to Sprint’s appeal and then Penncro’s cross-appeal. II The proper construction of a contract is a question of law we review de novo. See In re Villa West Assocs., 146 F.3d 798, 802 (10th Cir.1998); Liggatt v. Employers Mut. Cas. Co., 273 Kan. 915, 46 P.3d 1120, 1125 (2002). Under Kansas law, which the parties agree controls this dispute, see MSA § 17.6, a contract is ambiguous if reading its plain language yields doubtful or conflicting meanings. Liggatt, 46 P.3d at 1125. However, ambiguity will only be found where the meaning is genuinely uncertain, and even then not until the language has been given a reasonable and practical construction in light of the contract as a whole. Id.; Decatur County Feed Yard, Inc. v. Fahey, 266 Kan. 999, 974 P.2d 569, 574 (1999). In determining whether or not a contractual ambiguity exists, moreover, we are obliged under Kansas law to confine ourselves to the four corners of the contract. See Decatur County Feed Yard, Inc., 974 P.2d at 574; City of Wichita v. Sw. Bell Tel. Co., 24 F.3d 1282, 1287 (10th Cir.1994). Only after finding the presence of a contractual ambiguity may we look to extrinsic evidence-including the parties’ course of conduct-to construe the documents. Sw. Bell Tel. Co., 24 F.3d at 1287; Farrell v. Gen. Motors Corp., 249 Kan. 231, 815 P.2d 538, 546 (1991). It is with these standards fixed in mind that we approach the parties’ competing contentions. A 1. Sprint begins its appeal by renewing its argument that Penncro imper-missibly seeks lost profits. Section 13 of the MSA forbids the recovery of “eonse- *1156 quential damages,” specifying that they “include, but are not limited to, lost profits, lost revenues and lost business opportunities.” In Sprint’s parlance, this means that any lost profits are (forbidden) consequential damages. For several reasons, we are persuaded that Sprint’s interpretation is foreclosed by the unambiguous language of the MSA. Section 13’s syntax alone propels us in this direction. The parties’ language is not unlike a doctor’s prescription that “You really should not eat fried foods— and this includes, but is not limited to, meat and potatoes.” Ordinary usage and common experience does not suggest that the patient should avoid all meat and potatoes, but only those that are parts or components of the initial, larger group of fried foods (say, chicken fried steak and french fries). The dictionary underscores the point. Webster’s defines the term “to include” as meaning “to place, list, or rate as a part or component of a whole or of a larger group, class, or aggregate.” Webster’s Third New International Dictionary 1143 (2002). The more general term informs the subsequently listed examples, not the other way around, and so lost profits here refer only to those that are “a part or component” of the larger group or class of consequential damages. The common legal meaning of the terms involved confirms this reading. Direct damages refer to those which the party lost from the contract itself — in other words, the benefit of the bargain — while consequential damages refer to economic harm beyond the immediate scope of the contract. 5 Lost profits, under appropriate circumstances, can be recoverable as a component of either (and both) direct and consequential damages. 6 Thus, for example, if a services contract is breached and the plaintiff anticipated a profit under the contract, those profits would be recoverable as a component of direct, benefit of the bargain damages. If that same breach had the knock-on effect of causing the plaintiff to close its doors, precluding it from performing other work for which it had contracted and from which it expected to make a profit, those lost profits might be recovered as “consequential” to the breach. All of this is by way of saying that, under the circumstances we face here, a reading of Section 13 informed by the normal legal meaning of its terms suggests that it bars only the recovery of consequential lost profits, not direct lost profits. Section 13 says that no consequential damages are recoverable, “in-cludfing]” lost profits; it simply does not speak to direct damages, or to lost profits recoverable under such a theory. A review of Section 12 of the MSA further strengthens our confidence about the parties’ meaning in Section 13. In Section 12, Sprint obligated Penncro to indemnify Spring against “all claims, damages, losses, liabilities, costs, expenses, and reasonable attorney’s fees” arising from claims by third parties for work performed by Penncro. MSA § 12.1. By using the phrase “all ... damages,” the parties