US 10026100

I'll search for authoritative records on this specific patent number and its litigation history.

Let me retrieve the full claim set and check for any 2026 appellate activity.

Let me pull the full claim text and verify whether any 2026 appellate activity specifically involves the '100 patent.

US Patent 10,026,100 — Analyst Summary

Record verification. I searched on the literal number 10026100 / 10,026,100 (and did not substitute any similar number). Authoritative sources located: Google Patents (US10026100B2), the USPTO/PTAB document repository (ptacts.uspto.gov), the issued patent front page as filed in court (D. Del. 1:18‑cv‑00943, Doc. 9‑2, Exhibit E), FreePatentsOnline, and RPX Insight (which reproduces the issued claim set). The full text you supplied is consistent with these, except that the "Description" you supplied is the shared original 2000 specification (the "super‑saturation" disclosure), not the granted claims — the granted claims read on off‑site targeted Internet advertising. Note also that the date in my instructions (2026‑09‑28) differs from the date in your prompt (2026‑04‑26); this does not affect any bibliographic data below.

1. Bibliographic data

Continuation chain (from the cross-reference): 09/723,391 (filed 2000‑11‑28 → US 6,832,207) → 10/996,491 (2004‑11‑24 → US 7,822,639) → 12/911,601 (2010‑10‑25, abandoned) → 13/368,686 (2012‑02‑08 → US 8,244,586) → 13/585,765 (2012‑08‑14 → US 8,600,815) → 14/091,285 → US 10,026,100. It was itself continued as US 10,628,857 (filed 2018‑07‑11).

Classification: G06Q 30/02, 30/06; CPC G06Q 30/0269 (targeted ads based on user profile/attribute), 30/0251, 30/0214, 30/0277.

Abstract (verbatim): "A super-saturation method for information-media relates to directing information to an electronic media consumer (reader, surfer, viewer, listener, etc.). Embodiments of the present invention facilitate a first media entity substantially offering out of context information placement using a second media entity. A facilitator entity guarantees that a consumer of the second media is a known consumer of the first media. Accordingly, the second media entity presents an out of context information placement. For example, an Internet site 'AAA' is oversubscribed with potential paying advertising content at $100 CPM. This site offers advertisers an option to present their advertisements to certified 'AAA' viewers, albeit on a non-AAA Internet site, for $50 CPM. The facilitator locates certified AAA viewers at Internet site 'BBB' that normally charges $30 CPM. Contracts between AAA, BBB, and the facilitator divide a new revenue stream of $20 CPM among them."

2. Plain-language overview of the independent claims

The patent has three independent claims — claim 1 (method), claim 15 (independent claim, with dependents 16–28), and claim 29 (apparatus/system-style claim, with dependents 30–42). This grouping is corroborated by RPX's reproduction of the issued claims and by Roku's invalidity contentions, which quote the claim preambles. AlmondNet asserted claims 1–11 and 13–14 of the '100 patent against Roku.

Claim 1 — "A method of facilitating off-site targeted Internet advertising…"
The claim is performed "with a first computer controlled by a proprietor of a first Internet site." The preamble requires that this proprietor has already sold, for a first price, ad placements on other Internet sites, the ad being for a third‑party advertiser's offering. The steps, all "automatically," are:

• (a) the first computer places a cookie on each of many visitor computers visiting the first site;
• (b) the first computer detects which of those computers conducted a specific predetermined action on the first site (e.g., per the dependent claims, answering a particular question, reacting to a predetermined aspect of the site, remaining a predetermined amount of time, or expressing interest in a section — claims 2–5);
• (c) for each such detected computer, the first computer stores information reachable through the cookie indicating the action was performed (dependent claim 7 recites storing it in a database linked to the cookie);
• (d) later, working with a second computer that controls ad space on a second Internet site, the first computer detects visitors arriving at the second site bearing the cookie and, using the cookie‑accessible information, identifies which of them performed the action on the first site;
• (e) the first computer communicates with the second computer to cause the advertisement to be served to those identified computers, for display in ad space on the second Internet site (dependent claims 8–9 specify the second computer directing the visitor's browser to fetch the ad, and the second computer being controlled by an agency under contract);
• (f) the first computer causes the second site's proprietor to receive revenue for that display in an amount less than the first price (dependent claims 10–11 specify CPM‑ or fixed‑fee‑per‑ad‑based revenue); and
• (g) the first site's proprietor keeps at least part of the difference between the first price and what the second site received.

In short: one site pre‑qualifies its own visitors with a cookie, follows them to a third‑party site, and monetizes a discounted off‑site impression while retaining the spread.

Claim 15 — a second independent claim in the same family (dependents 16–28). I could not retrieve its verbatim text from the sources searched; based on the dependent claims charted against it (e.g., claim 19, "second computerized apparatus"; claim 23/24, display "using web television"; claim 28, ad subject matter relating to the first site's content but not the second's), it tracks the claim 1 subject matter with a different statutory framing. Treat this characterization as provisional.

Claim 29 — an independent claim reciting a "first computerized apparatus controlled by a proprietor of a first Internet site," with a parallel "second computerized apparatus," and reciting the same operative steps as claim 1 (cookie placement, detection of the specific predetermined action, storing cookie‑accessible information, later recognition at the second site, causing ad delivery, causing the second site to receive revenue less than the first price, and the first site retaining at least part of the difference). Its dependents 30–42 include the web‑television and "content of the first site but not the second site" limitations.

