[00:00:50] Speaker 00:
Okay, the next case, next target case is number 14, 1668, CNS Media Research against TiVo Research and Analytics.

[00:01:00] Speaker 00:
Mr. Goldberg, are you ready?

[00:01:07] Speaker 03:
Good morning, and may it please the court, and thank you for indulging my cap, which I need for the glare.

[00:01:14] Speaker 00:
Okay.

[00:01:18] Speaker 03:
The Supreme Court last year took the step of deciding an opinion regarding an affirmance of summary judgment and reiterated the fundamental bedrock principle that a judge's function at summary judgment is not to weigh the evidence and determine the truth of the matter, but to determine whether there's a genuine issue for trial.

[00:01:44] Speaker 03:
That was in the Toland versus Cotton case.

[00:01:48] Speaker 03:
This principle cuts across all the issues in this appeal.

[00:01:53] Speaker 03:
TRA is appealing from the district court's summary judgment against TRA on all of TRA's claims.

[00:02:03] Speaker 04:
Let me ask you about the stipulation as it relates to the purchase data.

[00:02:10] Speaker 04:
What did you understand the stipulation to mean when you entered into that stipulation?

[00:02:17] Speaker 03:
there was absolutely no suggestion at the time that that would be a basis for non-infringement.

[00:02:24] Speaker 03:
And to the contrary, at every moment up until that stipulation, it had been the position of Kantar that their products satisfied the purchase data limitation.

[00:02:37] Speaker 03:
It was an unbelievable gotcha when they took the position that the stipulated construction meant

[00:02:46] Speaker 03:
that their products did not infringe.

[00:02:48] Speaker 04:
Well, didn't you question why you were putting that language into the stipulation when the language doesn't appear in the claim anywhere?

[00:02:55] Speaker 03:
The language frankly seemed innocuous.

[00:02:59] Speaker 03:
It was also entered into by my predecessor counsel.

[00:03:03] Speaker 03:
But that language, in terms of what the purchase data meant, data about when a particular product is purchased and that it's purchased at a given time

[00:03:16] Speaker 03:
That language, as you say, is not in the claim.

[00:03:20] Speaker 03:
And for that language to now be further construed to mean anything specific as to the time element,

[00:03:31] Speaker 03:
is contrary to what's set forth in the patent.

[00:03:33] Speaker 04:
Well, did your predecessor counsel ever argue to the district court that the district court shouldn't be construing stipulated language, that district court should attempt to determine what the meeting of the minds was as it relates to that stipulation?

[00:03:48] Speaker 03:
At the time of the stipulation, there was not discussion about it.

[00:03:51] Speaker 03:
It was the party stipulated to it.

[00:03:53] Speaker 03:
And then when Kantar moved for summary judgment of non-infringement,

[00:03:59] Speaker 03:
What they said was their product discloses nothing whatsoever about time.

[00:04:07] Speaker 03:
In other words, they said their product does not even tell you what quarter a purchase was made.

[00:04:15] Speaker 03:
And so what they said, had summary judgment, was they weren't arguing that a given time means you have to know the precise date.

[00:04:25] Speaker 03:
What they were saying was even if a quarter

[00:04:29] Speaker 03:
would be a given time.

[00:04:31] Speaker 03:
They don't infringe because their product doesn't tell you the quarter.

[00:04:35] Speaker 03:
We submitted a mountain of evidence to show that their products disclose the quarter, at a minimum, the quarter in which a purchase was made.

[00:04:45] Speaker 03:
And so the issue that was presented to the district court was whether their product discloses quarterly data.

[00:04:52] Speaker 04:
They say it doesn't, we say it does.

[00:04:54] Speaker 04:
When you believe the district court surprised you by construing this stipulated language to have a very specific meaning, did you go back to the district court and say, Judge, you can't do that without having a Markman hearing or without making a determination as to what the court has understood the stipulation to mean?

[00:05:12] Speaker 03:
When I said there was a gotcha, I didn't mean that we were surprised by what the district court said.

[00:05:17] Speaker 03:
And the district court did not say you have to know the precise date.

[00:05:22] Speaker 01:
The district court agreed with Kantar.

[00:05:27] Speaker 01:
Why shouldn't we interpret the district court to have said that?

[00:05:30] Speaker 03:
The order says that Kantar's products disclose nothing at all about the time.

[00:05:37] Speaker 04:
But it also had a very clear statement about what she understood the time frame within these patents to be, which was a very tight time frame, right?

[00:05:49] Speaker 04:
Like the very point in time.

[00:05:51] Speaker 03:
To the extent that that's what the court decided, and that was not my understanding.

[00:05:56] Speaker 03:
My understanding is that it was based on what Cantor argued, which was that the products disclosed nothing at all about time.

[00:06:02] Speaker 01:
Where you're at now is we've come down to a claim instruction issue.

[00:06:06] Speaker 01:
Right.

[00:06:06] Speaker 01:
Because let's assume that I construed the claim, and I agree with your adversary, that there is a temporal requirement in the claims, which is the first question.

[00:06:16] Speaker 01:
And then I decided, well, and they're right.

[00:06:18] Speaker 01:
The purchase data has to be exactly specific to a single purchase transaction date and event.

[00:06:25] Speaker 01:
If that's the case, then summary judgment was properly entered.

[00:06:29] Speaker 01:
Your honor.

[00:06:30] Speaker 01:
So how do I decide this problem of claim construction?

[00:06:36] Speaker 01:
There hasn't been a Markman hearing.

[00:06:38] Speaker 01:
The parties haven't briefed back and forth what the patent tells us about the temporal restrictions.