manifested a clear intent to insulate Sprint from any damages, direct or indirect, claimed by third parties. When it came to insulating Sprint from liability to Penncro itself, however, the parties in the very next section precluded recovery only of the class of damages that are “consequential [or] indirect.” Id. § 13. When a contract uses different language in proximate and *1157 similar provisions, we commonly understand the provisions to illuminate one another and assume that the parties’ use of different language was intended to convey different meanings. See Decatur County Feed Yard, Inc., 974 P.2d at 574 (before finding ambiguity, court must view contested language in light of the contract as a whole); cf. Sosa v. Alvarez-Machain, 542 U.S. 692, 712 n. 9, 124 S.Ct. 2739, 159 L.Ed.2d 718 (2004) (“[W]hen the legislature uses certain language in one part of the statute and different language in another, the court assumes different meanings were intended.”) (internal quotation and citation omitted); Rogers v. Shanahan, 221 Kan. 221, 565 P.2d 1384, 1386 (1976) (presuming that “the legislature intended a different meaning when it used different language ... in different parts of a statute”). Finally, our understanding of the parties’ chosen language comports with how Kansas courts have interpreted similar terms. In Brennan v. Kunzle, 37 Kan. App.2d 365, 154 P.3d 1094, 1112-13 (2007), for example, a mortgage allowed the plaintiffs to recover expenses incurred to enforce the agreement, “including, but not limited to, reasonable attorneys’ fees.” The court held that this formulation authorized the recovery only of those attorneys’ fees incurred in pursuing remedies under the mortgage and simply did not speak, one way or the other, to other attorney’s fees, such as those incurred in defending counterclaims. The same logic applies here: Section 13 is all about consequential lost profits and does not tell us anything, one way or the other, about direct lost profits. Cf. Dikeman v. Nat’l Educators, Inc., 81 F.3d 949, 953 (10th Cir.1996) (allowing, in a statute where examples of a general term were listed, that the more general term limits the scope of the examples). 7 2. Sprint responds by observing that parties are free to diverge from linguistic and legal norms and to define terms in their contracts as they see fit. This, Sprint contends, is what the parties did here, defining “consequential damages,” to encompass any lost profits, including those normally thought to derive directly from a party’s breach. And, to be sure, parties to a contract are free to define their terms in any manner they wish. Up may be defined as down, right as left, day as night. But, while the parties may depart from the meanings associated with ordinary English and existing law, courts will recognize and give effect to such private definitions only where the parties’ intention to deviate from common usage is manifest. See Cor-bin on Contracts § 24.8; Restatement (Second) of Contracts § 202(3)(a). That is, we are reluctant to shed ordinary meanings and presume unusual private ones without some clear indication that the parties wished us to do so. Here, we are directed to no textual indicia that the parties intended such a departure. Neither do we find Sprint’s proffered authority of much help to its cause. In many of the cases Sprint cites, lost profits were not, as here, placed in an illustrative list of the sorts of consequential damages excluded; instead, they were singled out as a separate and distinct category of forbidden damages. 8 Sprint’s own authority thus il *1158 lustrates how parties easily can manifest an intent to preclude lost profits of any stripe and highlights how Section 13, by contrast, simply failed to do so. 9 B Even if Penncro is entitled to direct lost profits, Sprint submits that it is not obliged under the parties’ agreement to pay for 80,625 productive hours per month when Penncro never provided (or was able to provide) that amount of labor. At the very least, Sprint urges us to find ambiguity in the contract on this point so that we might step outside the four corners of the document and examine extrinsic evidence, including the parties’ course of conduct during the contract’s “start up” period when they agreed to provide and pay for fewer hours of work. Penncro responds that, while its willingness or ability to perform was an element of liability, liability is not contested and performance is irrelevant to the assessment of damages because the contract’s language unambiguously requires Sprint to pay for 80,625 productive hours per month. We are constrained to agree. 