MARVIN LUMBER AND CEDAR COMPANY MARVIN WINDOWS OF TENNESSEE, INC., — v. PPG INDUSTRIES, INC., — AMERICAN CHEMISTRY COUNCIL, AMICUS ON BEHALF OF MARVIN LUMBER AND CEDAR COMPANY MARVIN WINDOWS OF TENNESSEE, INC., — v. PPG INDUSTRIES, INC., —

PPG Industries, Inc., appeals from the judgment entered in favor of Marvin Lumber and Cedar Company and Marvin Windows of Tennessee, Inc. (collectively, Marvin), on Marvin's claim for breach of express warranty of future performance. Marvin cross appeals. We affirm in part and reverse in part.

This case is before us for the second time. See Marvin Lumber & Cedar Co. v. PPG Indus., Inc., 223 F.3d 873 (8th Cir.2000). As we explained in our first opinion, Marvin is a family-owned company that manufactures, among other things, millwork products — wooden doors and windows. PPG sells wood preservatives and coatings. The genesis of this lawsuit was Marvin's use, from 1985 to 1988, of PPG's wood treatment PILT (preservative in-line treatment) on Marvin's doors and windows. PILT replaced the industry standard in wood preservatives, products containing pentachlorophenol (Penta). Marvin had used Penta products successfully for years until environmental concerns were raised about the active ingredient.

In 1994, Marvin filed this diversity suit seeking damages on a number of legal theories, claiming that PILT had failed to prevent premature rot and decay in Marvin's wood products. The district court dismissed or granted summary judgment to PPG on all counts. On appeal, we affirmed in large part. But we remanded for trial Marvin's claim for breach of an express warranty of future performance, having concluded that there remained genuine issues of material fact on the claim. The warranty arose, Marvin said, from representations by PPG to Marvin employees that wood products treated with PILT would last as long or longer than Penta-treated products. See id. at 879-80.

On remand, at the request of PPG, the four-month jury trial was bifurcated. In part one, the jury found that PPG had given Marvin a warranty of future performance and that such warranty formed part of the basis of the bargain between the parties and was incorporated into their agreement. In the second phase, the jury found the warranty was breached and awarded damages: $53.6 million for out-of-pocket costs, $25.2 million for past lost profits; $27 million for future lost profits; and $30 million for loss of goodwill. The District Court entered judgment for Marvin, awarding $156,118,625.92 for damages (including third-party litigation costs) with interest (not including post-judgment interest). PPG appeals and Marvin cross appeals. We address the issues in the order presented.

For its first issue on appeal, PPG contends that the District Court erred when it granted judgment as a matter of law (JAML) in favor of Marvin on the question of notice.1 Under Minnesota law, after acceptance of goods, a "buyer must within a reasonable time after the buyer discovers or should have discovered any breach notify the seller of breach or be barred from any remedy." Minn.Stat. § 336.2-607(3)(a) (2002).2 We review de novo the District Court's decision to grant JAML to Marvin, applying the same standard as the District Court, that is, Rule 50(a) of the Federal Rules of Civil Procedure. Arabian Agric. Servs. Co. v. Chief Indus., Inc., 309 F.3d 479, 482 (8th Cir.2002). The question for us is: PPG having been "fully heard" on the issue of notice, was there a "legally sufficient evidentiary basis for a reasonable jury to find for" PPG so that granting JAML in Marvin's favor was error? Id. (quoting Fed.R.Civ.P. 50(a)(1)).

PPG claims that Marvin had discovered or should have discovered the breach no later than 1990; that the notice finally given to PPG in April 1993 was inadequate; and that adequate notice was not given until September 1997. According to PPG, whether the notice was given in 1993 or 1997, it was not, as a matter of law, given "within a reasonable time," and so Marvin is not entitled to any remedy for breach. PPG asks that we order judgment for PPG on the question of notice or at the very least remand for a new trial so the issue can be decided by a jury.