3. Prosecution / enforcement context (relevant to claim scope)

• The claim language reflects a long prosecution negotiation: applicant amended the predecessor claims (e.g., claim 108/121 of the '639 family) to make explicit the two‑price / first‑site‑retains‑the‑spread structure so as to distinguish the Roth/Engage prior art (see the '639 file‑history excerpts in the PTAB record). Claim 1(g) of the '100 patent carries that same "retains at least part of the difference" limitation.
• The '100 patent carries a terminal disclaimer and lists an adjusted expiration of 2023‑01‑23, meaning the document appears to be expired, notwithstanding its 2018 issue date. This is stated on the face of the record and is worth independent confirmation in USPTO PatentCenter if you need a legal conclusion.

4. Litigation and 2026 appellate activity

District court matters naming the '100 patent:

• AlmondNet, Inc. v. Oath Holdings Inc., D. Del. 1:18‑cv‑00943 (the '100 patent attached as Exhibit E).
• AlmondNet, Inc. & Intent IQ, LLC v. Roku, Inc., W.D. Tex. 6:21‑cv‑00731 — Roku's advertising platform accused; case stayed, transferred to D. Del. as 1:22‑cv‑01540‑MN, consolidated for pretrial into 1:21‑cv‑01035‑MN (Roku, Inc. v. AlmondNet, Inc., DJ action); docket closed via stipulation of dismissal in June 2025.
• AlmondNet v. Amazon.com (W.D. Tex. 6:21‑cv‑00898) — the AlmondNet correspondence identifies the '100 patent among those asserted regarding Amazon's DSP service (and in the earlier Delaware Yahoo!/Oath suit).
• PTAB: IPR2022‑01236 (Roku v. Intent IQ) is the vehicle in which AlmondNet served preliminary infringement contentions listing the '100 patent and its asserted claims (1–11, 13–14). The '100 patent itself was not, to my knowledge, the subject of an instituted IPR; the IPRs I found concern sibling patents ('639, '139, '878, '249, etc.).

Federal Circuit in 2026 (AlmondNet as party):

• 2024‑1834, AlmondNet, Inc. v. Meta Platforms, Inc. (appeal from IPR2022‑01436) — Rule 36 affirmance, April 13, 2026 (Lourie, Dyk, Taranto).
• 25‑1967 / 26‑1326, AlmondNet, Inc. v. Amazon.com, Inc. (appeal from W.D. Tex. 6:21‑cv‑00898, Judge Albright) — dismissed by joint agreement under Fed. R. App. P. 42(b), April 6, 2026.

Explicit uncertainty: I found no Federal Circuit docket in 2026 that names U.S. Patent 10,026,100 specifically. The two 2026 AlmondNet appeals above relate to other patents in the family/IPRs, and I could not confirm from the search results which patents IPR2022‑01436 covers. I therefore cannot state that any 2026 CAFC proceeding adjudicates the '100 patent. The '100 patent's 2023 adjusted expiration and the June 2025 Roku dismissal make a 2026 appeal on that patent unlikely, but that inference is not authoritative.

5. Items I would flag as not fully verified

1. Verbatim text of claim 15 (independent) — retrieved only indirectly; do not quote it from this summary.
2. Claim 29's exact wording — reconstructed from preamble fragments in Roku's invalidity contentions; the substance is reliable, the wording is not verbatim.
3. Whether the exact adjusted expiration (2023‑01‑23) and "expired" status are current — this comes from the Google Patents snapshot and RPX, not from a live PatentCenter query.
4. Whether the 2026 CAFC orders touch the '100 patent — no evidence that they do.

I'll research litigation involving US Patent 10026100. Let me run several searches.

Let me dig into the specific cases identified.

Let me search specifically for the '100 patent in court records and IPR proceedings.

I was able to complete several searches but hit my tool-call limit before exhausting docket-level verification, so I'll flag confidence levels carefully.

Litigation involving U.S. Patent No. 10,026,100 ("Methods and apparatus for facilitated off-site targeted internet advertising")

Owner/assignee: AlmondNet, Inc. (some co-asserted family patents are owned by Intent IQ, LLC). Inventor: Roy Shkedi. App. No. 14/091,285; granted July 17, 2018; priority Nov. 28, 2000. (Google Patents, https://patents.google.com/patent/[US10026100](/patent/US10026100)/en)

I found no appellate (CAFC) decision on this patent, and no PTAB final written decision that I could confirm as targeting the '100 patent specifically.

Confirmed asserted cases

1. AlmondNet, Inc. v. Oath Holdings Inc. (Yahoo!), D. Del., No. 1:18-cv-00943

• Plaintiff: AlmondNet, Inc. — Defendant: Oath Holdings Inc.
• Jurisdiction: U.S. District Court for the District of Delaware (Judge Richard G. Andrews / "CFC" designation in early filings)
• Filed: June 26, 2018
• The '100 patent was among the patents asserted. This is corroborated two ways: (a) Roku's later declaratory-judgment complaint states the Oath suits "contained infringement allegations as to at least the '398, '639, '586, and '100 Patents" (https://insight.rpxcorp.com/litigation_documents/14313998); and (b) AlmondNet's own notice letter references "the First Amended Complaint in the Delaware lawsuit against Yahoo! … which likewise explain[s] in detail why DSP activity infringes U.S. Patents 9,830,615 and 10,026,100" (https://www.docketalarm.com/cases/Texas_Western_District_Court/7--25-cv-00165/...).
• Outcome: Case docket flags CLOSED. Per AlmondNet's own account, "Verizon agreed to settle the litigation, and Oath obtained a license to the AlmondNet Group patent portfolio" (same Amazon notice letter). Oath also filed many post-grant reviews against AlmondNet patents; AlmondNet says it defeated "every one of the challenges."

2. AlmondNet, Inc. v. Oath Holdings Inc., D. Del., No. 1:19-cv-00247

• Same parties; a second Delaware suit filed February 6, 2019 (referenced in Roku's DJ complaint as also containing '100 allegations). Confidence: moderate — I confirmed the case exists and is tied to the same patent cohort, but I did not independently pull that docket's patent list.