1. The relevant contract language provides flatly that Penncro “agrees to maintain staffing levels” and that Sprint agrees “to pay for” 80,625 productive hours per month. Contract Order § C. Sprint’s obligation to pay thus is not conditioned in any way on how many hours Penncro was actually able to provide. Instead, the parties entered into a bilateral contract in which Penncro promised to provide Sprint with, and Sprint promised to pay Penncro for, a fixed amount of labor. See, e.g., Heller v. Martin, 14 Kan.App.2d 48, 782 P.2d 1241, 1243 (1989) (In “a classic bilateral contract!,] • • • each [party] promises future performance in consideration for the other’s promise of future performance.”); cf. People’s Exchange Bank of Elmdale v. Miller, 139 Kan. 3, 29 P.2d 1079, 1081 (1934). We acknowledge the apparent (at first blush, at least) inequity of this reading, as it affords Penncro damages for services it never rendered, and was unable to render. *1159 But a somewhat gentler light is shone on our interpretation when one recognizes that the parties’ contract was one for capacity. Penncro agreed to “maintain staffing levels” at 80,625 productive hours, and a productive hour includes time spent waiting for calls. Contract Order §§ B, C. Neither Penncro’s obligation nor its compensation was dependant on the number of calls, if any, that Sprint actually routed to Penncro. In return, Sprint agreed to pay for this set amount of call center capacity, whether or not Penncro’s phones were ringing. Contract Order § C. The terms of the parties’ incentive program, in Attachment A, confirm the point. As structured, the capacity hours agreed to by the parties could be changed only in the event of poor performance — and in no event by more than twenty percent. By contrast, under Sprint’s reading of the parties’ agreement, Sprint could reduce (or, presumably increase) the number of hours it sought from Penncro simply by making a phone call at the beginning of the month. Such a reading would make the agreement to reduce capacity only in the event of poor performance, and even then only by twenty percent, surplusage. Indeed, the entire incentive program would be meaningless if 80,625 hours were merely a forecast of the parties’ future work — rather than a binding capacity commitment on both sides — that Sprint could alter at its whim. 2. Sprint replies with three arguments. First, it stresses Penncro’s demonstrated inability to provide all the hours envisioned by the contract and the parties’ agreement to proceed in several months on different terms. But, having found the promise to pay for 80,625 hours of capacity to be unambiguous in the language of the contract itself, we are precluded by Kansas law from entertaining Sprint’s extrinsic evidence concerning the parties’ course of conduct. Farrell, 815 P.2d at 546. Of course, not every state is so restrictive on this score, see generally Farnsworth on Contracts § 7.12, but the parties deliberately contracted for application of Kansas law. Furthermore, the MSA explicitly prohibits modification of the contract’s terms except in writing, MSA § 17.15, so the parties themselves specifically agreed that there could be no modification of their obligations through the course of performance. Cf. Riley State Bank of Riley v. Spillman, 242 Kan. 696, 750 P.2d 1024, 1028 (1988) (holding that, where a clause required waiver to be in writing, course of conduct could not give rise to waiver). Of course, Sprint’s arguments about Penncro’s failure to perform its end of the bargain could well have offered a good basis for contesting liability. See, e.g., Fusion, Inc. v. Neb. Aluminum Castings, Inc., 962 F.Supp. 1392, 1395 (D.Kan.1997) (discussing plaintiffs material breach proffered as a defense to a breach of contract claim). But Sprint long ago made the tactical decision not to contest liability on this or any other basis — such as mutual breach, anticipatory breach, mistake, fraud, or material misrepresentation — and its decision to hang its hat at trial solely on the nature and quantum of Penncro’s damages cannot be reconsidered on appeal. See Cortez v. Wal-Mart Stores, Inc., 460 F.3d 1268, 1276 (10th Cir.2005); Wilson v. Muckala, 303 F.3d 1207, 1215 (10th Cir.2002); see also Hill v. Kemp, 478 F.3d 1236, 1250-51 (10th Cir.2007). Second, recognizing the weakness of its appeal to extrinsic evidence, Sprint argues that its obligation to pay only for hours that Penncro actually worked is manifest in the parties’ agreement itself, pointing us to Section 2.2 of the MSA. Section 2.2 