Skip to main content

California Cases October 03, 2022: MasterObjects, Inc. v. Meta Platforms, Inc.

Up to California Cases

Court: U.S. District Court — Northern District of California
Date: Oct. 3, 2022

Case Description

1

MASTEROBJECTS, INC., Plaintiff,
v.
META PLATFORMS, INC., Defendant.

No. C 21-05428 WHA

United States District Court, N.D. California

October 3, 2022

ORDER RE MOTION TO STRIKE AND MOTION FOR SANCTIONS

WILLIAM ALSUP, UNITED STATES DISTRICT JUDGE

INTRODUCTION

In this patent-infringement action, alleged infringer moves to strike patent owner's expert reports due to patent owner's purported failure to include the theories raised in those reports in its contentions. Patent owner, in turn, accuses alleged infringer of discovery chicanery that reportedly caused this issue in the first place, and moves for sanctions. Both motions are DENIED .

STATEMENT

This action between patent owner MasterObjects, Inc. and alleged infringer Meta Platforms, Inc. (formerly Facebook), involves autocomplete technology. Autocompletion suggests ways for the user to complete her search as she actively types it into a search bar.

MasterObjects originally brought this action in the United States District Court for the Western District of Texas, Waco Division, in February 2020. MasterObjects disclosed its preliminary infringement contentions in January 2021. The action proceeded apace in Texas

2

until Judge Albright granted Meta's motion to transfer it to our district in July 2021 (Dkt. No. 86). Important here, the parties agreed upon transfer that the contentions they had already served were sufficient under the patent local rules of our district.

MasterObjects disclosed its preliminary damages contentions on February 17, 2022. Now, on the eve of trial, Meta seeks to strike certain expert reports for asserting “new” infringement and damages theories not previously disclosed in MasterObjects' contentions. This order follows full briefing and oral argument.

ANALYSIS

Our patent local rules streamline discovery and serve to swiftly uncover the primary disputes in the action. They require the parties to disclose their theories of the case early but permit amendment of the contentions as new information comes to light in discovery. Belated, previously undisclosed theories may be struck. O2 Micro Int'l Ltd. v. Monolithic Power Sys., Inc. , 467 F.3d 1355, 1363, 1365-66 (Fed. Cir. 2006).

1. Meta ' S Motion to Strike Infringement Theories.

Patent Local Rule 3 governs disclosures. Generally, a party may not use an expert report or motion memoranda “to introduce new infringement theories, new infringing instrumentalities, new invalidity theories, or new prior art references not disclosed in the parties' infringement contentions or invalidity contentions.” ASUS Comput. Int'l v. Round Rock Research, LLC , 2014 WL 1463609, at *1 (N.D. Cal. Apr. 11, 2014) (Judge Nathaniel M. Cousins). The rules do not require identification of every evidentiary item of proof, but the contentions must provide reasonable notice why the disclosing party believes it has a reasonable chance of proving its theory at trial. See Oracle Am., Inc. v. Google Inc. , 2011 WL 4479305, at *3 (N.D. Cal. Sept. 26, 2011); Shared Memory Graphics LLC v. Apple, Inc. , 812 F.Supp.2d 1022, 1024 (N.D. Cal. 2010) (Judge Edward M. Chen).

A. Cache Limitations.

Meta moves to strike MasterObjects' expert John Peck's analysis of the patented system's cache limitations in his opening report on infringement.

3

Specifically, Meta argues that Expert Peck introduced a new theory regarding which structures in the accused system meet the server-side cache limitations of the patents-in-suit. In its first amended infringement contentions, served January 15, 2021, MasterObjects identified a “memcached-based query cache” on the Typeahead backend. Expert Peck, in his report dated July 11, 2022, now contends that the “Base Index” and “Realtime Index” functionalities from the “Unicorn” backend system constitute the “cache” recited in the patents-in-suit (Weston Decl. Exhs. 10, 11).

MasterObjects argues that Meta concealed these “caches” until the second Rule 30(b)(6) deposition of William Pei, which occurred on May 9, 2022. Meta contests this assertion. It avers its first document production - which contained only fourteen files - included a writeup that explicitly identified the Base and Realtime indices (Wang Decl. Exh. 5). Meta also highlights its response to MasterObjects' Interrogatory No. 7, which asked Meta to “Identify all components and features of the server (including where all such components are located) and describe in detail how these components and features operate, including how these components and features participate in a typeahead search session” (Hosie Decl. Exh. N). Meta's response, dated January 13, 2021, explained in relevant part:

To support these distinct types of items, the backend implementation that supports Tyepahead is split into multiple itemtype specific tiers - or verticals - managed by a top-aggregator, which coordinates searching of the verticals, . . . For each vertical a collection of information from which potential suggestions can be returned to the frontend is created. This collection includes various items . . . for that vertical, but does not contain any prior search information . These collections (also referred to as indexes) are created and updated as needed

( id. Exh. O, emphasis added). Meta notes how this response referenced “indexes” multiple times.

