RavenWhite Licensing LLC v. Walmart Inc

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: RavenWhite Licensing LLC (Delaware)
  • Defendant: Walmart Inc. (Delaware) and Wal-Mart Stores Texas, LLC (Delaware)
  • Plaintiff’s Counsel: Kramer Alberti Lim & Tonkovich LLP; Ward, Smith & Hill PLLC
• Case Identification: 2:24-cv-00689, E.D. Tex., 08/21/2024
• Venue Allegations: Plaintiff alleges venue is proper in the Eastern District of Texas due to Defendants’ continuous physical presence, including retail stores and fulfillment centers within the district, and because Defendants solicit and induce customers and users in the district via the Walmart website.
• Core Dispute: Plaintiff alleges that Defendant’s e-commerce website and its associated advertising platform infringe two patents, one related to a method for identifying client devices using different types of browser storage and another related to a system for targeted advertising based on user purchase history and inferred needs.
• Technical Context: The technologies at issue involve foundational aspects of modern e-commerce: persistent user identification for personalized experiences and sophisticated, behavior-based ad targeting to optimize marketing effectiveness.
• Key Procedural History: The complaint states that Plaintiff is the exclusive licensee of the patents-in-suit. The ’823 Patent is owned by RavenWhite Security, Inc., and the ’402 Patent is owned by Security Technology, LLC. No other significant procedural events are mentioned.

Case Timeline

Date	Event
2000-01-01	Walmart launches its eCommerce initiative (approximate)
2007-01-01	Walmart starts "Site to Store" service (approximate)
2014-12-04	Earliest Priority Date for ’823 Patent
2019-08-12	Earliest Priority Date for ’402 Patent
2020-03-17	’823 Patent Issued
2023-01-24	’402 Patent Issued
2024-08-21	Complaint Filed

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 10,594,823 - "Method and Apparatus for Storing Information in a Browser Storage Area of a Client Device," issued March 17, 2020

The Invention Explained

• Problem Addressed: The patent describes the privacy drawbacks and unreliability of traditional browser cookies, which users can easily block or delete. This prevents websites from consistently identifying returning users, which is critical for personalization and, in the case of financial institutions, for security and authentication ('823 Patent, col. 2:11-44).
• The Patented Solution: The invention proposes a method to store identifying information in more persistent browser storage areas, such as the history cache or Temporary Internet Files (TIFs), creating what it terms a "cache cookie." A server directs a user's browser to make a specific set of network resource requests, which causes data representing those requests to be stored locally. The server can later identify the device by observing which resources the browser retrieves from its local cache versus requesting anew from the server, effectively "reading" the stored data without relying on traditional cookies ('823 Patent, Abstract; col. 3:5-24).
• Technical Importance: This technique offered a more durable method for device fingerprinting than was available with standard cookies, providing a workaround to common user actions (like cookie deletion) to maintain session continuity and personalization ('823 Patent, col. 2:31-44).

Key Claims at a Glance

• The complaint asserts at least independent claim 1 (Compl. ¶40).
• Essential Elements of Claim 1 (A system):
  • Receiving a network resource request corresponding to a "first cookie of a first type" stored in a "first client device browser storage area" during a first session.
  • The system also involves a "second cookie of a second type" stored in a "second client device browser storage area different from the first" during a second session.
  • Based on the request, determining information that was encoded and stored on the client device.
  • Performing a "first identification" of the user/device using the first cookie.
  • Performing a "second identification" of the user/device using the second cookie.
  • Performing a determination based on the presence of one of the cookies and the absence of the other.
  • A memory coupled to the processor(s).

U.S. Patent No. 11,562,402 - "Advertising Model," issued January 24, 2023

The Invention Explained

• Problem Addressed: The patent identifies inefficiencies in typical online advertising schemes where ads are selected primarily based on the highest monetary bid, without sufficient regard for user context or intent (’402 Patent, col. 1:20-28).
• The Patented Solution: The invention describes a tiered advertising system that first determines a "quality level" for a user based on their profile, searches, and past purchases. A key aspect is determining that a user's "need" for a first product category has been met. Following this determination, the system identifies a user's interest in a second category and determines a second quality level. This allows for displaying more relevant ads (e.g., for cross-selling) based on a more nuanced understanding of the user's position in a purchasing journey (’402 Patent, Abstract; col. 20:3-49).
• Technical Importance: The technology enables a shift from simple, bid-based ad auctions to a more sophisticated model that tracks the user's lifecycle, potentially improving ad relevance and return on investment by targeting users at opportune moments, such as after a primary need has been fulfilled (’402 Patent, col. 1:26-30).