3. Roku, Inc. v. AlmondNet, Inc. and Intent IQ, LLC, D. Del., No. 1:21-cv-01035 (declaratory judgment)

• Plaintiff: Roku, Inc. — Defendants: AlmondNet, Inc.; Intent IQ, LLC (AlmondNet counterclaimed/consolidated as plaintiff)
• Jurisdiction: D. Del., Judge Maryellen Noreika; Wilmington
• Filed: July 15, 2021
• Patents at issue include: 10026100; 10321198; 10628857; 10715878; 7822639; 8244586; 8566164; 8595069; 8677398 (https://www.docketalarm.com/search/?q=&f=q_search_within-docket%3A(1%3A21-cv-01035)...)
• Outcome/status: Claim construction memorandum order May 29, 2024 (D.I. 193); stayed pending settlement April 22, 2025 (D.I. 244); stipulation of dismissal, case terminated June 20, 2025 (D.I. 252). Docket flags: CLOSED.

4. AlmondNet, Inc. and Intent IQ, LLC v. Roku, Inc., W.D. Tex., No. 6:21-cv-00731-ADA (Waco)

• Plaintiffs: AlmondNet, Inc.; Intent IQ, LLC — Defendant: Roku, Inc.
• Jurisdiction: U.S. District Court for the Western District of Texas, Waco Division; Judge Alan D. Albright
• Filed: July 15, 2021 (same day as the Delaware DJ action)
• Patents asserted include U.S. Pat. No. 10,026,100 (listed in the complaint and the AO-120 form) — https://www.docketalarm.com/cases/Texas_Western_District_Court/6--21-cv-00731/.../docs/29.pdf
• Outcome/status: Docket flags CLOSED, STAYED, PATENT. Joint motion to dismiss filed June 18, 2025 (D.I. 28); Order for Dismissal entered June 20, 2025 (D.I. 30). AlmondNet's claims dismissed with prejudice; Roku's defenses/counterclaims without prejudice; each party bears its own costs.

Not confirmed as asserting the '100 patent

AlmondNet/Intent IQ's other campaigns I surfaced — Meta (W.D. Tex. 6:21-cv-00896, filed Aug. 27, 2021), Amazon (W.D. Tex. 6:24-cv-00234, asserted only '445 and '904; and 7:25-cv-00165), Viant (D. Del. 1:23-cv-00174), FreeWheel/Beeswax (D. Del. 1:23-cv-00220), Lotame (D. Del. 1:24-cv-00376), Oracle (W.D. Tex. 6:24-cv-00303), Adobe (W.D. Tex. 7:25-cv-00487), and LinkedIn — asserted different family members ('398, '139, '639, '260, '146, '783, '582, '307, '249, '878, '445, '904, etc.). The '100 patent appears in AlmondNet's 2019–2025 notice letters to Amazon but not as an asserted claim in those pleadings.

Important caveats

• Google Patents itself lists three litigation family entries for US10026100: W.D. Tex. 6:21-cv-00731; D. Del. 1:21-cv-01035; and D. Del. 1:18-cv-00943 — matching the four cases above (with the 1:19-cv-00247 companion).
• I did not independently verify the 1:19-cv-00247 docket's patent list, and I did not determine whether any IPR/PGR petition (e.g., Roku's IPR2022-01236 or Samsung's IPR2022-01505) named the '100 patent as the challenged patent. Treat those as open questions rather than negatives.
• Litigation status is as of the most recent docket text I retrieved (June 2025 dismissals). Because the Roku dismissals were with prejudice as to AlmondNet's claims but involved settlement, there was no merits/validity ruling on the '100 patent in any case I found.

I'll research the PTAB history for this patent family.

PTAB Proceedings on US 10,026,100 — Analyst Report

Prepared: 2026-09-28 · Patent: US 10,026,100 B2 ("Methods and apparatus for facilitated off-site targeted internet advertising," AlmondNet, Inc., inventor Roy Shkedi; app. 14/091,285 filed 2013-11-26; granted 2018-07-17; earliest priority 2000-11-28)

Proceedings overview

There are zero AIA trial proceedings on file against US 10,026,100 — 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled after institution, and 0 institution denials. The canonical USPTO Open Data Portal block for this patent returns no IPR, PGR, or CBM proceeding, and my independent web searches surfaced no petition naming the '100 patent as the challenged patent (all AlmondNet PTAB activity I located targets sibling patents in the portfolio — chiefly the '398, '878, '615, '857, '307, '249, and '260 patents).

Bottom-line defensive posture: You cannot point to an IPR win because there isn't one. That cuts both ways. On the one hand, the '100 patent's own claims — notably independent claims 1 and 13 and their dependents — have never been tested by the Board, so there is no cancelation to hide behind and no § 315(e)(2) estoppel banked by anyone. On the other hand, the patent's asserted life is essentially over: Google Patents' structured record shows an adjusted expiration of 2023-01-23 with legal status "Expired – Lifetime" (the '100 is subject to a terminal disclaimer tying it to the family's chain from the 2000-11-28 priority filing). If that record is accurate, the only exposure today is pre-expiration damages subject to the 35 U.S.C. § 286 six-year lookback — not an injunction. Your realistic defenses are expiration + § 282 obviousness using the family-level PTAB record, not an IPR.

Proceedings on US 10,026,100

None found. No IPR, PGR, or CBM petition has ever been filed against this patent on the record I can access. Consequently, there are no judge panels, institution decisions, Final Written Decisions, settlements, or Federal Circuit appeals to report for the '100 itself. I am stating this as a negative finding rather than inferring it: the ODP block is silent, and no search result identified a proceeding docketed against patent number 10,026,100.