As it turns out, this interrogatory response was misleading and - as revealed by Mr. Pei - inaccurate. Meta acknowledges the interrogatory response described the Typeahead backend in generic terms without expressly identifying it as “Unicorn” (Sanctions Opp. 1920). (Indeed, Meta declined to identify Unicorn by name in all six of its relevant interrogatory responses.) But more problematically, in his second Rule 30(b)(6) deposition, Mr. Pei directly

4

refuted Meta's response to Interrogatory No. 7. He stated multiple times that Unicorn uses “search log” - which contains “what they [the users] actually searched eventually” - as a “data source” (Pei May 9, 2022 Dep. 28-30). He repeatedly acknowledged this:

Q. And how does Facebook use this information from the prior search logs in typeahead exactly?
A. So as I said, this is -- one of the data source in the data inventory, because we know what people actually search for.... So we just use it as a data source.
* * *
Q. So explain to me, how does Facebook use the prior search log information in typeahead? . . . What do you [do] with it?
A. So as I said, this is one of the data source for the suggestions. Imagine we don't have the data log -- the search log. It's possible we need to -- all the possible strings in our data source, but not all the possible strings make sense....So in order to narrow down or get a reasonable set of suggestions, we think search log is a good source because that represents what people searched before.
* * *
Q. And you agree that the search logs are of particular importance?
A. Well, I don't know what the author [of the document being reviewed] meant by “important,” but I will say that's a useful data source for the typeahead back end

( id. at 33-35, 77, objections omitted). Mr. Pei did qualify that the search log “is only for the suggestions; [it does] not include the prefix” ( id. at 36). Meta thus attempts to defend its response to Interrogatory No. 7 here on the grounds that “there was no reason to identify the search log because . . . that log contains completed searches, not Typeahead queries or suggestions” (Sanctions Opp. 20). It further specifies that “search log (1) is not used to return results to Typeahead queries and (2) does not identify to the Unicorn index any search suggestions previously returned in response to a Typeahead query” ( id. at 4).

These explanations are baseless. Meta made an unconditional statement: the backend contains no prior search information. It did not qualify or distinguish between Typeahead queries and non-Typeahead queries. Only here does it provide a complete explanation. Meta

5

does not now deny that its Unicorn system utilizes prior search information in some way for Typeahead.

True, MasterObjects did not revise its infringement contentions in the eight-week period between the second Pei deposition and its disclosure of Expert Peck's report. So “the court must revert to a simple question: will striking the report result in not just a trial, but an overall litigation, that is more fair, or less?” Apple Inc. v. Samsung Elecs. Co. , 2012 WL 2499929, at *1 (N.D. Cal. June 27, 2012) (Judge Paul S. Grewal). This was Meta's own system. It should know the parts most vulnerable to a theory of infringement. It cannot now express surprise that MasterObjects finally stumbled onto these (allegedly) infringing parts of the system despite Meta's interrogatory responses. Fairness dictates that Meta's motion as to the cache limitations be DENIED . The relief Meta seeks is too drastic given its own role in creating the issue in the first place.

B. Source Code Citations.

Meta next argues that MasterObjects' infringement contentions only cited source code from the web version of Facebook and did not contain any citations to source code from the iOS and Android Facebook platforms. Meta acknowledges, however, that the contentions accuse Facebook's mobile applications and cites both the iOS and Android versions of the app.

Meta also asserts its Interrogatory No. 25 sought source code citations on a claim element-by-element basis. MasterObjects' response to that interrogatory, dated April 29, 2022, merely cited its then-forthcoming expert reports (Weston Decl. Exh. 9).

If Meta had a problem with the infringement contentions not reciting source code for the iOS and Android versions of the Facebook app (despite citing their relevance) then Meta should not have agreed to the sufficiency of the contentions under our local rules after transfer of this action. If Meta had a problem with MasterObjects' interrogatory response, it should have filed a motion to compel, not wait more than three months to file a motion to strike the ensuing expert reports. The motion to strike as to the source code citations is DENIED .