Key Claims at a Glance

• The complaint asserts at least independent claim 19 (Compl. ¶65).
• Essential Elements of Claim 19 (A computer program product):
  • Determining a "first quality level" for a user profile based on a first search or purchase, using a unique identifier and clustering.
  • Determining an "indication of interest in a first category" based on a second search or purchase.
  • Storing this indication of interest.
  • Subsequently determining that a "need relative to the first category has been met" based on a third search or purchase.
  • Based on the need being met, determining an "indication of interest in a second category."
  • In response to the need being met, determining a "second quality level."
  • The quality level(s) are based on a "conversion assessment" which is based on "historical click behavior."
  • Displaying an advertisement based on at least one of the quality levels.

III. The Accused Instrumentality

• Product Identification: The complaint names the "Walmart system," which comprises Walmart's e-commerce website (walmart.com), its computer networks, and the "Walmart Connect" advertising platform and service (Compl. ¶¶38, 63).
• Functionality and Market Context:
  • The accused system provides an "omni-channel" retail experience, integrating Walmart's physical stores with its digital presence (Compl. ¶12). The walmart.com website allegedly uses various tracking technologies, including "session cookies and persistent cookies," as well as other trackers like pixels and beacons stored in different browser locations such as "Local Storage", "Session Storage", and "Cookies" (Compl. ¶¶41, 43). A screenshot provided in the complaint shows the developer view of a browser, illustrating that trackers for "walmart.com" are stored in distinct areas including "Local Storage", "Session Storage", "IndexedDB", and "Cookies" (Compl. ¶43, p. 16).
  • The Walmart Connect platform is an advertising service for placing sponsored product ads on Walmart's site (Compl. ¶63). It operates on a second-price, real-time auction model to determine which ads to show (Compl. ¶¶66, 69). Ad placement is based on a "relevancy" score, which considers factors like "customer's intent," cost-per-click (CPC) bid price, click frequency, and historical performance (Compl. ¶¶68, 76).

IV. Analysis of Infringement Allegations

’823 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
receive a network resource request from a client device, wherein the network resource request corresponds to a first cookie of a first type... stored... during a first previous network session... wherein a second cookie of a second type different from the first type was... stored... during a second previous network session... and wherein the first cookie... is stored in a first client device browser storage area and the second cookie... is stored in a second client device browser storage area different from the first client device browser storage area;	The Walmart system is alleged to use a "CID" cookie (first type) stored in the browser's "Cookies" area (first storage area) and a "glassCartIdMap" cookie (second type) stored in "Local Storage" (second storage area) to identify users across different sessions.	¶¶42, 45-47	col. 15:34-52
based at least in part on the network resource request from the client device... determine information that was encoded and stored in the client device;	When a client with a stored CID cookie makes a request, the system allegedly determines the user's identity and displays their account information without requiring a new login. A screenshot shows the user interface displaying "Hi, Ruby G Account" after recognizing the user.	¶¶48-50, p. 19	col. 15:53-58
perform a first identification of at least one of the client device and a user of the client device using the first cookie of the first type...;	The system allegedly performs identification using the persistent "CID" cookie, which contains the user's ID and is associated with their sign-in credentials.	¶51	col. 15:59-65
perform a second identification of at least one of the client device and the user of the client device using the second cookie of the second type;	The system allegedly performs a second identification using the "glassCartIdMap" cookie, which is stored in local storage and contains the user's cart information as well as the same user ID found in the CID cookie.	¶53	col. 16:1-3
perform a determination based at least in part on (1) a presence of a network resource request associated with one of the first cookie and the second cookie, and (2) an absence of a network resource request associated with the other of the first cookie and the second cookie; and	The complaint alleges that if the "glassCartIdMap" cookie (second cookie) is absent, the system uses the presence of the "CID" cookie (first cookie) to re-create the user's cart and download a new "glassCartIdMap" cookie.	¶¶54, 56	col. 16:4-9
a memory coupled to the one or more processors and configured to provide the one or more processors with instructions.	The Walmart system allegedly employs Envoy web servers, which require memory coupled to processors to function.	¶61	col. 16:10-12

