The Complaint No One Noticed

On May 7, 2020, a Delaware LLC called B.E. Technology filed two nearly identical complaints against Twitter and Google in the U.S. District Court for the District of Delaware. The complaints asserted nine patents covering methods for collecting user information, building behavioral profiles, and delivering advertisements reactively — in real time, based on what a user was doing at that precise moment. The filing barely registered in the trade press. It was 2020. The pandemic had emptied the news cycle. Twitter was in the early stages of what would become a chaotic decade. Google was defending against a dozen simultaneous regulatory threats. A patent case from an entity most people had never heard of was, at best, a legal footnote. But the footnotes of infrastructure are sometimes the most important ones. The nine patents at issue — all tracing to a single inventor, Martin David Hoyle, with a priority date of July 1998 — described the basic architecture of behavioral targeting as it exists today. User interaction data collected in real time. Profiles built from that data. Advertisements selected and delivered based on what the profile said and what the user was doing right now. This is not a description of any one company's product. It is a description of how programmatic advertising works, at the level of the machine that runs inside every ad exchange, every DSP, every retargeting pixel that has ever followed you across the web. The complaint asked the court to rule that Twitter had infringed these patents, to enjoin Twitter from continuing to infringe them, and to award damages. The damages were unspecified. That is common in patent complaints — the theory of the case is that the infringement is ongoing, the injury is ongoing, and the court will sort out the arithmetic later.

The Mechanism at Issue

The three patents that survived to the end of the case — U.S. Patent Nos. 8,549,410, 8,549,411, and 8,769,440 — all trace to a single patent family originating with a 1998 application by Hoyle, a Memphis-based programmer who, according to contemporaneous reporting, had spent years building systems that would let users access their own information across multiple computers. The advertising was meant to fund the infrastructure. What the patents describe, in the language of technical claims, is a method for targeting advertisements reactively: the system monitors user interaction with the computer, builds a profile from that interaction, and selects an advertisement based partly on that profile and partly on the current session behavior. This is not the same as cookies, though it predates cookies as a consumer technology. It is not the same as the pixel, though every pixel implements some version of it. It is the underlying logic — the idea that behavioral signals can be collected and used to match an advertisement to a person, in real time, without the person having to declare an intent. Real-time bidding systems implement this logic at millisecond scale. When you load a webpage, the ad exchange runs an auction in which your behavioral profile — your inferred demographics, your recent browsing, your inferred purchase intent — is matched against advertiser bids in roughly 100 milliseconds. The patents B.E. Technology asserted described that process before it existed commercially. The priority date is not incidental. An invention filed in July 1998, when the internet had approximately 120 million users and programmatic advertising did not exist, was still enforceable against systems built in 2020. The targeting logic was the same. The scale was different. The legal claim was the same.

The PTAB Proceedings

Twitter and Google responded to the complaint by doing what large technology companies do when they are served with patent infringement claims they consider specious: they went to the Patent Trial and Appeal Board and asked it to cancel the patents. In February 2021, the two companies jointly filed four inter partes review petitions at the PTAB, challenging the validity of all claims in the three asserted patents. The petitions argued that the claims were obvious — that a person of ordinary skill in the art, given the prior art available at the time of filing, would have combined the known elements in the claimed ways. This is the standard obviousness challenge at the PTAB, and it is the primary mechanism by which large technology companies defend against NPE patent assertions. The PTAB instituted review on all four petitions in September 2021. The case in district court was stayed pending the PTAB's final written decisions — the standard procedural posture when parallel validity challenges are underway at the Board. In September 2022, the PTAB issued its decisions. The results were largely favorable to Twitter and Google: claims 1 through 24 and 26 through 37 of the '440 patent were found to be unpatentable. Claims 1 through 19 of the '410 and '411 patents were also found to be unpatentable. The challengers had successfully invalidated the vast majority of the asserted claims. One claim survived. Claim 25 of U.S. Patent No. 8,769,440, covering "providing reactive targeting of advertising to the user in real time by selecting and presenting an advertisement based at least in part on user interaction with the computer," was found not to be unpatentable. The PTAB could not find prior art that anticipated or rendered obvious that specific combination. Both sides appealed. B.E. Technology appealed the invalidation of the bulk of its claims. Twitter and Google cross-appealed the survival of Claim 25. The Federal Circuit, in a Rule 36 affirmance with no written opinion, affirmed the PTAB decisions in their entirety on August 13, 2024. Claim 25 survived. By that point, B.E. Technology had already been limited, by order of the district court in May 2023, to asserting only Claim 25. The theory of the case had been reduced to a single claim, covering the core reactive targeting mechanism — the thing that arguably describes what every major ad platform does every time it serves an advertisement.

