Federal Circuit Eliminates Pre-Suit Damages for Failure of Licensee to Mark

by Dennis Crouch

This is an important case in a few ways.  The biggest holding, I think, is the court’s statements that “method patents” cannot be used to “circumvent” the marking requirement of Section 287.  In addition, although not covered in depth here, the software claims were found patent eligible. 

Packet Intelligence LLC v. NetScout Systems, Inc. (Fed. Cir. 2020)

A jury sided with the patentee Packet Intelligence–finding its three patents willfully infringed by NetScout and the claims not-invalid.  The jury also awarded the patentee $5.75 million in a running royalty. US6665725; US6839751; US6954789. Judge Gilstrap topped the awarded with extra $2.8 million in enhanced damages and also denied a variety of post-verdict JMOL motions.  On appeal, the Federal Circuit has affirmed most aspects of the decision, but has reversed the pre-suit portion of the damage award.

Pre-Suit Damages and Patent Marking:

35 U.S.C. § 287(a) provides a limitation on damages in patent cases.  If authorized patented articles are being sold or imported, they should be marked as patented (with the patent numbers). If not marked, then the statute bars any damages from before “the infringer was notified of the infringement.”  The statute was amended as part of the AIA to allow for “virtual marking” that has now become common.

Although Packet Intelligence doesn’t make any products itself, it has licensed its patents, and the marking requirement extends to authorized (i.e., licensed) products made under the patent.

In Arctic Cat Inc. v. Bombardier Recreational Prods. Inc., 876 F.3d 1350 (Fed. Cir. 2017), the Federal Circuit endorsed a burden-shifting process associated with patent marking:

  1. The infringer has the initial burden of production to provide evidence of unmarked products. “[T]his is a low bar” and one that would seemingly be confronted as summary judgment. Arctic Cat.
  2. However, if the case gets to trial, the patentee then has the burden of proving that it complied with Section 287 (either showing that the products were marked or they didn’t need to be marked).

Arctic Cat was decided after the jury verdict in this case and the district court followed a different approach — placing the entire burden on the defendant of proving the existence of a licensed product.  The question of marking was given to the jury with instructions for the jury to award six-years of back damages unless NetScout “show[s] the existence of a … a licensed product that practices one or more claims of the ’789 patent.” (Note below, NetScout did not object to this portion of the verdict).

On appeal, the Federal Circuit reversed the pre-suit damage award — finding that NetScout had met its burden-of-production by identifying a product that arguably embodied the invention and that the patentee had not proven that the product “does not practice the []patent.”

“Method Patents” and Patent Marking:  The claims in two of the patents were all method claims — not subject to the marking requirement which focuses on “patented articles.”  However, on appeal, the Federal Circuit also found that no substantial evidence supported the jury verdict for back damages of those method claims.

Packet Intelligence cannot simply count sales of the software accused of infringing the ’789 patent as sales of the method claimed in the ’725 and ’751 patents. Instead, Packet Intelligence was required to produce evidence that the claimed method was actually used and hence infringed.

Apparently some evidence was produced of infringement-by-use, but it was all regarding internal-use, sales, and training — and there is “no evidence supporting damages caused by or resulting from these pre-suit activities.”  Instead, all of the damages calculations focused on sales.

The court went-on to hold that it would be improper to allow the patentee to “circumvent § 287” by capturing sales of an unmarked product with its method claims:

Packet Intelligence is barred from recovering damages for pre-suit sales of the [accused] products because it failed to comply with the marking requirement. It cannot circumvent § 287 and include those products in its royalty base simply by arguing that NetScout’s infringement of related method claims drove sales. Because neither the record nor the law supports Packet Intelligence’s recovery of pre-suit damages for any of the asserted patents, NetScout is entitled to judgment as a matter of law on this issue.

Id. In the end, the court found that the infringer was entitled to judgment as a matter of law cancelling the pre-suit damage award.  I would have thought that a new trial would be preferred because of the faulty jury instructions.

= = = =

The jury instructions were in error because they placed the burden on the infringer of proving no-marking rather than on the patentee of proving that it properly marked.  However, the infringer (NetScout) did not object to those particular instructions.  Failure to object normally means no appeal, unless there is a “plain error” in the instructions.  Fed. R. Civ. Pro. R. 51(d). In this case, the court did not find “plain error” but instead pursued an alternative approach on appeal.