I will not manufacture proceeding numbers to fill this section. What follows is the adjacent record you must not confuse with the '100.

Adjacent proceedings in the same portfolio/family (NOT on the '100 — do not cite as '100 art)

These are the AlmondNet/Intent IQ PTAB matters that actually exist. They matter to a '100 defendant only as § 282 prior-art and obviousness ammunition and as a guide to which arguments the Board has already blessed or rejected. Verify each before filing.

Caveats I will not paper over: (i) I could not verify, from primary sources, the exact claim numbers and date for IPR2023-00384's FWD or the patent identification in IPR2022-01064 — treat those two rows as leads, not conclusions. (ii) AlmondNet has repeatedly stated that Yahoo!/Oath filed post-grant review petitions against "10 of the 11 patents" in the first Yahoo! action and that AlmondNet "defeated every one of the challenges." I could not confirm that the '100 was among those PGRs (the '100 issued 2018-07-17, and a PGR would have had to be filed within nine months, i.e., by roughly April 2019, while the Oath case was live). If a PGR2018/PGR2019 naming the '100 exists, it would have to be run down — this is the single most important open verification item.

Strategic summary

Claim status: everything is UNTESTED. No claim of US 10,026,100 has been canceled, confirmed, or even instituted upon. AlmondNet has asserted the '100 in litigation on a broad claim set — its preliminary infringement contentions against Roku (D. Del. 1:21-cv-01035) identified claims 1–11 and 13–14, and a Roku pleading recites AlmondNet's broader allegation of claims 1–3, 5, 7, 9–11, 13–17, 19, 21, 23–25, 27–31, 33, 35, 37–39, and 41–42. The independent claims that anchor any demand letter are therefore claim 1 (method) and claim 13 (system), plus whatever independent claims appear at 27/35/41 in the issued claim set. All of them are live on the face of the patent record. The countervailing fact is the 2023-01-23 adjusted expiration shown in the structured record — check this against the fee/terminal-disclaimer history in Patent Center before you rely on it, because if it holds it converts a validity fight into a damages-only, six-year-lookback fight.

Estoppel landscape: essentially empty, and that is good news for you. Because no IPR was ever instituted on the '100, no petitioner is estopped under § 315(e)(2) as to this patent, and no SAS-style all-claims/all-grounds narrowing applies. Family-level IPRs do not create statutory estoppel against a different patent, so a defendant sued on the '100 in 2026 faces no estoppel bar from the Meta/Roku/Samsung/Microsoft/Amazon IPRs on the '398 and '878. The flip side: Petitioners who instituted on sibling patents (Meta in IPR2022-00773; Roku in IPR2022-01236) burned their shot on those patents only. Discretionary-denial doctrine is likewise weak leverage against you — General Plastic and Fintiv cut against repeat filers, and a first-time petitioner on the '100 with no parallel advanced litigation (most Roku and Oath cases are now dismissed or stayed) is in a comparatively clean position procedurally.

Pattern signals: (1) The portfolio is a serial-assertion target, with defendants including Roku, Microsoft, Yahoo!/Oath, Samsung, Amazon, Meta, LinkedIn, Viant, and Oracle; litigation posture is highly settlement-driven (Roku stipulated dismissals; Meta terminated its instituted '398 IPR on settlement; Oath took a portfolio license). (2) No defensive aggregator appears to have driven an IPR against the '100 — Unified Patents' portal surfaces the patent only as prior-art context; the family IPRs were filed by operating companies (Roku, Meta, Samsung, Microsoft, Amazon), not by a standalone aggregator. (3) AlmondNet defends aggressively — the successful CBM-eligibility knockouts (CBM2017-00049/00050) and the Roku Director-Review denial show the patent owner winning at the threshold, not just on the merits.

Recommended next steps

If you are a defendant today:

• Lead with expiration, not invalidity. Confirm the 2023-01-23 adjusted expiration in USPTO Patent Center for 14/091,285 and the terminal-disclaimer chain. If confirmed, plead that prospective relief is unavailable and that damages are capped by 35 U.S.C. § 286 (six years back from the complaint). Contemporaneously check whether the '857 (US 10,628,857, the direct continuation of the '100) carries a later term — the '100's own front page notes it is "subject to a terminal disclaimer" in the related '857 document.
• Run an IPR anyway only if the patent is still within its term and the claim set you're accused of is claim 1/13-based. There is no estoppel and no prior institution to fight off. Budget the statutory one-year trial clock from institution (35 U.S.C. § 316(a)(11)) and note that § 315(b) gives you one year from service of the complaint — a hard bar, as Amazon learned in IPR2023-00227 (denied as time-barred).
• Reuse the family record for § 282 obviousness. The Board's institution papers and the Board's reasoning on the '398 and '878 — including the Roku denial on the merits and the Meta institution on the same specification family — are free expert-scaffolding. Because the '100 shares the specification lineage, much of the analysis transfers, and § 282 gives you no estoppel ceiling in district court for art the Board never reached on this patent.

If the patent is expired and you are only facing a demand letter:

Say plainly that claims 1–42 of the '100 have never been adjudicated invalid, but that the patent record shows expiration on 2023-01-23 and that any claim is limited to pre-expiration past damages. Do not assert that any claim "has been canceled" — that would be a fabrication.

Open verification items (do these before relying on this memo):

1. Run the '100 through PTAB E2E / PTAB Center (https://ptacts.uspto.gov/ptacts/) directly — the ODP feed may lag.
2. Search PGR2018-* and PGR2019-* for any Yahoo!/Oath petition naming the '100 (nine-month PGR window closing ~2019-04-17).
3. Pull the IPR2023-00384 FWD and the IPR2022-01064 FWD to fix the exact patents, claims, and dispositions before citing them.
4. Check CourtListener / CAFC for any appeal from those FWDs — I found none tied to the '100.