[00:06:47] Speaker 01:
You argued, before someone judge them, that the patents to define purchase large data and inquiry, yada, yada, yada, would be broad enough to include court dates.

[00:07:01] Speaker 01:
But there hasn't been an adjudication on the claim construction issue.

[00:07:07] Speaker 03:
Your Honor, the claim construction of purchase data was stipulated

[00:07:14] Speaker 01:
And so the issue about what that... Well, what does the claim construction purchase data is circular because it brings in the temporal limitation that talks about at a given time.

[00:07:25] Speaker 01:
And there's a massiveness of room for things at a given time.

[00:07:30] Speaker 03:
Right.

[00:07:30] Speaker 03:
So I guess I have two comments I'd like to make about that.

[00:07:33] Speaker 03:
One of them is when you have a construction, my understanding is that once the claim has been construed, the issue of whether the product

[00:07:42] Speaker 03:
falls within that limitation is something for the jury.

[00:07:46] Speaker 01:
But we just went through a minute ago.

[00:07:49] Speaker 01:
If claim construction feels that purchase data is limited to a specific transaction on a specific date on a specific product, then you lose some of the judgment because the accused product don't have that.

[00:08:03] Speaker 01:
Right?

[00:08:04] Speaker 01:
So the claim construction issue drive is a bottom line result on the non-automotive product.

[00:08:10] Speaker 01:
And my question is, how do I know?

[00:08:13] Speaker 01:
For example, can you point to me in the patent the language that convinces me that quarterly data is sufficient for purposes of a temporal restriction?

[00:08:28] Speaker 01:
The best I can find in the patent, I'll tell you after you tell me what it is, where is the definition of purchase data?

[00:08:37] Speaker 01:
That will include quarterly data.

[00:08:46] Speaker 01:
I'm looking at the 940, but you can look at 891.

[00:08:49] Speaker 01:
There are three patents, and they're essentially the same in terms of specification.

[00:09:07] Speaker 03:
I guess the point I'm trying to make, Your Honor, and I will try to address your question.

[00:09:13] Speaker 03:
On our opening brief on page 23, footnote five, I think really cuts to the heart of it.

[00:09:19] Speaker 03:
And I would urge the court to look at that particular footnote.

[00:09:24] Speaker 03:
And it is a very, I agree, it's a challenging issue.

[00:09:29] Speaker 03:
We have, it's nested claim constructions, and there's construction of the construction, and it gets very messy.

[00:09:38] Speaker 03:
We come into issues about notice and waiver.

[00:09:43] Speaker 03:
And what was the argument that Kantar was putting forward?

[00:09:45] Speaker 03:
And as we explained it, but no.

[00:09:47] Speaker 01:
Well, this is your whole argument that they conceded this over and it changed their mind.

[00:09:50] Speaker 01:
Let's throw that aside.

[00:09:51] Speaker 03:
Well, that's our judicial estoppel argument, which is different.

[00:09:55] Speaker 01:
Right.

[00:09:55] Speaker 01:
But I come back to the problem, which is I don't know.

[00:10:00] Speaker 01:
Having spent quite a lot of time on this, I think there probably is a temporal requirement with regard to purchase data.

[00:10:10] Speaker 01:
Just generally speaking, there has to be some relationship of time to a purchase to make this whole patent make any sense.

[00:10:18] Speaker 01:
But I'm trying then to decide, what is the nature of the temporal requirement?

[00:10:23] Speaker 01:
Is anything having to do with time, qualifying, quarterly, semi-annually, or is it more specific?

[00:10:33] Speaker 01:
I think when you... And that's a claim construction issue.

[00:10:36] Speaker 01:
Do you agree with me that's a claim construction issue?

[00:10:38] Speaker 03:
With respect, Your Honor, I don't.

[00:10:41] Speaker 03:
For the reason that we fight to the PPG case in footnote five.

[00:10:46] Speaker 01:
And what's that trying to tell me?

[00:10:48] Speaker 03:
And we quote that and we say, the task of determining whether the construed claim... Right, so here's the problem.

[00:10:59] Speaker 04:
When a court is reconstruing its own language,

[00:11:03] Speaker 04:
That's one thing.

[00:11:05] Speaker 04:
But you've got extraneous language that you all stipulated into the claim.

[00:11:10] Speaker 04:
And how does the court determine, without reference to the patent and without a markman hearing, what those stipulated words mean?

[00:11:21] Speaker 04:
So it's a very different situation when you cite these cases where we said, OK, a court's got its own construction, and then later the court says, what's the scope of that construction?

[00:11:30] Speaker 04:
That's different than trying to figure out what a stipulation meant without ever any inquiry into what the parties understood their meeting of the minds to be.

[00:11:38] Speaker 03:
I guess, I think you're right, Your Honor.

[00:11:40] Speaker 03:
And all ambiguities are supposed to be resolved in favor of the non-moving party.

[00:11:47] Speaker 03:
And so it was the burden of Kantar to prove that there are no disputed issues of material fact regarding non-infringement.

[00:11:56] Speaker 01:
And they told us- That depends on what the claim instruction is.

[00:11:59] Speaker 04:
So in other words, if your argument that all they ever argued was they don't consider time at all, and since there's evidence that they at least consider time on a quarterly basis, summer judging is not appropriate.

[00:12:13] Speaker 04:
But even if that's true, what goes to the jury?

[00:12:17] Speaker 04:
I mean, doesn't there have to be some assessment of what time frames are contemplated by the patent?