Initially, PPG asserts in its brief that "this Court held in Marvin I that a reasonable jury could find that Marvin discovered or should have discovered its claim for breach" in early 1990, suggesting that our comment foreclosed JAML on the question of notice. Brief of Appellant at 23. We did note that fact questions existed on the issue of when Marvin knew or should have known of the breach for purposes of a statute of limitations issue — was suit filed within four years of the time when the alleged breach was known or should have been known?3 But the District Court's conclusion as a matter of law that Marvin gave reasonable notice of breach to PPG is not in any way inconsistent with our determination in Marvin I, on the record before the Court at that time, that a jury could find Marvin knew or should have known of the problems with PILT before April 1990. The language from Marvin I noting that fact questions made summary judgment inappropriate on the statute of limitations question on the record before us in that case does not preclude JAML on the notice issue on the record before us now. The District Court's decision on notice followed a full-blown trial. Marvin, no doubt having read our opinion in Marvin I as carefully as did PPG, presumably worked to tighten up its evidence in hopes of winning a jury verdict — or better yet, JAML — on the question of notice.

It is true that the sufficiency of notice under § 2-607 ordinarily is a question of fact to be determined by a jury. Church of the Nativity of Our Lord v. WatPro, Inc., 491 N.W.2d 1, 5 (Minn.1992), overruled on other grounds, Ly v. Nystrom, 615 N.W.2d 302 (Minn.2000). But if, as the District Court decided, "no reasonable jury could have determined" that Marvin failed to give the requisite notice, it was not error for that court to decide the issue and enter judgment as a matter of law in favor of Marvin on the issue. Order and Memorandum of June 7, 2002, at 3. We now look at the evidence before the court on the question of notice.

Under § 2-607, Marvin's duty to give notice to PPG arose when Marvin discovered or should have discovered the breach — by early 1990, according to PPG. Marvin's position is that it was unaware of the breach until just before it gave PPG notice in April 1993. The undisputed evidence at trial showed that Marvin indeed had some concerns about wood deterioration before 1993 but that company employees did not know that the rot problems correlated with the switch to PILT until 1993. Moreover, given the representations about PILT's superior ability to prevent rot, it was reasonable that Marvin did not come quickly to the realization that the wood preservative might be the cause of the problem until it had explored and ruled out other potential causes — which it had done by 1993. Marvin's duty under the "should have known" standard, once it became aware of excessive rot problems, was to investigate further and determine whether PILT was the problem. This it did.

But PPG cites evidence in the record from which, it says, a reasonable jury would have found (or at the very least, could have found) that Marvin knew or should have known about the breach of warranty by early 1990. First, PPG identifies a Marvin document dated December 12, 1989, that lists thirty-five "anticipated concern[s]," among which is number twenty-nine, "wood rotting complaints," with "wood treatment systems" under the corresponding "comment." Defendant's Exhibit 624. But the document does not indicate that complaints were escalating or that they were primarily in the newer, PILT-treated products. And as we have learned from our review of the trial transcript, wood treatment "systems" would refer not only to the preservative pretreatment, but also to coatings that go on top of the treated wood (e.g., paint) and the method by which the pretreatment and the coatings are applied. In other words, a reasonable jury could not have found that the Marvin employees who saw this document knew or should have known — because of the content of this document — that PILT had failed to live up to PPG's promises regarding its future performance.

PPG also brings to this Court's attention Marvin's "Wood Deterioration Project" of 1990. Defendant's Exhibit 1576. Significantly, the bulk of the report lists the specific millwork products about which complaints had been received and the location of the deterioration within those products. Exterior finish is noted, but type of pretreatment is not. Instead, PPG suggests that Marvin should have known PILT was the problem because the report noted that of the 554 windows and doors with reported deterioration problems, 139 of the units reported upon were manufactured from 1985 to 1988, during which time Marvin was using PILT. But the Wood Deterioration Project is simply a report of raw numbers; no statistical analysis accompanied the project. For example, the numbers do not reflect the actual percentage of production for a given year that was affected with deterioration problems. In truth, the most dramatic correlation, if there is one, is between deterioration and geography: 373 of the reports came from the northeast, more than the other three geographical areas combined. No reasonable jury could find from the results of this project that Marvin knew or should have assumed in 1990 that the use of PILT was causing an increase in wood-rot complaints. The report does suggest that additional inquiry might be in order, and as we have said, Marvin undertook a further investigation.