Sources consulted: structured Google Patents record for US 10,026,100 (https://patents.google.com/patent/US10026100/en); Docket Alarm PTAB dockets for IPR2022-01236, IPR2022-00773, IPR2023-00227, IPR2022-01420; PTAB institution decisions reproduced in AlmondNet v. Oracle, 6:24-cv-00303 (W.D. Tex.), D.I. 32-14; AlmondNet's preliminary infringement contentions served as Ex. 2012 in IPR2022-01236; D. Del. 1:21-cv-01035 and W.D. Tex. 6:21-cv-00731 / 6:21-cv-00731 AO-120 forms; AlmondNet v. Oath Holdings, 1:18-cv-00943 docket.

I'll research the assignment record for US 10026100. Let me run several targeted searches.

Let me dig for the specific assignment reel/frame records and correspondent details.

Let me search for more specific assignment record data and family transfer history.

I hit my tool-call budget before I could pull docket-level confirmation of the reel/frame for this specific patent, so I'll flag confidence levels explicitly rather than fill gaps with guesses.

Assignment / Ownership Analysis — U.S. Patent No. 10,026,100

Patent: "Methods and apparatus for facilitated off-site targeted internet advertising"
App. No.: 14/091,285 (filed 2013‑11‑26; divisional of 13/585,765 → 13/368,686 → 12/911,601 → 10/996,491 → 09/723,391, filed 2000‑11‑28)
Granted: 2018‑07‑17
Verification entry point: https://assignmentcenter.uspto.gov/ (mirror: https://assignment.uspto.gov/patent/index.html) — search by patent number 10026100 or app. 14/091,285.

Inventors

Roy Shkedi — sole named inventor.

• Employer/affiliation at time of filing: AlmondNet (founder/CEO). Shkedi founded AlmondNet in 1998 (with later subsidiaries Intent IQ and Datonics); he relocated to the U.S. in 2000, the same year the priority application (09/723,391) was filed. Address of record appears variously as Bnei‑Brak, Israel and Forest Hills, NY across the family.
• Unusual-pattern check: The "all inventors leave the assignee within 12 months" fire‑sale tell is not applicable and not present. This is a single‑inventor family, and the inventor is also the founder and controlling principal of the assignee. There is no separation of inventor from assignee. If anything, the inverse tell appears: a founder assigns into his own company and later builds a 170+‑patent portfolio on top of it (Shkedi interview, IAM/PTAB exhibit; https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1558243](/patent/1558243)/...). No inventor‑departure signal.

Original assignee

AlmondNet, Inc. — named on the face of the '100 patent and the sole recorded assignee.

• Entity type: Delaware corporation (per the inventor‑to‑AlmondNet assignment language, "sell, assign, and transfer to AlmondNet, Inc., a Delaware corporation…"). U.S. addresses of record: New York, NY; Long Island City, NY.
• Primary line of business: Behavioral‑targeting / targeted online advertising. AlmondNet is the original developer of the "off‑site targeted advertising" model that this patent claims (a publisher selling its audience to advertisers while the audience is on another site). The group operates subsidiaries Intent IQ (identity resolution / cross‑device) and Datonics (data). Self‑description: parent "focused on R&D and the licensing of its extensive portfolio."
• Did it ship a product embodying the claims? Yes, at least historically — AlmondNet was a behavioral‑targeting ad company and the '100 family originates in its commercial off‑site advertising offering. The group today markets productized offerings through Intent IQ. Note the nuance: the operating model is now dominated by licensing (see below), so product revenue is not the primary revenue line.
• Current status: Operating, not dissolved, not in bankruptcy. Founder Roy Shkedi remains CEO/Chairman. The group has monetized via licensing (a 2013 Google license — described by Shkedi as a take‑it‑or‑leave‑it deal made to avoid insolvency — and licenses/verdicts with Roku, Comcast, Samsung, Oracle, and Amazon; 2024 $122M jury verdict vs. Amazon on the '639/'139 patents, later settled in 2026). No Chapter 7/11 event found.

Assignment timeline

⚠️ Important limitation up front. The USPTO Assignment Center is JavaScript‑gated and I was not able to retrieve the reel/frame for the '100 patent's own record within my search budget. Google Patents' legal‑events feed for this patent surfaces exactly one post‑filing assignment record — verified in the authoritative page text:

"2014‑01‑24 — Assigned to ALMONDNET, INC. — reassignment — ALMONDNET, INC. — ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SHKEDI, ROY"

Because I could not open the Assignment Center entry, I am reporting the recorded date and parties (confirmed) but not inventing a reel/frame.

• 2013‑11‑26 (exec. date not separately confirmed) / recorded 2014‑01‑24 — Reel/Frame: not retrieved
  • Conveyance: Assignment (assignment of assignors' interest)
  • Assignor: Shkedi, Roy (sole inventor)
  • Assignee: AlmondNet, Inc.
  • Correspondent: Not confirmed on this record. The AlmondNet group's attorney of record across the portfolio is Louis J. Hoffman, P.C. (Reg. No. 38,918; USPTO Customer No. 26362; Scottsdale, AZ — formerly 14301 N. 87th St. Ste. 312, now 7689 E. Paradise Lane Ste. 2). He is the group's prosecution counsel and litigation lead counsel and is the likely correspondent here, but treat that as probable, not verified.
  • Context: Founder‑to‑own‑company assignment (confirmatory/housekeeping). It coincides with a batch of late‑2013/early‑2014 AlmondNet recordation activity (e.g., the same‑period loss‑of‑small‑entity filing dated 2013‑11‑08, and the family transfers below), consistent with a housekeeping sweep rather than a strategic transfer.