• Identified Points of Contention:
  • Scope Questions: A central question may be whether the claim term "cookie" can be interpreted to cover different types of browser storage technologies as alleged. The defense could argue that a JSON data structure in "Local Storage" (the alleged "glassCartIdMap") is technically distinct from an HTTP cookie stored in the "Cookies" area (the alleged "CID" cookie) and falls outside the scope of the term "cookie" as understood in the patent.
  • Technical Questions: The complaint alleges the "CID" and "glassCartIdMap" objects perform first and second identifications, but also states they are linked by the same customer ID (Compl. ¶53). This raises the question of whether they constitute two truly distinct identifications as required by the claim, or if the second is merely derivative data dependent on the first.

’402 Patent Infringement Allegations

Claim Element (from Independent Claim 19)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
determining a first quality level associated with a user profile... based at least in part on... (1) a first search... or (2) a first purchase... and wherein the determining... is based at least in part on a unique identifier and clustering;	Walmart's ad platform allegedly determines a "relevancy score" (first quality level) based on user searches and purchases. The complaint cites a Walmart research paper on using "clustering ads intentions" for its search engine marketing.	¶¶67-68, 72, 75	col. 20:7-13
determining, for a user... an indication of interest in a first category, wherein the indication of interest... is determined based at least in part on... (1) a second search... or (2) a second purchase...;	Walmart allegedly determines "customer's intent" (indication of interest) for a product category based on user search and purchase history. A screenshot from a Walmart guide highlights that the platform considers "how closely the advertised product matches the customer's intent."	¶¶76, 79, p. 36	col. 20:14-20
storing, in a record associated with the user, the indication of interest in the first category;	The system is alleged to store users' indications of interest, including browsing history, search activity, and purchase history, in records associated with the user profile.	¶82	col. 20:21-23
subsequent to determining the indication of interest... determining, for the user, that a need relative to the first category has been met based at least in part on... (1) a third search... or (2) a third purchase...;	The complaint alleges Walmart determines a need has been met by observing subsequent user behavior, such as searching for a complementary product. It cites a Walmart paper describing "multi-level cobought models to understand her shopping tasks," which implies determining when a task (need) is complete.	¶84, 88	col. 20:24-30
based at least in part on the determination that the need... has been met, determining an indication of interest in a second category...;	After a user's initial shopping tasks are understood to be complete (need met), Walmart allegedly recommends "a series of items at checkout," which is presented as determining interest in a second category (the checkout items).	¶87-88	col. 20:31-36
in response to determining that the need... has been met, determining a second quality level associated with the user, wherein the second quality level is determined with respect to the first category;	The complaint alleges that after a need is met, Walmart determines a second quality level by assessing "the customer's intent," which includes product category and purchase history, to serve further relevant ads.	¶90	col. 20:37-41
wherein at least one of the first quality level or the second quality level is based at least in part on a conversion assessment... and wherein the conversion assessment is based at least in part on historical click behavior; and	Walmart's system allegedly tracks "Conversion Rate" as a key performance metric, which is a conversion assessment. The relevancy score is based on "click frequency," which is historical click behavior. A screenshot shows "Conversion Rate" as a "Key Performance Metric."	¶¶94-95, p. 52	col. 20:42-46
displaying an advertisement to the user based at least in part on the at least one of the first quality level or the second quality level.	The Walmart Connect platform allegedly runs a real-time auction to select and display an advertisement to the user based on ad relevance (the quality level).	¶95	col. 20:47-49

• Identified Points of Contention:
  • Scope Questions: A primary issue may be whether Walmart’s system performs the specific, sequential logic of claim 19: determine interest in category A -> determine need for A is met -> determine interest in category B. The defense could argue its ad platform uses a holistic, real-time scoring model where all user data points are weighed concurrently, rather than following the discrete, linear process recited in the claim.
  • Technical Questions: What evidence demonstrates that Walmart's "relevancy score" is separated into a "first quality level" and a "second quality level," as opposed to being a single, continuously updated metric? The defense may contend that the accused system does not make the distinct determinations at the separate process stages required by the claim language.