The Settlement That Produced Nothing

On April 25, 2025, the parties filed a stipulated dismissal with prejudice under Federal Rule of Civil Procedure 41(a)(1)(A)(ii). The dismissal was mutual. Each party agreed to bear its own attorneys' fees and costs. No damages were awarded. No injunction was granted or denied. No court made any finding about whether Twitter had infringed Claim 25, or whether Claim 25 is valid, or what a reasonable royalty would be if infringement were found. The case simply ended. As it ended, the infrastructure continued operating. A stipulated dismissal with prejudice, in patent litigation, typically signals a private settlement — a licensing agreement reached without public disclosure of its terms. The absence of any fee-shifting provision (each party bearing its own costs) is consistent with a negotiated resolution in which some form of payment changed hands. But there is no way to know from the public record what the settlement contained, or whether Twitter and Google took different licenses, or whether the terms were structured to avoid creating any precedential signal about how the surviving Claim 25 should be valued. The case ran 1,814 days. It consumed an enormous amount of attorney time on both sides. It produced no precedent, no judicial interpretation of Claim 25's scope, and no ruling on the central question: whether the foundational mechanism of reactive behavioral targeting, as practiced across the entire industry, infringes a patent filed by a Memphis programmer in 1998.

The Industry That Didn't Notice

The industry, in the relevant sense, is every company that runs programmatic advertisements. Google Ads runs them. Meta Ads runs them. Every DSP and SSP in the programmatic ecosystem runs them. The RTB process — the real-time auction in which your behavioral profile is matched to an advertiser's targeting criteria in 100 milliseconds — implements, at the functional level, exactly what Claim 25 of the '440 patent describes. None of those companies were party to this case. None of them received any ruling from it. None of them received any license, any release, or any notice that their infrastructure is now operating under some cloud of potential infringement that was asserted against Twitter and Google and then privately resolved without disclosure. This is the standard condition of the industry. Patent portfolios change hands quietly. NPEs assert patents against one or two major defendants as a proxy for the entire industry. If the defendants settle, the NPE may have extracted a license fee that implicitly values the patent portfolio — but the fee is confidential, the license terms are confidential, and the release, if any, applies only to the settling parties. Every other company in the industry continues operating in the same legal position it occupied before the case began: no precedent, no guidance, no resolution. The Federal Circuit's affirmance of Claim 25's survival, incidentally, did not escape notice in patent law circles. IPWatchdog, covering the decision, noted that the surviving claim "is arguably foundational to internet advertising and to Google's ad revenue, which in 2023 was $237.86 billion." That is a precise observation. The mechanism described in Claim 25 — selecting and presenting an advertisement based at least in part on user interaction with the computer, in real time — is not Google's alone. It is the mechanism of the industry. Whether that mechanism is valid, infringed, or both, no court has ever ruled.

The Five Years That Weren't

There is a specific quality to the ending of a patent case that resolves by stipulated dismissal. It has the form of a legal proceeding — the docket entries, the briefing, the oral arguments, the decisions, the appeals — and the substantive content of a legal proceeding. But it produces none of the outputs that a legal proceeding is supposed to produce. No rule emerges from it. No principle is established. No future court will cite this case as authority for anything. It is as if the proceeding occurred in a sealed room, and when the door opened, the room was empty. What happened in that room was a negotiation. The negotiation was between the holders of a patent on reactive targeting and two companies whose advertising infrastructure implements reactive targeting. One claim survived the invalidity challenge. That claim was presumably discussed in the settlement. The discussion may have involved a license fee. The fee may have been substantial. The fee may have been structured as an ongoing royalty, or as a lump sum, or as cross-licenses involving other patents in the B.E. Technology portfolio. None of this is public. What is public is that the industry continues to use reactive targeting. That Google Ads continues to operate. That every DSP and SSP and ad exchange continues to implement the mechanism Claim 25 describes. That no company has received a legal obligation to do otherwise, and no company has received a legal certificate that it is in the clear. This is not a scandal. It is the ordinary condition of infrastructure law in the United States, where patent cases resolve privately and the resolution of one case provides no reliable signal to the rest of the industry. But it is worth noting, because the infrastructure at issue here is not peripheral. It is the mechanism by which attention is priced and sold. The five years of B.E. Technology v. Twitter produced a substantial body of procedural history and one surviving claim that arguably describes a foundational advertising technology — and the industry that uses that technology every day learned nothing from any of it. The bill came due. The bill was paid. The bill has not been shared. Sources - Delaware District Court Case 1:20-cv-00621 (B.E. Technology LLC v. Twitter, Inc.) — complaint, PTAB decisions, stipulated dismissal - PTAB IPR2021-00482 through IPR2021-00485 — Final Written Decisions, September 2022 - U.S. Patent No. 8,769,440, Claim 25 — reactive targeting of advertising in real time based on user interaction - U.S. Patent Nos. 8,549,410, 8,549,411 — computer-implemented advertising methods - Federal Circuit Rule 36 affirmance, August 13, 2024 (consolidated appeals) - IPWatchdog, "Federal Circuit Affirms PTAB Decisions in B.E. Technology v. Google/Twitter" — claim 25 survival, $237.86B Google ad revenue reference - PatSnap case analysis — case timeline, settlement structure, NPE litigation patterns - Stanford NPE Litigation Database — B.E. Technology prior suits against LinkedIn, Groupon, Pandora, Samsung, Google (2012, W.D. Tennessee)