[00:12:23] Speaker 04:
Because maybe they'll come back and say, OK,

[00:12:25] Speaker 04:
All right, we'll concede that we consider it on a quarterly basis, but we don't have tight timeframes, like the patent process.

[00:12:32] Speaker 03:
Right.

[00:12:35] Speaker 03:
If the court were to construe the patent as requiring that you need to know the exact date, that will be one situation.

[00:12:44] Speaker 03:
We don't think that's what the patent says.

[00:12:46] Speaker 03:
We don't think that's what the stipulation was.

[00:12:47] Speaker 01:
If that were the case, then in general it would be correct, case seven, go to the jury.

[00:12:51] Speaker 03:
As to literal infringement.

[00:12:53] Speaker 01:
Let's stick with that for the moment, because we're hung up with this.

[00:12:58] Speaker 01:
So if the case went back and went to the jury on the claim construction we have now, the judge is going to have to charge the jury as to what had a given time means, whether it means to a specific date and time, or whether it can be more generous and include, say, quarterly time.

[00:13:18] Speaker 01:
Because the two of you very much disagree on that point of claim construction.

[00:13:23] Speaker 01:
The jury has to be told one way or the other.

[00:13:25] Speaker 01:
Do you agree?

[00:13:28] Speaker 03:
With respect, Your Honor, I don't know that the court would or should construe the stipulated construction.

[00:13:35] Speaker 03:
So if the court were to do that, we would have argument about how the jury should be instructed.

[00:13:41] Speaker 03:
We haven't had to do that given the way that all the arguments

[00:13:46] Speaker 03:
have come out.

[00:13:47] Speaker 01:
If there had been a Markman hearing and this discussion had been going on from the district court, the district court I'm certain would have leaned over and said, well I know that you all want to say that at a given time defines purchase data, but why don't you tell me why pick between one between the two of you?

[00:14:08] Speaker 01:
about whether or not, you know, it's a precise point in time or whether it can be a general point in time.

[00:14:13] Speaker 01:
Right.

[00:14:14] Speaker 01:
And then the two of you would then retreat back into the specification, right?

[00:14:19] Speaker 01:
And you would hunt for information that supported either of your points of view.

[00:14:24] Speaker 01:
And the judge would then decide.

[00:14:27] Speaker 01:
And the dispute would be over.

[00:14:29] Speaker 00:
And then the jury would... This is the answer to this discussion.

[00:14:32] Speaker 00:
This is on summary judgment.

[00:14:34] Speaker 00:
If it went back, there would be a Bachman hearing.

[00:14:36] Speaker 00:
There would be a trial.

[00:14:38] Speaker 00:
There'd be jury instructions.

[00:14:41] Speaker 00:
Am I missing the point of our body discussion?

[00:14:44] Speaker 01:
I think you're right, Your Honor.

[00:14:45] Speaker 01:
You were suggesting that there's no necessity for a court hearing because you've already construed the claims yourself.

[00:14:51] Speaker 03:
I think that there is a reasonable argument on both sides.

[00:14:56] Speaker 03:
And I think it would be reasonable either way.

[00:14:58] Speaker 03:
The precedent I was referring the court to would suggest that once there's a construction, now you leave it to the jury, even if there's some ambiguity in the construction.

[00:15:06] Speaker 03:
But I also think it would be very reasonable to have some further argument on this and let the court decide so you don't get

[00:15:13] Speaker 03:
Confusion.

[00:15:14] Speaker 04:
Is there an issue with the fact that even if Cantor ultimately has to consider, has to have a time component because otherwise you can't measure frequency.

[00:15:25] Speaker 04:
I mean, you have to have some sense of how, what time frame you're working in to determine frequency.

[00:15:31] Speaker 04:
Is there a problem that they don't actually collect that data?

[00:15:35] Speaker 04:
That someone else collects the data?

[00:15:37] Speaker 03:
I don't think so.

[00:15:38] Speaker 03:
I don't think that that mattered.

[00:15:40] Speaker 03:
They receive the data.

[00:15:42] Speaker 03:
They have that information.

[00:15:43] Speaker 03:
The whole purpose of their product is to match up the information about what people were watching with what people were buying.

[00:15:52] Speaker 01:
And they do this and... But do they perform the steps of the claim or does somebody else perform it?

[00:16:00] Speaker 03:
They perform those steps.

[00:16:03] Speaker 03:
Your Honor, I'll just point out...

[00:16:05] Speaker 03:
Clarification, my time is down to two seconds.

[00:16:07] Speaker 03:
I'm not sure if that's my first 10 minutes.

[00:16:09] Speaker 00:
Well, do answer the question and we'll restore your rebuttal time.

[00:16:14] Speaker 00:
Okay.

[00:16:14] Speaker 00:
Proceed.

[00:16:16] Speaker 03:
Thank you.

[00:16:17] Speaker 03:
I would like to make sure I'm addressing all the court's questions about purchase data, but this also is a case that involves many claims.

[00:16:27] Speaker 03:
We have, of course, the three patents.

[00:16:30] Speaker 03:
Patents are in the case involving literal infringement,

[00:16:33] Speaker 03:
and doctrinal equivalents with respect to two separate groups of products.

[00:16:37] Speaker 03:
There are also five different groups of trade secrets.

[00:16:41] Speaker 03:
There are four separate contracts that were breached and those breaches have to be presumed on summary judgment because the other side, Kantar, did not move for summary judgment as to liability.

[00:16:54] Speaker 00:
We also have to do sharing duty.

[00:16:56] Speaker 00:
Let's hear from the other side on this question of summary judgment.