According to the court, jury instructions do not become “law of the case” for evaluating sufficiency of the evidence. Rather, the appellate panel will not look to see whether the jury followed the instructions as presented but rather they should have been presented:

We are bound by the law, not by the jury charge, even if the charge was not objected to. Markman v. Westview Instruments, Inc., 52 F.3d 967, 975 n.5 (Fed. Cir. 1995) (en banc). And NetScout’s failure to object to the district court’s jury instruction does not render the instruction law of the case for evaluating the sufficiency of the evidence. Boyle v. United Techs. Corp., 487 U.S. 500, 514 (1988) (citing City of St. Louis v. Praprotnik, 485 U.S. 112, 120 (1988) (plurality opinion)).

I’ll note that on this point that the Federal Circuit cites its own law and not 5th Circuit as it did for other procedural matters in the case. One reason for this is that the Federal Circuit’s approach is not in the majority. See:

When a party fails to object to instructions in the trial court, those instructions normally become the law of the case, and appellate review is precluded unless plain or fundamental error can be found.

Corpus Juris Secundum, 89 C.J.S. Trial § 876 (2020). A number of sources disagree on this front.

=  == = =

Finally I’ll note that Judge Reyna dissented-in-part. He would have found the asserted claims ineligible under Section 101:

In my view, the claims are directed to the abstract idea of identifying data packets as belonging to “conversational flows” rather than discrete “connection flows.” While the claimed implementations of this idea may ultimately contain inventive concepts that save the claims, it was clear error for the district court to base its finding of inventiveness on the abstract idea itself and its attendant benefits.

Slip Op.  The majority disagreed — holding that “The claim solves a technological problem by identifying and refining a conversational flow such that different connection flows can be associated with each other and ultimately with an underlying application or protocol.”

7 thoughts on “Federal Circuit Eliminates Pre-Suit Damages for Failure of Licensee to Mark

  1. 1

    I recall an earlier Fed. Cir. decision in which the patent owner circumvented the marking statute by only suing on the method claims even though the patent also had apparatus claims. Was it cited and distinguished here?

      1. 1.1.1

        Indeed, it looks like this decision does not cite prior Fed. Cir. marking cases? But I might be confusing the above with another Fed. Cir. decision [was it Rembrant?] which you discussed in an earlier blog on marking in which the patent owner removed method claims from the patent by disclaimer after the marking defense was raised? As I recall, that did not work.
        BTW, I have been told that some patent owners have granted “covenants not to sue” instead of normal licenses to avoid having to try to contractually force their product selling licensees to mark when they resist that? Does that really work?
        The marking statute is a fascinating patent law subject that seems to sometimes come as a surprise to patent owner’s counsel when [as required] it is affirmatively pleaded in the answer to the compliant. Of course it is not of any concern to the many PAE lawsuits nowadays, since almost all are on patents never used or licensed by any patent owner. I.e., illogically under the marking statute, only patent owners who make and sell and/or license real products get reduced infringement damages, while “paper patents” get more.



          It does seem odd at first blush, but it makes sense if you consider that reducing damages isn’t actually the end goal, but only a means to ensuring that anyone who encounters an article manufactured under a patent knows that they cannot just copy it without risking patent infringement. This would be really useful upon seeing a new type of plow, so you realize that you can’t just go ask your blacksmith to make it. It can perhaps lead to absurd results in the modern economy.

          Incidentally, the original patent marking statute included a statutory penalty which any person could sue to recover, rather than a limitation on damages. See Act of August 29, 1842, ch. 263, §§ 5-6, 5 Star. 543, 544. See also Wagner v. Corn Prods. Ref. Co., 28 F.2d 617, 618 (D.NJ. 1928). This was changed to the current structure in the Act of March 2, 1861, oh. 88, § 13, 12 Stat. 246, 249.



            The reason for the change has a rather twisted reflection in the AIA itself. The view that ‘just anybody’ might be able to ‘cash in’ was peddled as a ‘bad thing,’ at the same time that allowing ‘just anybody’ to challenge a granted patent in the non-Article III forum afforded by the first part of the post-grant review proceedings was peddled as a ‘good thing.’

            Enforcement of patent protections… ‘bad.’
            Destruction of granted patents… ‘good.’

            Connect the dots.


            Yes, the marking statute makes partial sense in that the patent owner can collect more [pre-suit] infringement damages from someone copying their marked product, because that is someone more likely to be doing reckless or willful infringement. But, as noted, only those patent owners who make, advertise and sell and/or license real products [orders of magnitude more capital investment than just getting a patent] get punished [and their infringers pay less] under the marking statute, even if their product was never seen or copied.


              Can we separate the notion of virtue signaling in the “but they make product” comments from the actual impetus for patents?

              Certainly, the ‘positive’ action of making products is ‘nice’ (and may even lead to further innovation), but let’s be abundantly clear that having a patent system is NOT about that ‘positive’ aspect and that the patent right is a ‘negative’ right.

Comments are closed.