indicates that “[t]he execution of a Contract Order ... is [Penncroj’s agreement to provide and Sprint’s agreement to accept and pay for Services and Deliverables *1160 in accordance with this Agreement and the applicable Contract Order.... ” Rather than support Sprint’s position, however, this section merely confirms, as we have already indicated, that the MSA did not obligate either party to supply or purchase services, deferring the nature of the parties’ (possible) future obligations to the Contract Order. As the document itself explains, “Sprint has no obligation to accept and pay for Services or Deliverables that are not set forth in an executed Contract Order.” MSA § 2.2; see also supra Part I.A. And the Contract Order unambiguously reflects Sprint’s promise to pay for 80,625 productive hours. Third, Sprint argues that the Addendum to Attachment A, which removed the term “guaranteed” before the words “productive hours outlined in Section C of the Contract Order” from Attachment A, commands its construction. Specifically, Sprint argues that the removal of the word “guaranteed” shows that, under the agreement in effect at the time of the breach, the number of productive hours were not, if they ever were, guaranteed by Sprint. For its part, the district court refused to credit the removal of “guaranteed.” The court noted that Sprint made this change unilaterally, sending the Addendum as an email attachment to Penncro, and without even noting the alteration in its cover email which, notably, highlighted other changes. See supra Part I.A. But even assuming that the Addendum effectively modified Attachment A to remove the word “guaranteed,” the presence or absence of the word is immaterial to our conclusion that Sprint promised to pay for 80,625 hours of call center capacity. Attachment A, whether or not modified by the Addendum, merely sets out the details of the incentive plan Sprint used to evaluate its vendors. Under its terms, vendors can be rewarded for good performance by monetary bonuses, and they can be punished for poor performance by a reduction in productive hours. The performance plan, however, simply does not speak to the critical question before us — namely, the nature of the “productive hours” agreed to by the parties for alteration under Attachment A’s incentive plan. Put plainly, was “80,625 productive hours,” merely a forecast of demand? Or was it a fixed amount of labor capacity? Attachment A is mute on this dispositive question. Instead, it merely points the reader back to Section C of the Contract Order: In the event of poor performance, Attachment A says that Sprint “may permanently reduce the number of productive hours requested of [Penncro],” and “[t]his reduction will result in a corresponding reduction of the amount of productive hours outlined in Section C of the Contract Order.” Addendum at 3. Thus, under the terms of the incentive plan, the parties could modify Penncro’s promised productive hours outlined in Section C based on its success or failure under certain tests, but it simply does not speak to the nature (capacity v. forecast) of the productive hours promised in Section C. 10 Ill In calculating damages, the district court found that, after Sprint moved to terminate the parties’ first-party inbound collections contract, Penncro managed to avoid $7,665,472 in losses by taking on *1161 work for AT & T and American Water, as well as third-party outbound collections work for Sprint. In its cross-appeal, Penncro does not challenge the district court’s finding that it had sufficient capacity to take on the new Sprint work only by virtue of Sprint’s termination of its initial contract. But Penncro does contest the district court’s conclusion that Penncro’s damages should be reduced by the amounts associated with its AT & T and American Water contracts, $6.5 million in all. In Penncro’s view, it easily could have handled these new jobs in addition to fulfilling its first-party inbound collections work for Sprint — and, thus, the amounts it earned from AT & T and American Water should not qualify as avoided losses. Accordingly, resolution of Penncro’s cross-appeal fairly hinges on the resolution of a single factual question: Could Penncro have taken on this additional work and still performed its initial contract with Sprint? Putting the point differently, was Penncro what contract law calls a lost volume seller? See Bill’s Coal Co., v. Bd. of Pub. Util. of Springfield, Mo., 887 F.2d 242, 245 (10th Cir.1989) (“A lost volume seller is one who has the capacity to perform the contract which was breached as