[00:16:59] Speaker 00:
We'll see if we need to get into the other issues.

[00:17:02] Speaker 03:
Okay.

[00:17:03] Speaker 00:
Thank you very much.

[00:17:13] Speaker 02:
Morning, your honor and may it please the court.

[00:17:17] Speaker 02:
Your honor, the patent distinguishes between two types of concepts that TRA is trying to conflate.

[00:17:23] Speaker 02:
And I think that's what underlies the problem here.

[00:17:25] Speaker 02:
The patent on the one hand refers to what's called user type.

[00:17:29] Speaker 02:
User type is whether someone is a light, medium, or heavy purchaser of a particular product, segment, or brand, like yogurt, or Danon yogurt.

[00:17:38] Speaker 04:
But how do you know if someone's a heavy, light, I mean, I may buy 20 yogurts.

[00:17:43] Speaker 04:
If that's over the course of two years, that's not a heavy user.

[00:17:47] Speaker 02:
Exactly.

[00:17:47] Speaker 04:
Over the course of one week, it is.

[00:17:49] Speaker 02:
Precisely, Your Honor.

[00:17:50] Speaker 02:
And that's why, as we indicated in our briefing, and there was no dispute about this below, that has to be measured over a window of time.

[00:17:57] Speaker 02:
TRA is trying to argue that user type, which is this question of whether you're a light, medium, or heavy purchaser over a window, like a quarter, they're trying to argue that that's the same as purchase data, and there are two problems with that.

[00:18:09] Speaker 02:
First is that they stipulated to a construction of purchase data, and I'll come back to that one in a minute.

[00:18:14] Speaker 02:
There's a fundamental problem though, that even if they somehow were to prevail on that, still doesn't show infringement.

[00:18:21] Speaker 04:
They stipulated to add a given time, and then you all totally disagree as to what that stipulation means.

[00:18:26] Speaker 02:
Right, and I'm happy to address that momentarily, but even if they... It's going so fast.

[00:18:31] Speaker 01:
You now say that you agree.

[00:18:34] Speaker 01:
that your products involve a window of time.

[00:18:37] Speaker 02:
No, this is the point I was about to make, Your Honor.

[00:18:39] Speaker 02:
Kantar media audiences, the defendant whose product is at issue here, the CPG product, do not collect data even about purchases made at a particular window of time.

[00:18:52] Speaker 01:
They don't do even that.

[00:18:53] Speaker 01:
That's your assertion.

[00:18:55] Speaker 01:
And that's in the record.

[00:18:56] Speaker 01:
But the assertion on the other side is that they've got quarterly data about purchasers in the screenshots and other places.

[00:19:03] Speaker 02:
And the problem is they have no evidence whatsoever from which a fact finder could determine that.

[00:19:08] Speaker 02:
Why not?

[00:19:09] Speaker 02:
Because they have an expert, a technical expert, who spent three days looking at our software, sitting with the system.

[00:19:16] Speaker 02:
He wrote a 236 page expert report.

[00:19:20] Speaker 02:
He never said, because it's not true,

[00:19:22] Speaker 02:
that KMA's system, the Kantar system, collects purchase data as construed by the court.

[00:19:29] Speaker 02:
The only thing they pointed to was these two screenshots.

[00:19:33] Speaker 04:
And your own witness said repeatedly that it's on a quarterly basis.

[00:19:39] Speaker 02:
Mr. North said that what they do is that over the screenshots that they're referring to, and Mr. North said at best he could speculate about it.

[00:19:48] Speaker 02:
So there's no, again, they bear the burden of proof, so there's no evidence.

[00:19:52] Speaker 02:
beyond what Mr. North said, which is he believed that those were the quarters during which the match was made, not during which purchases were made.

[00:19:59] Speaker 01:
You didn't say believe, you said it might be.

[00:20:00] Speaker 01:
It might be.

[00:20:01] Speaker 01:
And there's no- There's a difference between believe and might be.

[00:20:03] Speaker 02:
Well, the problem is for them, because as the burden of bearing the burden of proof, there's no evidence that it does that.

[00:20:09] Speaker 01:
They're attempting with one speculative statement- Would you agree that it is possible, possible in this world that with regard to user data, you could have temporal information?

[00:20:22] Speaker 02:
So with user type, absolutely.

[00:20:24] Speaker 01:
If what you're doing is you're trying to get user data, is Clevenger a heavy user, a light user, or a minor user of Dan and yogurt?

[00:20:32] Speaker 01:
Could that assessment involve temporal information?

[00:20:37] Speaker 02:
It does.

[00:20:38] Speaker 02:
Yes, Your Honor.

[00:20:38] Speaker 02:
And the reason that Kantar Media Audiences has that information is that it gets it from a vendor.

[00:20:45] Speaker 01:
You're going too fast.

[00:20:47] Speaker 01:
So you are willing to agree that user information can include temporal information.

[00:20:56] Speaker 01:
So if I have to use this, OK?

[00:20:58] Speaker 02:
Right.

[00:20:58] Speaker 02:
So if I have to, if the patent says I have to collect apple juice.

[00:21:02] Speaker 01:
I'm trying to get some place.

[00:21:04] Speaker 01:
Yes, your honor.

[00:21:04] Speaker 01:
In this case, and you're resisting that.

[00:21:06] Speaker 01:
I'm sorry, your honor.

[00:21:07] Speaker 01:
If user information, which your current system has.

[00:21:10] Speaker 01:
User types, yes.

[00:21:11] Speaker 01:
User types.

[00:21:14] Speaker 01:
include temporal information?