6

C. TCP Port Numbers and the Claimed System'S Usability Test.

Meta next argues that passages from Expert Peck's reply report and the reply declaration of MasterObjects' source-code expert Dr. Trevor Smedley should be struck because they specifically identify TCP port numbers as supplying the server response information required for the “usability test” recited in the claims-in-suit.

Meta acknowledges that MasterObjects' broad identification of HTTP header information in its infringement contentions did encompass TCP port numbers. Here is the relevant passage from the contentions (emphasis added):

The XMLHTTP Request and response contain additional information beyond what can be seen while viewing the content of these messages, such as header information , cookies, a JavaScript identifier of the request, etc. The additional information may comprise a request identification . Furthermore, the return message contains coded identifiers (e.g., “kwEntId.,” “keywordid,” and “itemloggingid”). The ID may also include one or more values or parameters associated with the HTTP message. One or more of these coded identifiers may constitute a request identification

(Weston Decl. Exh. 10 at 76). MasterObjects itself emphasized the bolded portions above in its correspondence with Meta regarding where it identified TCP port numbers in its contentions ( id. Exh. 20). Meta concedes that HTTP headers may contain TCP port numbers, but maintains that headers also contain myriad other information (Br. 14). So the issue here is whether MasterObjects' general identification gave Meta reasonable notice of its infringement theory.

Meta's problem is that it agreed that MasterObjects' broad contentions were sufficient upon transfer of this case over a year ago. Meta now argues “the contentions' vague reference to header information did not put Meta on notice of MasterObjects' actual infringement theory” (Reply Br. 14). But the time for these sorts of complaints has passed. If “header information” included too many possibilities, then Meta should not have agreed to the adequacy of the infringement contentions upon transfer and should have moved to strike the contentions then. A contrary conclusion would risk gamesmanship by encouraging parties to leave broad patent disclosures unchallenged in order to challenge expert reports instead.

7

Meta has also not demonstrated that Experts Peck and Smedley failed to compose “true” reply materials. In their opening report and declaration, Experts Peck and Smedley, respectively, discuss how “[t]he server system asynchronously receives and responds to the query messages from the client object over a network. For example, the HTTP GET method includes a response from the server system” (Peck Rep. ¶¶ 63-82, Weston Decl. Exh. 11). In his rebuttal report, Meta's expert John Black states “Multiple XMLHTTP objects can exist at the same time, each awaiting a response for different requests, but they will use different connections to the server ( i.e. , each request is made over a separate TLS connection that was established in advance)” (Black Rebuttal Rep. ¶¶ 45-46, Hosie Decl. Exh. D). In response, Expert Smedley replied that “[t]he connections Dr. Black is referring to are TCP connections, as provided through the HTTP and TCP/IP protocols....TCP keeps all these connections straight through port numbers ” (Smedley Reply Decl. at 1, Weston Decl. Exh. 18). Expert Smedley specifies the “connection” underlying the method discussed in his original declaration in order to rebut Expert Black's response. Expert Peck's reply report briefly echoes Expert Smedley's analysis.

The motion to strike as to this material is DENIED . This order cannot conclude that Experts Smedley and Peck introduced new theories in their reply papers.

2. Meta ' S Motion to Strike Damages Theories.

This order now switches gears to discuss MasterObjects' damages contentions, which are governed by Patent Local Rule 3-8. The rule requires the patent owner to “[i]dentify each of the category(-ies) of damages it is seeking for the asserted infringement, as well as its theories of recovery, factual support for those theories, and computations of damages within each category.” This order adopts the position that a party has a duty to supplement or amend damages contentions when the party's damages theory shifts “in some material respect.” Netfuel, Inc. v. Cisco Sys. Inc. , 2020 WL 4381768, at *3 (N.D. Cal. July 31, 2020) (Judge Edward J. Davila); accord Looksmart Grp., Inc. v. Microsoft Corp. , 386 F.Supp.3d 1222, 1227 (N.D. Cal. 2019) (Judge Jon S. Tigar).

8

MasterObjects' February 2022 preliminary damages contentions disclosed a damages theory based on a reasonable royalty (Damages Contentions at 1, Weston Decl. Exh. 1). Specifically:

MasterObjects anticipates that its claim for reasonable royalty damages will be based, in part, on the portion of Meta's sales and/or incremental profits that are attributed to the patented invention. This may take the form of a running royalty rate based on a percentage of Meta's revenues or a per-unit royalty based on the usage of the Accused Instrumentalities

( id. at 8). The question presented here is whether the theory of recovery recited by MasterObjects' Expert William Latham, in his report disclosed on July 11, 2022, shifted away from this statement in some material respect.