In addition, PPG points to an internal Marvin memorandum from a territory manager to the customer service manager dated September 14, 1990, that notes rotting problems in its products installed in buildings in southern states. Defendant's Exhibit 170. But the memo says that the problems are "due to sash separation, which then leads to a rotting problem." Id. The proposed corrective measures listed in the memo all relate to the joints and attachments of the sashes; there is no indication that a failure of wood pretreatment, whether Penta or PILT (we cannot tell from the memo when the sashes at issue were manufactured), was or should have been suspected as the root cause of the problem. Indeed, the memo references a wholly unrelated cause — sash separation.

Finally, PPG notes a report sent to a PPG employee from a Marvin employee on July 14, 1993, showing that from 1990 through 1992, deterioration complaints about PILT-treated products had outstripped complaints about the older Penta-treated products. Plaintiffs Exhibit 1586. The undisputed testimony, however, was that numerous factors could have contributed to the rot problems. And it was not until a committee of Marvin employees studied the problem in 1992 and 1993 that a determination was made that PILT was the common denominator in the products about which Marvin was receiving complaints.

"In deciding whether the notice requirement has been complied with, the jury must evaluate `the factual setting of each case and the circumstances of the parties involved.'" WatPro, Inc., 491 N.W.2d at 5 (quoting Wagmeister v. A.H. Robins Co., 64 Ill.App.3d 964, 21 Ill.Dec. 729, 382 N.E.2d 23, 25 (1978)). We must agree with the District Court that, on these facts, Marvin was entitled to JAML on the issue of § 2-607 notice. Even taken together, the documents upon which PPG relies in its argument are not sufficient to show that Marvin knew or should have known that the wood-rot problems it was experiencing beginning in 1989 through 1992 were the result of pretreatment with PILT, and no reasonable jury could so find. While PPG receives the benefit of all reasonable inferences that may be drawn from the evidence, an inference is not reasonable unless it may be drawn "without resort to speculation." Arabian Agric. Servs. Co., 309 F.3d at 482 (citations to quoted cases omitted). And it would be pure speculation for a jury to conclude that the evidence of rot that Marvin had in its possession in 1990 was such that Marvin knew or should have known that PILT was the singular or even primary culprit. There was no conceivable reason for Marvin not to advise PPG as soon as it was confirmed that PILT was the problem. It was in Marvin's best interests to identify the cause, remedy the problem, and seek redress from the responsible party as soon as possible. Shortly after Marvin realized that the pattern of rot complaints it was receiving had only PILT in common, in spite of the promises of superior performance it had received from PPG regarding the wood treatment, it notified PPG — well within a "reasonable" time.

PPG contends that even if we agree with the District Court that Marvin was entitled to JAML because it did not know, nor should it have known, there was a breach of warranty until 1993, the notice given at that time was inadequate because it did not mention specifically the warranty upon which Marvin ultimately prevailed. Marvin did not rely upon an express warranty of future performance for recovery of damages from PPG until it amended its complaint to add the claim in 1997. The 1993 notice therefore failed as a matter of law, according to PPG. But "[t]he notice provision is not intended to `operate as a technical procedural barrier to deny claimants the opportunity to litigate the case on the merits.'" WatPro, 491 N.W.2d at 5 (quoting Prutch v. Ford Motor Co., 618 P.2d 657, 661 (Colo.1980) (en banc)). We think PPG demands more of the notice than Minnesota law requires. See State v. Patten, 416 N.W.2d 168, 172 (Minn.Ct.App.1987) (citing Moosbrugger v. McGraw-Edison Co., 284 Minn. 143, 170 N.W.2d 72, 80 (1969)).