[00:21:16] Speaker 02:
You're saying it doesn't?

[00:21:17] Speaker 02:
No, we agree that user type is measured over a window.

[00:21:21] Speaker 04:
Is your argument that the patent requires both the use of temporal data and the collection of temporal data?

[00:21:28] Speaker 02:
The patent requires the collection of purchase data and user type, and we don't have a dispute that we collect user type.

[00:21:35] Speaker 02:
What they're doing is they're trying to put the two together.

[00:21:37] Speaker 02:
They're conflating them.

[00:21:38] Speaker 02:
They're saying user type

[00:21:40] Speaker 04:
Because of all the evidence that we... But our point is, I think, that you can't establish user type without knowing temporal data.

[00:21:49] Speaker 02:
Right.

[00:21:49] Speaker 02:
And that's not something KMA does.

[00:21:51] Speaker 02:
That's something a vendor does, and they buy that data.

[00:21:54] Speaker 04:
So you're saying that requires the collection?

[00:21:55] Speaker 02:
Yes.

[00:21:57] Speaker 04:
So you don't really dispute that you ultimately use some aspects of purchase data in determining user type?

[00:22:05] Speaker 02:
Purchase data is used to determine purchase type.

[00:22:07] Speaker 02:
That's not done by canter media audiences.

[00:22:10] Speaker 02:
That's acquired from a vendor.

[00:22:11] Speaker 01:
Can you show me in the patent where that says exactly what you just said to me?

[00:22:18] Speaker 02:
I'll try to direct you.

[00:22:19] Speaker 01:
Purchase data is used to do with a purchase.

[00:22:21] Speaker 01:
It has nothing to do with a user.

[00:22:23] Speaker 01:
Well, so there are a couple of places.

[00:22:25] Speaker 01:
Yeah, where does the patent say that it has to be collected?

[00:22:28] Speaker 01:
Take the 940 patent, for example, and tell me in the 940 patent where purchase type data will exclude this quarterly information.

[00:22:43] Speaker 02:
Excuse me.

[00:22:57] Speaker 02:
And this was undisputed.

[00:22:58] Speaker 01:
I'm asking you to look at the patent, sir.

[00:23:00] Speaker 01:
Yes, your honor.

[00:23:01] Speaker 01:
Where in the patent is there a definition of purchased data that would exclude the quarterly information that you're asking for at point two?

[00:23:10] Speaker 02:
There are two.

[00:23:12] Speaker 02:
13 to 15, say, product purchase data derived from those same households.

[00:23:19] Speaker 01:
This is in the 7, 940?

[00:23:21] Speaker 01:
940, yes, your honor.

[00:23:22] Speaker 01:
So that the... 12 to 15.

[00:23:27] Speaker 02:
And there's another point, which is... Yeah, it says purchase data... Yes, so in other words, the idea is that the purchase data would be compared before and after the stimulus.

[00:23:39] Speaker 02:
You would know the date on which the purchase happened

[00:23:41] Speaker 02:
and you would know the date on which the advertisement happened, and you would see if the ad worked.

[00:23:44] Speaker 02:
You can't do that unless you know whether it was before or after the ad happened.

[00:23:48] Speaker 02:
That's the key point, which is why they stipulated that purchase data has to be a particular product at a given time.

[00:23:55] Speaker 02:
And we have neither of those.

[00:23:57] Speaker 02:
I would direct your honor to figure 43 of the patent.

[00:24:00] Speaker 01:
Is there any place else in the patent that you claim that that excludes formally?

[00:24:06] Speaker 02:
Yes, excluding the quarter is only part of it.

[00:24:09] Speaker 02:
There's no particular product.

[00:24:11] Speaker 02:
But as to quarter, it does exclude that.

[00:24:13] Speaker 01:
So what else in the patent is there an express exclusion of scope of quarterly temporal information from purchasing it?

[00:24:22] Speaker 02:
Column 6, lines 19 to 25, which says, by matching what programming each household is viewed or tuned to on a television with products the household has purchased, the advertiser can know which advertising has worked.

[00:24:36] Speaker 02:
Again, it's this temporal concept that the ad appears on Tuesday.

[00:24:41] Speaker 02:
You now know whether somebody makes a purchase on Wednesday.

[00:24:43] Speaker 02:
You can tell whether it's worked.

[00:24:45] Speaker 02:
And again, this is why it was articulated too.

[00:24:46] Speaker 04:
So why can't you tell whether something's worked if you know that an ad airs in a given quarter and that in a given quarter there were increased purchases of the particular product?

[00:24:58] Speaker 02:
That's just not the way that the patent describes the system working.

[00:25:01] Speaker 02:
And that's not the way the system...

[00:25:08] Speaker 01:
Again, it doesn't specify... What I'm looking for is a definition in the patent of your purchase data.

[00:25:16] Speaker 02:
Right, and we don't have that, Your Honor.

[00:25:17] Speaker 01:
There's no express definition.

[00:25:19] Speaker 01:
You look at column 19, sir.

[00:25:22] Speaker 01:
You look at column 19.

[00:25:23] Speaker 01:
Yes, Your Honor.

[00:25:23] Speaker 01:
And you look down there, line 45.

[00:25:27] Speaker 01:
Purchase data such as purchase type data, purchase type data when we're talking about specific stuff.

[00:25:35] Speaker 01:
Purchase information data, how about that?

[00:25:38] Speaker 01:
That's broader.

[00:25:40] Speaker 01:
Product category data, that's also purchase data.

[00:25:43] Speaker 01:
Demographics data, that's also purchase data.