First , Expert Latham determined the “per query” royalty rate based on “the price list MasterObjects actually used in its business” (Latham Rep. ¶ 107, Weston Decl. Exh. 4). The price list Expert Latham refers to is MasterObjects' price list for its QuestField product from 2010. MasterObjects sold QuestField but discontinued the product before the asserted patents issued.

The damages disclosure was barebones, but it did disclose a per-unit royalty theory of damages. No motion to require more specificity was made, and here we are mere weeks before the scheduled trial and Meta is now complaining about the barebones nature of the contentions. The contention, brief as it is, necessarily disclosed damages based upon a royalty rate as well as the frequency of use by Meta of the patents. The royalty rate would necessarily have to be based on some analogous historical expense. The 2010 price list should have been an obvious candidate. As ordered at the hearing, the parties jointly filed excerpts of deposition transcripts where counsel had the opportunity to pose questions on the price list.

Upon review of those excerpts, this order finds that Meta had a reasonable opportunity to probe the issues related to the price list. In MasterObjects' CEO Mr. Smit's deposition that occurred on June 25, 2021, he specifically cited the price lists: “We had price lists and pricing strategies that we devised later than our sale to Hewlett Packard. We considered Hewlett Packard to be our reference customer” (Smit June 25, 2021 Dep. 344-45, Dkt. No. 249-5). Mr.

9

Smit made this statement in the context of a customer-by-customer review that touched on the licenses for Hewlett Packard, Siemens, and Princeton University ( id. at 308, 309, 348-50). Meta's counsel, notably, asked Mr. Smit in that deposition about the discount Hewlett Packard received over the “normal retail pricing” for MasterObjects' software ( id. at 345).

Meta also deposed MasterObjects' former Chief Marketing Officer William Hassebrock in June 2022, who mentioned that MasterObjects gave Hewlett Packard a discount off the “list price,” and explained that MasterObjects had several iterations of a price list (Hassebrock Dep. 81-84, Dkt. No. 249-7). Meta had the opportunity to take another deposition of Mr. Smit regarding damages issues on August 22, 2022, after Expert Latham had served his damages report in July.

At this August deposition, Meta had free rein to ask any and all questions it wanted of Mr. Smit regarding the price list, including its past application and whether MasterObjects inflated the listed prices therein. Meta asserts that Mr. Smit does not qualify as an objective witness; it also complains it did not have the opportunity to seek discovery from MasterObjects' customers about why they never paid per-query pricing. But Meta knew about those customers well before Expert Latham's report and never bothered to subpoena them. Meta had a reasonable opportunity to take discovery into the history of the price list and customers licensed thereunder. The frequency of use information used in Expert Latham's report is taken directly from Meta's own records, which Meta produced only ten days before the deadline for opening expert reports. Meta's motion to strike the royalty rate based on the price list is accordingly DENIED .

Second , Expert Latham presents another damages theory regarding MasterObjects' prior license with a third-party tech company from 2016. Expert Latham presents this as an alternative theory. As explained in the hearing, MasterObjects will not be permitted to introduce this theory at trial but, should Meta open the door and raise this theory itself, MasterObjects would then be fully entitled to present testimony on this theory. Accordingly, with this understanding, Meta's motion to strike as to this patent-license theory is DENIED .

10

3. MasterObjects ' Motion for Sanctions.

MasterObjects, to counter Meta's motion to strike, moved for sanctions for discovery misconduct pursuant to the Court's inherent authority and Rule 37. As for Rule 37, MasterObjects asserts Meta violated Paragraph 25 of the supplemental CMC order, which requires counsel that learns of unproduced material that contradict produced discovery that would reasonably be credited by the judge or jury or opposing counsel to correct their prior statement “very promptly.” Because the parties filed highly curated excerpts of the relevant transcripts of the depositions of Meta's witnesses, a prior order requested complete, unabridged versions of the transcripts.

Upon review, this order finds that while some witnesses equivocated and others at times appeared to feign incomprehension, this order cannot conclude that Meta's behavior during discovery warrants sanctions. MasterObjects seeks drastic relief: that MasterObjects has proven the accused system practices the cache limitations; so instruct the jury; and preclude any contentions otherwise. That goes too far. As explained above, while Meta's obfuscation warrants denial of certain aspects of its motion to strike no further action is warranted. MasterObjects' motion is DENIED . Finally, this order makes no determination now regarding Meta's complaints regarding opposing counsel. Counsel for both sides are reminded of their obligation to follow our district's standards of professional conduct.

CONCLUSION

For the foregoing reasons, both parties' motions are DENIED.

IT IS SO ORDERED.