On the subject of § 336.2-607 notice and its adequacy, the Minnesota Supreme Court has said that it "need merely be sufficient to let the seller know that the transaction is still troublesome and must be watched." WatPro, 491 N.W.2d at 5 (quoting Minn.Stat. Ann. § 336.2-607 U.C.C. cmt. 4 (West 1966)). Marvin's 1993 notice to PPG did just that, advising PPG that the common factor in the spate of rot complaints it was receiving was wood pretreatment with PILT. In fact, a December 28, 1993, internal PPG memorandum demonstrates that PPG knew full well that Marvin was claiming warranty, breach, and damages. That document summarized a December 14, 1993, meeting "Regarding Rot Claims" and described Marvin's position as PPG understood it: "In 1984 PILT was sold to Marvin as a superior product and better than anything previously used [the warranty]. Now they have rot problems [the breach] and would like PPG to ante up [the damages]." Plaintiffs Exhibit 1664. As we have said, Marvin had every reason to get notice to PPG as soon as Marvin knew that PILT was the problem. PPG thus cannot claim "commercial bad faith" on the part of Marvin, so Marvin should not, on the basis of the notice given here, be denied its remedy for breach. WatPro, 491 N.W.2d at 5 (quoting Minn.Stat. Ann. § 336.2-607 U.C.C. cmt. 4 (West 1966)).

We affirm the District Court's decision to grant JAML to Marvin on the issue of notice.

PPG next argues that the District Court erred when it declined to enforce a term that appeared in the order acknowledgments for PILT that PPG sent to Marvin. That term purports to limit damages to the price Marvin paid PPG for PILT, $1.6 million.

PPG sent Marvin sixty-six acknowledgments of Marvin purchase orders for PILT. On the face of the form acknowledgment, PPG avers that it accepts the order with the understanding that the only terms and conditions to which it consents are set forth in the acknowledgment. Although there is a line for an "authorized signature" just below this statement, none of the acknowledgments is signed anywhere by anyone. On the back of the acknowledgment, in fine print, in the last item under "TERMS AND CONDITIONS," is this language: "In no event shall Seller's liability for damages in respect to products sold hereunder, or otherwise exceed the purchase price attributable to the specific product as to which a claim is made." E.g., Defendant's Exhibit 7. On its face, this disavowal would seem to limit Marvin's damages as PPG claims.

Under Minnesota law and the U.C.C., a written order acknowledgment from one merchant to another that contains terms in addition to those originally agreed upon by the parties will be considered part of the agreement unless:

(a) The offer expressly limits acceptance to the terms of the offer;

(b) [The new terms] materially alter [the agreement]; or

(c) Notification of objection to them has already been given or is given within a reasonable time after notice of them is received.

Minn.Stat. § 336.2-207(2) (2002). There is no contention here that Marvin's purchase order (its offer) expressly stated that acceptance was limited "to the terms of the offer" (subsection (a)) nor that Marvin ever objected to the limitation of damages in the course of its dealings with PPG (subsection (c)). But as for subsection (b), Marvin contends that the damages limitation was a material alteration to the contract for purchase and sale.4 PPG disagrees, relying on U.C.C. commentary that says "a clause ... limiting remedy in a reasonable manner" (citing U.C.C. § 2-719) involves "no element of unreasonable surprise" and will become a part of the contract unless an objection is raised. Minn.Stat. Ann. § 336.2-207 U.C.C. cmt. 5 (West 1966). The "unreasonable surprise" is a reference to a previous comment wherein the drafters gave examples of clauses that "would normally `materially alter' the contract and so result in surprise or hardship if incorporated without express awareness." Id. cmt. 4. Quoting this Circuit and U.C.C. § 2-207 Comment 4, the Minnesota Court of Appeals has said, "An agreement is materially altered if an addition would `result in surprise or hardship if incorporated without express awareness by the other party.'" TRWL Fin. Establishment v. Select Int'l, Inc., 527 N.W.2d 573, 579 (Minn.Ct.App.1995) (quoting N & D Fashions, Inc. v. DHJ Indus., Inc., 548 F.2d 722, 726 (8th Cir.1976) (quoting Comment 4 to U.C.C. § 2-207)).5 Likewise, in applying § 2-207, we have said, "Considerations of surprise and hardship must remain a part of the [material alteration] analysis." Shur-Value Stamps, Inc. v. Phillips Petroleum Co., 50 F.3d 592, 599 (8th Cir.1995) (applying Texas law); see Johnson v. Murray, 648 N.W.2d 664, 670 (Minn.2002) ("Uniform laws are interpreted to effect their general purpose to make uniform the laws of those states that enact them. Accordingly, we give great weight to other states' interpretations of a uniform law." (citation omitted)).6