[00:25:47] Speaker 01:
Other data and many other types of data, which would suggest that it includes user data.

[00:25:53] Speaker 01:
Right?

[00:25:54] Speaker 02:
No, Your Honor, because the patent actually distinguishes user type from purchase data.

[00:26:03] Speaker 01:
User type is mentioned in all these patents.

[00:26:06] Speaker 01:
User type is mentioned twice, only twice, where it refers to heavy user, light user, medium user.

[00:26:14] Speaker 02:
Correct, Your Honor.

[00:26:15] Speaker 01:
So what I'm asking you to do, sir, is to come back to column 19.

[00:26:19] Speaker 01:
Well, I would come back to the stipulation.

[00:26:33] Speaker 02:
Yes, Your Honor, I would point to the stipulation of the parties.

[00:26:38] Speaker 01:
This is assumed that I'm trying to do this as a judge, trying to decide what claim construction is.

[00:26:43] Speaker 01:
And I want to know if there's a definition of purchase data in the patent that either includes or excludes user data.

[00:26:53] Speaker 02:
I don't think there's a definition that answers that question in the patent, Your Honor.

[00:26:56] Speaker 01:
I think what we're working on.

[00:26:57] Speaker 01:
I pointed you to column 19, which is the closest I can find to a specific definition of purchase data.

[00:27:05] Speaker 01:
It sure looks like a definition of purchase data.

[00:27:08] Speaker 02:
Your Honor, I don't think anything in there... I have no argument that that column or line indicates what the time frame is.

[00:27:18] Speaker 02:
I do have an argument that even if they were to prevail on a quarter being a given time, there also isn't a particular purchase.

[00:27:24] Speaker 02:
Again, we're dealing with a stipulated construction.

[00:27:27] Speaker 02:
This court held in digital vending services... Have you missed?

[00:27:31] Speaker 01:
We're trying to understand what the stipulation meant.

[00:27:34] Speaker 02:
Yes, Your Honor.

[00:27:35] Speaker 02:
And one of the things in the stipulation is it has to be a particular product.

[00:27:39] Speaker 01:
And the patent does refer to that.

[00:27:40] Speaker 01:
In the stipulation, where does it say particular product?

[00:27:43] Speaker 02:
Yes, Your Honor.

[00:27:44] Speaker 02:
The stipulation says specifically data describing the purchase of a particular product at a given time.

[00:27:49] Speaker 02:
Now, figure 43 of the patent actually clarifies precisely this distinction.

[00:27:55] Speaker 01:
If Your Honor looks at figure 43... Aren't the screening shots for particular products?

[00:27:59] Speaker 02:
No, Your Honor.

[00:28:00] Speaker 02:
That's exactly the point.

[00:28:01] Speaker 02:
They are not.

[00:28:02] Speaker 02:
They are for categories of products.

[00:28:04] Speaker 02:
A particular product would be, let's say, a 19.1 ounce Apple Jacks.

[00:28:09] Speaker 04:
They are for categories... Again, you're asking us to understand what you understood by the situation.

[00:28:17] Speaker 04:
To me, a yogurt could be a particular product.

[00:28:19] Speaker 02:
That's in Figure 43, Your Honor, which specifically shows brand as Apple Jacks and then product, Apple Jacks 19.1 ounce.

[00:28:28] Speaker 02:
It's in there.

[00:28:30] Speaker 02:
And that's the thing is that the patent makes sense.

[00:28:31] Speaker 04:
You're saying the word particular.

[00:28:33] Speaker 04:
It's something that you stipulated to.

[00:28:34] Speaker 04:
And now you're saying, and we should go back now and ask you, Court of Appeals, to construe the patent so that you can understand what we in that project are doing.

[00:28:44] Speaker 02:
I understand the concern, Your Honor.

[00:28:45] Speaker 02:
And I think it does, under my mail, come down to the claim construction issue.

[00:28:49] Speaker 02:
There's no dispute as to how the system works.

[00:28:51] Speaker 02:
Their expert went through it in detail.

[00:28:53] Speaker 02:
We don't collect particular data about a particular product.

[00:28:57] Speaker 02:
That is defined as clearly as possible in figure 43, which distinguishes brand from product.

[00:29:02] Speaker 04:
How is Chobani yogurt not a particular product?

[00:29:05] Speaker 02:
Because a particular product, the point of this is that you can take particular product purchase information and you can compile it, as Your Honor mentioned earlier, into user type.

[00:29:15] Speaker 02:
We do the second, we don't do the first.

[00:29:17] Speaker 02:
We buy the user type information from a vendor who may have done that, but the argument that Cantor Media audience infringes

[00:29:26] Speaker 02:
would require cancer media audience to do that.

[00:29:28] Speaker 02:
That's what they don't do.

[00:29:30] Speaker 02:
And they stipulated that it has to be a particular product at a given time.

[00:29:34] Speaker 02:
I admit that there's some lack of clarity as to what that means.

[00:29:37] Speaker 02:
It ultimately, though, under my mail is a claim construction question.

[00:29:41] Speaker 02:
And it is clear from figure 43 that a particular product refers not just to a brand, like Chobani yogurt, but to a particular, here it would say 19.1 ounce Apple Jack.

[00:29:52] Speaker 02:
And that purchase data is not gathered by KMA.

[00:29:57] Speaker 02:
That's undisputed.

[00:29:59] Speaker 04:
So there's no factual issue.

[00:30:00] Speaker 04:
Let me ask you a question about the sanctions.

[00:30:03] Speaker 04:
After you stipulated that the narrowing was going to, if we're going to talk about stipulations, that it would have no prejudice at all to your opponent from them coming in and narrowing.