well as other potential contracts,” without resource or capacity constraints.). Whether a party to a contract is a lost volume seller is a question of fact. Rodriguez v. Learjet, Inc., 24 Kan.App.2d 461, 946 P.2d 1010, 1014 (1997). We therefore will not disturb the district court’s factual findings about Penncro’s ability to service AT & T, American Water, and its original Sprint contracts simultaneously unless they are clearly erroneous. For a factual finding to be clearly erroneous, it “must be more than possibly or even probably wrong; the error must be pellucid to any objective observer.” United States v. Cardenas-Alatorre, 485 F.3d 1111, 1118—19 (10th Cir.2007). After extensive proceedings, the district court found that Penncro had the capacity to assume additional collections work only by virtue of Sprint’s contract termination — and, thus, that it was not a lost volume seller. The district court had before it ample evidence to support this conclusion. Unconstrained by the strictures against extrinsic evidence in this inquiry, the district court was able to observe and take full account of Penncro’s continual staffing difficulties and significant performance problems throughout the life of its first-party inbound contract with Sprint. It noted, too, that Penncro solicited and was awarded the AT & T contract expressly on the basis of the telecommunications experience it garnered from having worked for Sprint and on the understanding that those who had performed Sprint’s work would service AT & T’s contract. As to American Water, the district court found that its contract was awarded only after Sprint’s breach and all of the work was performed in the same facility where Sprint’s work had been performed; this allowed Penncro to let its leases at other facilities lapse and to reallocate existing resources to American Water’s work, thereby saving significant start-up and training costs. To be sure, Penncro points to competing evidence and testimony from its employees suggesting that it had adequate capacity, or could have found adequate capacity, to cope with Sprint’s contract on top of the new work it received. But pointing to conflicting evidence inconsistent with the district court’s finding is insufficient, standing alone, to establish clear error, for “every trial is replete with conflicting evidence, and in a bench trial, it is the district court, which enjoys the benefit of live testimony and has the opportunity firsthand to weigh credibility and evidence, that has the task of sorting through and making sense of the parties’ competing narratives.” Watson v. United States, 485 F.3d *1162 1100, 1108 (10th Cir.2007). The district court, moreover, found Penncro’s evidence unpersuasive in several ways we find illuminating. For example, the district court discounted Penncro’s assertions that it had the capacity to handle AT & T’s work at a different facility because that contract had to be performed by the same personnel who handled Sprint’s work. The court also found that Penncro could not have performed the American Water contract at a different facility because Penncro could not have opened and staffed a new facility in time to take on American Water’s work. And the district court found that Penncro’s staffing woes throughout the life of its initial contract with Sprint spoke volumes to its capacity to take on additional work. Though reasonable factfinders might come to different conclusions on the facts associated with Penncro’s cross-appeal, we are persuaded that the district court considered them thoughtfully and came to a well-reasoned result free of clear error. * * * We hold that, in keeping with plain meaning and legal norms, where parties to an agreement exclude liability only for consequential damages, profits lost as a direct result of a breach may be recovered. As to the amount of profits lost, we hold that the parties’ fixed-capacity agreement obliged Sprint to pay for that amount of capacity, whether utilized or not. Finally, we discern no clear error in the district court’s finding in this case that Penncro avoided losses as a result of Sprint’s breach by taking on work from AT & T and American Water. Accordingly, we affirm the district court’s judgment to Penn-cro in the amount of $17,136,612. Sprint’s appeal on the award of attorneys’ fees is dismissed as moot. Footnotes [1] It is first party because Penncro's employees identify themselves as Sprint's agents, and inbound because Sprint's customers' calls are routed to Penncro rather than being initiated by Penncro. This is in contrast with third- *1153 party