No further assignments found. Google Patents lists AlmondNet, Inc. as current assignee, and I found no recorded transfer of the '100 patent to Intent IQ, Datonics, or any third party. This is itself a finding: the original assignee still owns this patent.

Family‑level reel/frame context (NOT the '100 record — clearly labeled)

These reels appear on sibling patents in the same AlmondNet group and show how the group records title generally. They are not the '100 patent's chain, and I list them only as pattern evidence:

The 026952/0137 AlmondNet→Intent IQ transfer and Datonics's ownership of the '574 patent (CBM2017‑00059 mandatory notices) show the group does use subsidiary LLCs — but that pattern was not applied to the '100 patent, which stayed with the parent.

Timeline diagram

timeline
    title Ownership of US 10026100
    2000 : Inventor files original application
    2013 : Divisional application filed
         : Assigned to AlmondNet Inc
    2018 : Patent granted
         : Suit filed against Oath Holdings
    2021 : Suits filed against Roku
    2025 : Roku cases dismissed after settlement

NPE / troll-pattern signals

Verdict

Operating‑company assertion (with an explicit hybrid‑monetization caveat).

The ownership record for the '100 patent is a single assignment from sole inventor Roy Shkedi to AlmondNet, Inc., recorded 2014‑01‑24 (Google Patents legal events; reel/frame not retrievable), with no subsequent transfer — the patent is still held by the Delaware operating company the inventor founded in 1998. None of the classic troll‑chain signals fire: no shell‑LLC transfer, no cascading LLC chain, no pre‑litigation transfer, no bankruptcy sale, no privateering, no defensive‑aggregator termination. The one genuine recurrence (Louis J. Hoffman, P.C. as family correspondent) evidences long‑standing outside counsel, not a shell‑factory. Caveat for the record: AlmondNet Group's business model is now predominantly R&D‑plus‑licensing, and it is one of the most prolific patent plaintiffs in ad tech, so a purely functional "who monetizes this?" test could argue for NPE‑adjacent status — but the ownership chain itself is original‑owner assertion, not an NPE acquisition structure, which is the question this analysis is asked to answer.

Recommendation before relying on this: pull the Assignment Center entry for 10026100 to capture (a) the exact reel/frame of the 2014 recordation, and (b) the correspondent of record on that instrument, to confirm or refute the Hoffman inference. URL: https://assignmentcenter.uspto.gov/

Cross‑reference note

No contradiction with the previously generated Litigation summary: that section likewise found no merits/validity ruling on the '100 patent and no CAFC decision, and its "AlmondNet remains the asserting owner" framing is consistent with the single‑assignment ownership record here. The hybrid‑monetization caveat above is meant to sharpen, not contradict, that summary.

I'll research the prior-art citations for US 10,026,100. Let me run several targeted searches in parallel.

Let me retrieve the actual front-page references cited on the '100 patent and its file history.

Prior Art for U.S. Patent No. 10,026,100 — Reference-by-Reference Analysis

0. Scope, sources, and confidence (read this first)

What I was asked to do vs. what I could verify. I attempted to pull the verbatim front-page "References Cited" list from the USPTO full-text database and Google Patents for the literal number 10026100 (not any similar number). My search budget was exhausted before I could retrieve the complete verbatim list from a primary database page. The reference list below is therefore reconstructed from three authoritative reproductions of the '100 patent itself and its litigation record:

1. The printed front page of the '100 patent as filed in court — AlmondNet, Inc. v. Oath Holdings Inc., D. Del. 1:18‑cv‑00943, Doc. 9‑2 (Exhibit E), which reproduces US 10,026,100 B2 at pp. 2–3 (https://www.docketalarm.com/cases/Delaware_District_Court/1--18-cv-00943/AlmondNet_Inc._v._Oath_Holdings_Inc/docs/9/2.pdf).
2. FreePatentsOnline's "US Patent References" field for 10026100 (https://www.freepatentsonline.com/10026100.html).
3. The closely related continuation US 10,628,857 (same title, same inventor, filed 2018‑07‑11 as a continuation of the '100 patent and sharing its "References Cited" lineage), as reproduced in AlmondNet v. Roku, W.D. Tex. 6:21‑cv‑00731, Doc. 1‑14 (https://www.docketalarm.com/cases/Texas_Western_District_Court/6--21-cv-00731/AlmondNet_Inc._et_al_v._Roku_Inc/docs/1/14.pdf).

Consequences of that limitation — please treat as explicit uncertainty:

• The list below is near-complete but not certified verbatim; issue dates come from the OCR'd front-page fragments and may contain OCR digit errors (the Google-hosted OCR of this family visibly garbles several names and one patent number — e.g., "Bleveux"/Blewett, "Oswann", "Munteverde"/Monteverde, "Gilford"/Gifford, "5,870,556"/"5,870,550"). I have not auto-corrected these; where I flag a probable OCR error I say so.
• I could not retrieve the '100 patent's prosecution history (USPTO PatentCenter / public PAIR) in this session, so I cannot confirm which of these references the Examiner actually applied in a §102 or §103 rejection, as opposed to merely listing as of record. That distinction matters enormously and is discussed in §4.
• Date conflict flag: this task states the current date is April 26, 2026, whereas my system date is 2026‑09‑28. As already flagged in the previously generated summary, this does not change any bibliographic data and I proceed on the bibliographic facts independently.