V. Key Claim Terms for Construction

For U.S. Patent No. 10,594,823:

• The Term: "cookie"
• Context and Importance: The infringement theory for the ’823 patent depends on mapping the "CID" tracker to a "first cookie of a first type" and the "glassCartIdMap" tracker to a "second cookie of a second type." As these trackers are allegedly stored in different browser locations ("Cookies" vs. "Local Storage"), the scope of the term "cookie" will be critical to determining whether the accused system meets this claim limitation.
• Intrinsic Evidence for Interpretation:
  • Evidence for a Broader Interpretation: The patent repeatedly refers to its invention as a "cache cookie" and describes storing information in various browser locations beyond traditional cookie files, including the "history cache" and "Temporary Internet Files (TIFs) area" ('823 Patent, col. 5:4-8, 5:55-57). This may support an argument that "cookie" should be construed functionally to mean any data stored on a client device for identification purposes.
  • Evidence for a Narrower Interpretation: The specification distinguishes its "cache cookie" from "traditional cookies" ('823 Patent, col. 5:16-24), yet the claim itself uses the unmodified term "cookie." A defendant may argue that this term should be given its plain and ordinary meaning in the art, which typically refers to HTTP cookies, and that other technologies like Local Storage are distinct and not covered.

For U.S. Patent No. 11,562,402:

• The Term: "determining... that a need relative to the first category has been met"
• Context and Importance: This limitation is a central, sequential step in the claimed method. Infringement requires showing that the accused system performs this specific determination as a predicate to identifying interest in a second category. Practitioners may focus on this term because it defines the core logic and structure of the claimed process.
• Intrinsic Evidence for Interpretation:
  • Evidence for a Broader Interpretation: The specification describes this functionally, providing an example where a user's search for a camera lens implies that their prior "need" for a camera has been met ('402 Patent, col. 5:29-36). This could support a construction where any algorithmic inference of a completed purchase or task satisfies the limitation.
  • Evidence for a Narrower Interpretation: The claim language recites an affirmative "determining" step. A defendant could argue this requires a discrete, binary decision ("need met: yes/no"), not merely a probabilistic shift in a user's interest score within a complex algorithm. The patent's examples focus on concrete events like a purchase, which may support a narrower construction requiring an actual or clearly inferred transaction event.

VI. Other Allegations

• Indirect Infringement: For the ’402 Patent, the complaint alleges induced infringement under 35 U.S.C. § 271(b). It alleges Walmart induces its clients, customers, and end users to infringe by providing "instruction materials, support training, and services" and by "actively inducing its customers to sign up for online accounts" to use the Walmart Connect advertising system (Compl. ¶¶96-97).
• Willful Infringement: The complaint does not contain an explicit count for willful infringement. However, for the ’402 Patent, it alleges inducement and knowledge "From at least the time Walmart received notice of the '402 patent," which lays the groundwork for a claim of post-filing willfulness (Compl. ¶¶96, 98). No similar allegation is made for the '823 Patent.

VII. Analyst’s Conclusion: Key Questions for the Case

This case presents two distinct technological disputes, one centered on the definition of a web technology and the other on the specific logic of an advertising algorithm. The key questions for the court will likely be:

• A core issue will be one of definitional scope: For the ’823 patent, can the term "cookie" be construed broadly to cover functionally similar but technically distinct browser storage mechanisms like HTML5 Local Storage, or is its meaning limited by the patent's specific examples and the conventional understanding in the art?
• A key evidentiary question will be one of process structure: For the ’402 patent, does the accused Walmart Connect platform perform the specific, sequential process of claim 19—making a discrete determination that a "need has been met" before assessing interest in a new category—or does its real-time, auction-based system employ a holistic algorithm that fundamentally differs from the claimed method's step-by-step logic?