Before we move on to our legal analysis, we must address the matter of our standard of review. The Minnesota Court of Appeals has said that "[w]hether an additional term materially alters an agreement is a question of fact which must be resolved on a case by case basis." TRWL Fin. Establishment, 527 N.W.2d at 579 (citing N & D Fashions, 548 F.2d at 726). In such a case, we would review the District Court's finding of material alteration only for clear error. But the District Court itself concluded that the damages limitation term was a material alteration "as a matter of law." Order & Memorandum of June 7, 2002, at 5. If that is the case, we should review the decision de novo, as PPG acknowledges in its main brief. See Salve Regina Coll. v. Russell, 499 U.S. 225, 231, 111 S.Ct. 1217, 113 L.Ed.2d 190 (1991) ("[A] court of appeals should review de novo a district court's determination of state law."). In its reply brief, PPG cites TRWL for the first time for the proposition that the question of material alteration is one of fact but then makes no effort to clarify our standard of review. Further, PPG does not challenge on appeal the District Court's having usurped the role of the jury, if indeed TRWL is correct. Marvin ignores the standard of review altogether. We decline to sort out the difficulties of ascertaining our standard of review because it is not necessary for us to do so. The result we reach would be the same if we reviewed the decision for clear error as a question of fact or if we applied de novo review as though the decision were a conclusion of law.

The facts of this case concerning material alteration, that is, those which go to prove hardship or surprise, are really without dispute. After considering them, we conclude that enforcement of the damages limitation provision would result in both hardship and surprise to Marvin.

Economically, the hardship to Marvin is clear — the $1.6 million PPG says it owes under the terms of the acknowledgment is a fraction of the out-of-pocket costs to Marvin resulting from PILT's failure, not even considering any consequential damages. Under the terms of the boilerplate limitation, the distribution between Marvin and PPG of risk of the product's failure is dramatically altered, with Marvin bearing the brunt of it. The warranty that the jury found was for future performance. The failure of the product in the future — after it was applied to the wood and the wood was made into a window and installed in a building — would be, and was, catastrophic in terms of damages. Given that the warranty was for PILT's future performance, the damages should remedy a failure of that future performance, not just refund the money spent for the product. Reimbursing the cost of the product — essentially "replacing" the product — would not enforce the warranty that PPG gave to Marvin. Moreover, most of the damages at issue were unquestionably foreseeable.7

As for surprise, the provision was not negotiated by PPG and Marvin. While it appeared in all the acknowledgments, it was boilerplate language, in small print, on the back side of a form that was dense with other small print. Cf. TRWL Fin. Establishment, 527 N.W.2d at 579 ("If the forum selection clause is contained in the boilerplate language of a confirmatory memorandum, and is not bargained for, this presumption [that a party adversely affected by the clause has received consideration at the time of contracting] is untenable."). Marvin witnesses testified concerning their surprise that a damages limitation clause may have become part of the contract with PPG. While this testimony is self-serving, it was uncontradicted and not without support elsewhere in the record. Marvin also proffered evidence that PPG and other coatings manufacturers had always made things right when a product of theirs failed with consequences, not just refunding the purchase price of the product sold but replacing the Marvin product that was damaged as a result of the failure. Granted, the catastrophic nature of PILT's failure was new to the relationship between PPG and Marvin, but that would be all the more reason that PPG's refusal to pay consequential damages would be both a surprise and a hardship.

We have said that "Uniform Commercial Code remedies should be liberally administered." Soo Line R.R. v. Fruehauf Corp., 547 F.2d 1365, 1373 (8th Cir.1977) (applying Minnesota's § 2-719). It is true that Comment 5 to U.C.C. § 2-207 lists "a clause ... limiting remedy in a reasonable manner" as one that will "involve no element of unreasonable surprise" and therefore will "be incorporated in the contract." Apart from the fact that the commentary does not have the force of law, we conclude on the facts of this case that the provision in reality did involve "an element of unreasonable surprise" to Marvin. The damages limitation language from the PPG acknowledgment forms therefore did not become a part of the contract between PPG and Marvin and is not enforceable.

For its next issue, PPG challenges the awards to Marvin of goodwill damages ($30 million) and damages for past and future lost profits ($25.2 million and $27 million, respectively).