1. Critical framing: most "citations" on the '100 patent are the applicant's own patents

The single most important analytical point about this reference list is that the overwhelming majority of the examiner-cited U.S. patents are Roy Shkedi's / AlmondNet's own patents. Examples visible on the '100 front page (Doc. 9‑2, p. 3) and in FPO's list:

7,085,682; 7,428,493; 7,454,364; 7,747,745; 7,822,637; 7,861,260; 7,890,609; 8,051,444; 8,204,822; 8,204,788; 8,204,965; 8,239,264; 8,244,574; 8,244,582; 8,244,586; 8,244,588; 8,280,758; 8,281,336; 8,341,747; 8,494,904; 8,566,164; 8,589,210; 8,595,069; 8,607,267 — all Shkedi/AlmondNet.

Why this matters for a §102 question: a reference patent or application is prior art under §102(a)/(e) only if it is "by another." Where Shkedi is the sole or common inventor (and the '586/'582/'574/'822/'965/'264/'758/'336/'747/'904 members are in the same priority chain), these are not §102(a) or §102(e) art against claim 1 of the '100 patent. They may appear in the "References Cited" block as a matter of examiner diligence / double-patenting hygiene, not as anticipation. The earlier-generated bibliographic section already noted the '100 patent carries a terminal disclaimer.

Therefore the references that can actually bear on §102 are the third-party/non-Shkedi entries, set out below.

2. Examiner-cited U.S. patent documents (third-party / non-Shkedi)

Dates are issue dates as they appear in the OCR'd front page. "Desc." is my working understanding of the reference's subject matter; an asterisk (*) marks descriptions I could not independently verify in this session — treat those as provisional.

3. Foreign documents cited

Note: WO 97/21183 appears in the front-page list of the sibling US 10,628,857; I could not confirm in this session that it appears on the '100 front page itself. Treat as provisional.

4. §102 legal analysis — what these references can and cannot do

Threshold point: because these references are of record on a granted patent, the Examiner did not treat any of them as anticipating. A §102 anticipation case for claims 1–14 would have to be made de novo by a challenger, and the two §102 candidates that come closest fall short of the claim-1 preamble/steps in identifiable ways:

Claim 1 (method; asserted against Roku as claims 1–11, 13–14). The claim requires, in combination:

• (preamble) the first-site proprietor has already sold off-site ad placements for a first price;
• (a) first computer places a cookie on visitor computers;
• (b) detects a specific predetermined action by those visitors on the first site;
• (c) stores cookie-reachable information recording that action;
• (d) with a second computer controlling ad space on a second site, detects the cookie-bearers arriving there and identifies the action-performers;
• (e) causes the ad to be served for display on the second site;
• (f) causes the second site's proprietor to receive revenue less than the first price; and
• (g) the first site keeps at least part of the difference.

Mapping the strongest candidates:

Claim 15 (independent; verbatim text not retrievable here — see the earlier summary's flag) and claim 29 ("first computerized apparatus controlled by a proprietor of a first Internet site" + parallel second apparatus). Both re-recite the same operative sequence, so the same references map the same way and fail on the same two-price/retained-spread limitations (claims 1(f)–(g)/29 equivalents). On claim 29 the analysis is additionally weakened because the reference must disclose a structural first apparatus controlled by the proprietor of the first Internet site — even Merriman's ad-server architecture does not disclose that particular ownership/control configuration.

Net conclusion on §102: on the record I could retrieve, none of the cited references anticipates any of claims 1–14 or 15–42. The closest single-reference candidates are Montulli '670 (for the cookie limitations) and Merriman '061/'637 (for the ad-serving limitations), and even combined they would be a §103 rather than §102 theory. The discounting/retained-spread revenue limitations (1(f)–(g)) were the express point of novelty on which the family distinguished the earlier art (Goldhaber-style attention brokerage and the Roth/Engage material referenced in the '639 file history), and none of the cited references discloses them.

5. Prior art actually asserted against the '100 patent in litigation (non-examiner-cited, but the real §102 battleground)

Because the front-page list is largely the applicant's own patents, the references with real §102 bite are those charted by defendants:

• Roku, Inc. — Appendix 1 ("State of the Art References for Asserted Patents") filed with Roku's Invalidity Contentions in the consolidated Delaware action (USPTO PTAB document server, Roku/Intent IQ filing): https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1555761](/patent/1555761)/... . It lists, inter alia, U.S. 5,588,119 (Vincent), 5,724,424 (Gifford), 5,740,252 (Minor), 5,754,939 (Herz), 5,787,411 (Groff), 5,796,393 (MacNaughton), 5,826,241 (Stein), 5,918,014 (Robinson), 5,933,811 (Angles), 5,945,988 (Williams), 5,948,061 (Merriman), 5,960,409 (Wexler), 5,991,740 (Messer), 5,999,971 (Buckland), 6,009,409 (Adler), 6,014,638 (Burge) — all asserted as §102(a)/(b)/(e) art.
• Meta Platforms — "Exhibit G," Defendant's Preliminary Invalidity Contentions in AlmondNet v. Meta, W.D. Tex. 6:21‑cv‑00896 (reproduced in IPR2022‑01262): https://www.docketalarm.com/cases/PTAB/IPR2022-01262/Meta_Platforms_Inc._v._AlmondNet_Inc/docs/11-08-2022-Patent_Owner/Exhibit-2012-Exhibit_2012.pdf . It includes U.S. 6,973,436 (Shkedi) and U.S. 7,861,260 (Shkedi) among the charted items — note the irony that even the defendants' charts lean on the Shkedi family.
• The prior-art "Roth/Engage" material the applicant overcame during prosecution of the '639 precursor (as referenced in the earlier-generated summary) is the truly substantive §102/§103 challenge to this family — but I could not retrieve its identifying citation in this session and therefore decline to characterize it.