outbound collections, where Penncro’s employees initiate calls to customers with past-due accounts and introduce themselves as third-party collections agents. See infra p. 1154. [2] Each full time employee ("FTE”), or equivalent, counted as 161.25 productive hours per month. Contract Order § C. The total number of employee-hours (80,625) divided by hours per employee (161.25) equals 500 employees. [3] To oversee the operation, Penncro agreed to provide one supervisor for every 15 FTEs and one manager for every 5 supervisors. Contract Order § C. The 500 FTEs required 33 (33.33) supervisors, who required 7 (6.67) managers, for a total of 40 management-level employees. For these services, Sprint agreed to pay a flat monthly fee of $4,500 per management-level employee. [4] The MSA indicated that Sprint could propose changes to services by providing a change notice (which may, per § 17.1, be via email); if Penncro did not respond within seven days, the changes were deemed effective. Id. § 2.5. The email from Sprint asked for both an email and an initialed hard-copy reply. It appears that Penncro did not, in fact, respond. Sprint Opening Br. at 7 n. 2. [5] See Restatement (Second) of Contracts § 347(a) & cmts. a, c (1981); Black's Law Dictionary 416-17 (8th ed.2004); 24 Samuel Williston & Richard A. Lord, A Treatise on the Law of Contracts § 64.12 (4th ed.1993). [6] See, e.g., Source Direct, Inc. v. Mantell, 19 Kan.App.2d 399, 870 P.2d 686, 693 (1994); Restatement (Second) of Contracts § 347. [7] See also Coremetrics, Inc. v. Atomic Park. com, LLC, 2005 WL 3310093, at *4 (N.D.Cal. Dec. 7, 2005) (“Under the plain reading of this [similarly worded] provision, 'loss of profits’ is referenced only as a subset or species of indirect damages."); In re Ardent, Inc., 305 B.R. 133, 138-39 & n. 3 (Bankr.D.D.C.2003); Combustion Sys. Servs. v. Schuykill Energy Resources, Inc., 1993 WL 496946, at *2-5 (E.D.Pa. Nov. 19, 1993); Canal Elec. Co. v. Westinghouse Elec. Corp., 756 F.Supp. 620, 623 n. 2, 627 (D.Mass.1990). [8] See Vaulting & Cash Servs. v. Diebold, 199 F.3d 440 (5th Cir.1999) (unpub.) (lost profits deemed to be separate category of precluded *1158 damages, as "consequential” modifies only "damages,” not "lost profits”); Imaging Sys. Intern., Inc. v. Magnetic Resonance Plus, Inc., 227 Ga.App. 641, 490 S.E.2d 124, 127 (1997) (lost profits set off from consequential damages by disjunctive "or"); CompuSpa, Inc. v. IBM, 228 F.Supp.2d 613, 626 (D.Md.2002) (same). [9] Sprint’s remaining cases rely on other, entirely independent grounds for denying lost profits that do not usefully inform our analysis. See ALLTEL Info. Setvs., Inc. v. F.D.I.C., 194 F.3d 1036, 1039-40 (9th Cir.1999) (denying lost profits for breach of contract because a federal statute specifically defined direct compensatory damages to exclude lost profits); Continental Holdings, Ltd. v. Leahy, 132 S.W.3d 471, 475, 476-77 (Tex.App.2003) (denying recovery for direct lost profits under contract precluding liability for "loss of profits, loss of business or any other indirect or consequential damages” where a separate early termination provision limited recovery to the time the contract was actually in effect, i.e., only up to its termination, not to the end of the written contract period). The case that perhaps comes closest to supporting Sprint is In re Arbitration Between InterCarbon Bermuda, Ltd. and Caltex Trading and Transport Corporation, 146 F.R.D. 64 (S.D.N.Y.1993). But there, the court never purported to construe the parties' contract de novo, as we must do here, but was merely called on to review an arbitrator's handling of the case for "misconduct” under a very different standard of review (asking merely whether a "party was denied a fundamentally fair hearing”). Id. at 72 (internal quotation and citation omitted). Further, to the extent the court addressed the issue, it summarily concluded, without analysis, that all lost profits are consequential damages. Id. at 73. Even Sprint does not make this claim; Sprint argues that consequential damages include direct lost profits in this case only because of a specific, defining contractual clause. [10] The same problem attends Sprint's argument that the word “requested” in Attachment A evinces an intent only to pay for the capacity actually supplied. Any reference to the hours that Penncro was to provide, whether described as “requested,” "guaranteed,” or otherwise, simply refers the reader back to Section C of the Contract Order, which sets forth the parties’ binding commitment to provide and pay for a fixed amount of capacity.