[00:30:15] Speaker 04:
And then the judge issues a sanction and says, you lose your whole claim because you narrowed your claim.

[00:30:22] Speaker 04:
I mean how is that inappropriate?

[00:30:25] Speaker 02:
So your honor, from day one Cantor was trying to determine what the reported trade secrets were.

[00:30:29] Speaker 02:
A year into the case, in June of 2012, we requested that the court compel them to disclose

[00:30:35] Speaker 04:
In response to interrogatory transcripts where the court specifically said, you know what?

[00:30:41] Speaker 04:
You're going to get all that information because they're going to narrow their claims.

[00:30:44] Speaker 04:
And then she turned around and sanctioned them for doing exactly what she asked them to do.

[00:30:47] Speaker 02:
Your honor, what they've said, and this is a little bit misleading.

[00:30:49] Speaker 02:
They said that they had narrowed their claims and they did.

[00:30:52] Speaker 02:
They dropped the number of claims, but they kept a broad swath of documents that they were purportedly basing it on.

[00:30:58] Speaker 02:
So their trade secret claim, for example, is speed and reliability of the system.

[00:31:03] Speaker 02:
We were never able to determine what is it about the speed.

[00:31:05] Speaker 02:
And they said, well, look at our documents.

[00:31:07] Speaker 02:
So we looked.

[00:31:07] Speaker 02:
And for example, one of them was, well, we use the Cognizio database.

[00:31:11] Speaker 02:
Well, Cantor doesn't.

[00:31:12] Speaker 02:
Cantor uses an Oracle database.

[00:31:13] Speaker 02:
So they said, well, that's not our trade secret.

[00:31:15] Speaker 02:
At each step of the way, they were never particular about what the trade secrets were.

[00:31:19] Speaker 04:
But you never moved for a more particular rule, 2016 disclosure.

[00:31:23] Speaker 02:
So two things on that, Your Honor.

[00:31:25] Speaker 02:
First of all, TRA is

[00:31:28] Speaker 02:
suggesting that there's a requirement for a motion, but that's only under 37D.

[00:31:31] Speaker 04:
Well, 26E is self-executing in the sense that if you fail to disclose something, that you then can't later rely on it.

[00:31:40] Speaker 04:
This is a situation where you say they disclosed too much, and they ultimately later narrowed it.

[00:31:44] Speaker 04:
You're not saying under 26E that there was some piece of information that they didn't disclose.

[00:31:49] Speaker 02:
They failed to pinpoint what their trade secrets were.

[00:31:52] Speaker 02:
And that's precisely what you need to know in order to be able to defend against it.

[00:31:55] Speaker 02:
And the judge did warn them at a hearing that if they didn't, this was in April 2013, if they didn't pinpoint what their trade secrets were, they were going to, quote, lose quickly.

[00:32:06] Speaker 04:
Now they're experts.

[00:32:07] Speaker 04:
And she specifically said the reason you lose quickly is that once you get to the merits, then you can't just be vague anymore.

[00:32:13] Speaker 04:
Right.

[00:32:14] Speaker 04:
So they then come in, in response to summary judgment, and they're not vague anymore.

[00:32:17] Speaker 04:
because she's going to get the merits.

[00:32:19] Speaker 04:
And she says, you lose because you didn't do this sooner.

[00:32:22] Speaker 02:
The problem was they still were vague, because they said, well, here it is in a somewhat narrower bunch of documents.

[00:32:27] Speaker 02:
And they never specified what they were.

[00:32:30] Speaker 02:
Entirely separate from the sanction, the court also found that the Kantar hadn't used.

[00:32:34] Speaker 02:
There was simply no evidence that Kantar had used any of these trade secrets.

[00:32:37] Speaker 02:
Kantar's system does not work the way TRA's system is.

[00:32:40] Speaker 02:
There's no allegation that it does.

[00:32:42] Speaker 02:
Their 236-page expert report doesn't say our system works the way theirs does.

[00:32:47] Speaker 02:
It's just these vague sort of hand-waving arguments that, well, speed and reliability is a trade secret.

[00:32:52] Speaker 04:
And the judge said that's not a trade secret.

[00:32:54] Speaker 04:
Well, you don't really argue that the TRA analytics wouldn't be a protectable trade secret, do you?

[00:33:00] Speaker 02:
No, absolutely not.

[00:33:02] Speaker 02:
They have a protectable trade secret in their software.

[00:33:04] Speaker 02:
We didn't copy it.

[00:33:04] Speaker 02:
We didn't even have access to it.

[00:33:06] Speaker 02:
The people who developed the Kantar system were not looking at their software and never saw it.

[00:33:10] Speaker 02:
That's not even alleged.

[00:33:11] Speaker 02:
The allegation is that speed, reliability, and performance is a trade secret.

[00:33:16] Speaker 02:
And the judge correctly held.

[00:33:17] Speaker 02:
and this is an abuse of discretion standard, that that's simply not a trade secret under... Why would that be an abuse of discretion?

[00:33:26] Speaker 02:
Well, I'm sorry, I can say the two things.

[00:33:28] Speaker 02:
The abuse of discretion standard is to sanction.

[00:33:30] Speaker 02:
So the judge is an abuse of discretion standard as to whether she correctly sanctioned them under Rule 37.

[00:33:36] Speaker 02:
And that is an abuse of discretion standard.

[00:33:39] Speaker 02:
Quite apart from that, she also correctly ruled as a matter of law.