In arguing that the damages for lost goodwill should be vacated, PPG raises three points: the evidence was insufficient to sustain the award, the award duplicated the damages for future lost profits, and the claim was submitted to the jury without proper notice to PPG. We are somewhat troubled by the short notice given to PPG that Marvin was seeking damages specifically for lost goodwill and that the claim would be submitted to the jury in addition to a claim for future lost profits. But we reverse because we determine that the record cannot support awards for both $30 million in lost goodwill and $27 million for future lost profits. The fact is, the evidence supporting an award of goodwill damages to which Marvin directs our attention is the very same evidence used to support an award of damages for future lost profits. We conclude that this evidence warranted an instruction on, and ultimately an award of, damages for future lost profits, as we will explain, but not an instruction on (or an award of) damages for lost goodwill in addition.

We review de novo the District Court's decision to submit an instruction on lost goodwill. Porous Media Corp. v. Midland Brake, Inc., 220 F.3d 954, 961 (8th Cir.2000). Because we conclude that the court erred in giving the instruction, we will not consider sufficiency of the evidence per se. Nonetheless, our analysis will entail a review of the evidence that Marvin claims supports the award. See C.L. Maddox, Inc. v. Benham Group, Inc., 88 F.3d 592, 601, 602 (8th Cir.1996) (noting that certainty of damages and sufficiency of evidence are "very closely related issues" but "analytically distinct").

Marvin presented evidence of the importance to a wooden window and door manufacturer of a good reputation among distributors, builders, homeowners, and others. The undisputed testimony was that damage to reputation would be a serious problem. Before the problems with PILT, Marvin had an excellent reputation in the industry. But when the volume of rot problems became so great that Marvin was unable to continue to replace windows and repair incidental damage that resulted from the deteriorated millwork, customers became "quite hostile," even threatening lawsuits, according to Gary Daniels, a Marvin employee who handled PILT-related complaints. Transcript at 5223. President Susan Marvin testified to spending "hours on the phone with customers who are so angry and so distressed." Id. at 7333. Susan attributed stagnant growth at Marvin after the PILT problems "to homeowners telling friends and neighbors and associates that the product isn't good." Id. at 7462. She said, "Our reputation has been seriously damaged. And we have lost a lot of loyal customers." Id. As for specifics, Susan testified that an institutional customer, having become aware of the problems others were having with Marvin millwork products, required Marvin to purchase a performance bond before the customer would proceed with a project, already underway, to replace existing windows with Marvin windows.

On the basis of this evidence, the jury was instructed that it could consider whether Marvin "suffered any measurable loss to its goodwill" and award damages for such loss. Id. at 10846. The court instructed the jury that "[t]he measure of Marvin's damage is the difference between such goodwill before and after its experience with wood rot." Id. at 10847. The problem we see with Marvin's claim is in the quantification of goodwill damages, translating angry or cautious customers into dollars and cents. While we are not suggesting that mathematical certainty is required, the jury's decision that Marvin suffered a $30 million loss of goodwill was conjecture. Marvin does not — indeed, it cannot — point to any evidence in the record separate and apart from its evidence of future lost profits to support an award of $30 million in lost goodwill.8 Marvin's trial counsel admitted as much, suggesting to the jury that goodwill damages are "something that you just have to go into your tummy and get." Id. at 10814. It was goodwill, not the damages for a loss thereof, that was described for the jury as an "intangible business value." Id. at 10846. The damages must be measurable. The problem is not so much that lost goodwill and future lost profits are duplicative in this case (in fact, the jury was instructed that they are different from each other). The problem is that Marvin did not make a submissible case of lost goodwill damages. There was no testimony that any Marvin entity was worth less because of the problems with PILT and certainly no evidence giving a measure of, or a way of measuring, how much less. A loss of goodwill resulting in a quantifiable diminution in the value of Marvin's millwork businesses was not shown with sufficient certainty, as a matter of law, to have been submitted to the jury. The award of goodwill damages must be vacated.

As for lost profits, past and future, PPG also challenges the sufficiency of the evidence to support those awards. PPG maintains there was only a scintilla of evidence of past lost profits and no evidence of lost profits from 2002 through 2004. On appellate review, when the issue is sufficiency, we view the evidence in the light most favorable to the verdict and will reverse only if a reasonable jury could not have found as this jury did. United States v. Larry Reed & Sons P'ship, 280 F.3d 1212, 1214 (8th Cir.2002). Consequential damages such as the lost profits in question must be proved "with a reasonable degree of certainty and exactness." County of Blue Earth v. Wingen, 684 N.W.2d 919, 924 (Minn.Ct.App.2004) (citation to quoted case omitted). For future losses specifically, Marvin cannot collect damages that are "remote, speculative, or conjectural." Pietrzak v. Eggen, 295 N.W.2d 504, 507 (Minn.1980). Absolute exactitude is not required, however. "Instead, the plaintiff must prove the reasonable certainty of future damages by a fair preponderance of the evidence." Id. We think the record evidence satisfies these requirements.

We have carefully reviewed the record (no small task in this case). There most definitely was evidence that Marvin's profits had suffered since the problems with PILT began to surface, and that they likely would remain flat. PPG's quarrel with the sufficiency of that evidence appears to have its roots in the fact that the evidence came in largely through the testimony of Marvin senior executives Susan Marvin and Jake Marvin, who based their testimony on their review of the company's financial records.9 PPG also complains that the witnesses focused only on selected years, instead of all years, resulting in skewed calculations of Marvin's projected growth but for the PILT problems. But PPG had the opportunity to — and did — cross-examine Marvin's witnesses on these points, using the demonstrative exhibit prepared by Marvin that noted profits for all the years in question. To the extent that PPG is challenging the court's admission of the evidence in question, we see no abuse of discretion. And we think the evidence before the jury proved lost profits, past and future, and the amounts thereof, with the required degree of certainty. The jury heard it all and determined an award was in order. We cannot say that no reasonable juror could have found that Marvin's damages from the failure of PILT to work as warranted included $52.2 million in past and future lost profits.

PPG contends that Marvin's evidence of breach and causation was not admissible under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). See Kumho Tire Co. v. Carmichael, 526 U.S. 137, 141, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999) ("We conclude that Daubert's general holding — setting forth the trial judge's general `gatekeeping' obligation — applies not only to [expert] testimony based on `scientific' knowledge, but also to [expert] testimony based on `technical' and `other specialized' knowledge." (citing Fed.R.Evid. 702)). According to PPG, the District Court failed in its gatekeeping role when it allowed testimony regarding the results of statistical studies that were conducted by one of Marvin's experts, Dr. Frank Martin, which PPG claims was the only evidence of breach and causation.

PPG's arguments are drawn largely from purported deviations from the Federal Judicial Center's Reference Manual on Scientific Evidence 2d (2000). It goes without saying that the Manual does not have the force of law, nor are judges required to follow it. The reference guides therein, including the one on statistics, "are not intended to instruct judges concerning what evidence should be admissible or to establish minimum standards for acceptable scientific testimony." Reference Manual at vi. We do not review Martin's adherence vel non to the standards recommended in the guide. Our only question is whether the District Court abused its discretion in allowing Martin to testify about the results of his studies. Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997) (standard of review). That is, was the evidence sufficiently relevant and reliable to be put before the jury. See Kumho Tire, 526 U.S. at 152, 119 S.Ct. 1167 ("The objective of [Daubert's gatekeeping] requirement is to ensure the reliability and relevancy of expert testimony.").

Martin was a statistician who was hired by Marvin in 1994 to evaluate the mounting wood-rot data Marvin was collecting. Over time, he conducted a series of statistical studies analyzing the information. He was permitted to testify to two conclusions based on those studies: first, PILT-treated windows did not outlast Penta-treated windows, and second, the excessive wood-rot problems that Marvin was experiencing in its products were the result of the failure of PILT as a wood preservative.

PPG challenges Martin's "methodology" and the "reliability" of the studies' results, invoking these catchwords relating to the admissibility of expert opinion testimony under Rule 702 of the Federal Rules of Evidence. See, e.g., Bonner v. ISP Techs., Inc., 259 F.3d 924, 929 (8th Cir.2001). Specifically, PPG complains that the studies were done in anticipation of litigation; that data was collected by an employee o