[00:33:43] Speaker 02:
that there was insufficient evidence from which a jury could find that Kantar had ever actually used any of these trade secrets.

[00:33:50] Speaker 01:
As to some, she says use.

[00:33:51] Speaker 01:
As to others, she said the trade secret doesn't qualify.

[00:33:54] Speaker 01:
Right.

[00:33:54] Speaker 01:
So under New York law.

[00:33:55] Speaker 01:
It's a mixture of grounds, alternative grounds.

[00:33:58] Speaker 02:
Correct.

[00:33:58] Speaker 02:
On the technical trade secret not used.

[00:34:00] Speaker 01:
Does she recognize that her sanction ruling must subject to action?

[00:34:05] Speaker 02:
Well, I think the sanction, again, with respect, we've been trying since the beginning of the case to determine what these trade secrets were.

[00:34:12] Speaker 02:
And they said, it's in this massive document.

[00:34:14] Speaker 02:
At one point, they narrowed the massive document.

[00:34:17] Speaker 02:
And then they include, nonetheless, a reservation of right, saying, well, it could be in other documents as well.

[00:34:24] Speaker 02:
And she said, that's just not enough.

[00:34:25] Speaker 02:
And there's case law in the Second Circuit that that's the case that we cited.

[00:34:29] Speaker 02:
So at the end, they not only failed to identify which documents the trade secrets were in, they failed to give us an indication of where in those documents what specific things were protectable, and in any event,

[00:34:42] Speaker 02:
quite apart from that, there was no evidence that any of those trade secrets had been used.

[00:34:50] Speaker 03:
Regarding use of the trade secrets, one quick point is that there is no dispute that Kantar actually received

[00:35:11] Speaker 03:
all of the items that TRA contends for trade secrets.

[00:35:15] Speaker 03:
These were all received under NDAs.

[00:35:17] Speaker 03:
And regarding whether TRA's expert report details use, it actually does in quite a lot of detail at A1160 to 67.

[00:35:30] Speaker 03:
There's a thorough discussion by Dr. Mila of the use of the trade secrets.

[00:35:36] Speaker 04:
Maybe it's partially a function of throwing too many claims at the wall here, but the response to the summary judgment was

[00:35:44] Speaker 04:
pretty short on facts as it relates to the use.

[00:35:48] Speaker 04:
And isn't that your responsibility is to respond to a well-supported summary judgment and to at least show the judge that there are meaningful facts from which you are attempting to make this argument or create a material issue of fact?

[00:36:02] Speaker 03:
Absolutely, Your Honor.

[00:36:04] Speaker 03:
You're 100% correct, and I believe that we did do that.

[00:36:07] Speaker 03:
There is direct evidence of use in the record, but it's

[00:36:14] Speaker 03:
very clear that under the law you don't need to have direct evidence of use and that the more common situation is you have circumstantial evidence of use.

[00:36:22] Speaker 03:
And there is very abundant evidence, very abundant circumstantial evidence of use in this case, including Mr. Salama saying that it would take them two years to replicate TRA's capabilities and they were able to come out with the product in a substantially shorter amount of time.

[00:36:40] Speaker 03:
There is also abundant evidence

[00:36:42] Speaker 03:
that TRA's trade secrets actually were disseminated widely within TNS to people who were not supposed to receive the information.

[00:36:51] Speaker 03:
And there's evidence that they understood they weren't supposed to receive the information.

[00:36:58] Speaker 04:
What about the argument that the patent requires not just the use of purchase data, but the actual collection of all the purchase data, which would include data that would reflect the time frame in which it was purchased?

[00:37:12] Speaker 03:
Right.

[00:37:15] Speaker 04:
So regardless of the time frame we're talking about, even if it's a bigger time frame, if there's no collection of it, is there infringement?

[00:37:24] Speaker 03:
I don't think that was the basis for them moving for summary judgment on the collection point.

[00:37:29] Speaker 03:
What they said was their product doesn't disclose anything at all about the time.

[00:37:34] Speaker 03:
And so it has been a moving target where they've now said, well, what about a particular product?

[00:37:40] Speaker 03:
They say that they don't have any data about a particular product.

[00:37:43] Speaker 03:
And as Your Honor pointed out, they do have very specific products that they talk about, particular brands of yogurt, for example.

[00:37:50] Speaker 03:
So it's a little bit harrowing, frankly, to be defending against a summary judgment motion that keeps shifting and moving all around, and the positions keep changing.

[00:38:00] Speaker 03:
And all of the ambiguities are supposed to be resolved in our favor.

[00:38:04] Speaker 03:
It's hard to imagine what TRA could have done better, frankly, in terms of protecting its proprietary information.

[00:38:11] Speaker 03:
Mark Lieberman, who started TRA, he and his co-founders came up with an incredible, incredible advance, what's been called the holy grail in this industry, and Cantor itself acknowledged that TRA succeeded where every other company in the industry had failed for three decades, and they sought patent protection, they protected their information with NDAs and kept the trade secret,

[00:38:34] Speaker 03:
They have contracts.

[00:38:35] Speaker 03:
They disclose information to their board members under fiduciary duty law.

[00:38:40] Speaker 03:
And every one of those things was violated.

[00:38:44] Speaker 03:
And the court has said that there's not even a disputed issue, that the court can rule in favor of Cantor on every single thing on summary judgment.

[00:38:55] Speaker 03:
And with respect to the district court, the district court was acting as if this were a bench trial as opposed to summary judgment.

[00:39:02] Speaker 03:
And that's precisely what the Supreme Court just this past year reiterated cannot be done.

[00:39:08] Speaker 00:
Thank you.