Result of that litigation: the Roku matters terminated by settlement/stipulated dismissal (June 2025), so there is no merits or validity ruling on claims 1–14 of the '100 patent; and the '100 patent was not, to my knowledge, the patent instituted in an IPR. That is consistent with the § 4 conclusion.

6. What I could not verify (explicit gaps)

1. The complete verbatim "References Cited" list on the '100 front page, and whether WO 97/21183 / any non-patent literature citations appear on the '100 patent itself rather than only on the sibling '857.
2. The '100 patent's prosecution history — i.e., which of these references the Examiner actually applied in a §102/§103 rejection, and the examiner's statement of reasons. This is the single most important missing input for a defensible §102 opinion.
3. Verbatim claim 15 (flagged in the earlier summary as unresolved) and verbatim claim 29 (reconstructed). Any §102 mapping to those claims is provisional.
4. Descriptions marked with (*) in the tables above are from general knowledge and were not verified against the references themselves in this session; several OCR name/number corruptions in the Google-hosted copies of this family remain un-reconciled (e.g., "Bleveux"/Blewett; "Oswann"; "Munteverde"; "Gilford"/Gifford; "5,870,556"/"5,870,550").

Primary record links: Google Patents https://patents.google.com/patent/US10026100/en · FreePatentsOnline https://www.freepatentsonline.com/10026100.html · USPTO PatentCenter https://patentcenter.uspto.gov (for the certified front page and file wrapper).

I'll ground this obviousness analysis in the actual prior-art references cited against this family. Let me search.

Let me pin down the specific prior-art references and the grounds actually asserted against this family.

Let me verify the primary pre-2000 references precisely before building the grounds.

Obviousness Analysis — U.S. Patent No. 10,026,100 under 35 U.S.C. § 103

Scope note. The claim-1 limitation labels used below ([1P], [1a]–[1g]) are my reconstruction from the earlier "Patent summary" section of this analysis, cross-checked against the asserted-claim list (claims 1–11, 13–14) in AlmondNet's preliminary infringement contentions and Roku's invalidity contentions. I did not independently retrieve verbatim claim 1 or verbatim claim 15 in this task, so treat the wording of each limitation as characterization, not quotation. The legal conclusions below are analytical opinions, not legal advice.

1. Contradiction flagged: the primary reference is mislabeled in the earlier section

The earlier "Patent summary" section describes the prosecution amendment as distinguishing "the Roth/Engage prior art" and the earlier "Litigation summary" speculates about "Merriman (US 5,855,008)."

My verification search produced a direct conflict with both:

• US 5,948,061, Method of delivery, targeting, and measuring advertising over networks — inventors: Dwight Allen Merriman and Kevin Joseph O'Connor; applicant/assignee: DoubleClick, Inc.; application US 19960738634, filed 1996‑10‑29, published 1999‑09‑07 (Espacenet bibliographic record, https://worldwide.espacenet.com/publicationDetails/biblio?CC=US&NR=[5948061A](/patent/5948061A); Google Patents, https://patents.google.com/patent/[US5948061A](/patent/US5948061A)/en).

So "Merriman" is the DoubleClick DART reference (US 5,948,061), not US 5,855,008. And US 5,855,008 is a Goldhaber et al. reference — it appears as 5,855,008 A 12/1998 Goldhaber et al. on the face of Shkedi's own US 7,428,493 (D. Del. 1:19‑cv‑00247, Doc. 109‑16, https://www.docketalarm.com/cases/Delaware_District_Court/1--19-cv-00247/AlmondNet_Inc._et_al_v._Oath_Holdings_Inc/docs/109/16.pdf). I could not complete a separate bibliographic verification of US 5,855,008 before exhausting my search budget, so I rely on that reference list plus training data.

I could not verify the existence of a distinct "Roth" reference in this family's prosecution. I therefore treat the earlier section's "Roth/Engage" label as unverified shorthand, and I build the analysis on the verified US 5,948,061 (Merriman/O'Connor, DoubleClick) plus the Engage Technologies S‑1 and Acxiom materials that actually appear in the family record. If a separate "Roth" reference exists, the analysis below should be re-run against it.

2. Governing law and framework

Pre‑AIA § 103(a). The '100 patent's earliest priority is 2000‑11‑28 (application 09/723,391). The application that issued as the '100 patent (14/091,285) was filed 2013‑11‑26, but because every application in the chain was filed before 2013‑03‑16 and claims that 2000 priority, the pre‑AIA statute and case law apply (AIA § 3(n)(1)). This matters substantially: pre‑AIA § 102(e) prior art is limited, and there is no IPR eligibility for pre‑AIA patents over one year old — which is exactly why Roku's and Meta's IPR art was routed through the later family members ('878, '139, '398, '260) rather than the '100 patent.

KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), supplies the operative rationales: (i) combining prior-art elements according to known methods to yield predictable results; (ii) simple substitution of one known element for another; (iii) use of a known technique to improve similar devices in the same way; (iv) application of a known technique to a known device ready for improvement; and (v) "obvious to try" where design incentives and market forces drive the variation. The Graham v. John Deere factors (scope/content of art, differences, PHOSITA level, secondary considerations) frame the analysis.

Critical date rule. All prior art relied on below must predate 2000‑11‑28. This eliminates most of the art the Board and courts actually litigated against the AlmondNet family, as discussed in §7.

3. Level of ordinary skill in the art (PHOSITA)

The Board adopted (and Patent Owner did not contest) a formulation in the family IPRs: a bachelor's degree in electrical engineering, computer engineering, or computer science, plus two years of work experience in network-based computer systems, including systems for sending and receiving information over the Internet/World Wide Web. A parallel petition version adds "data communications over the Internet (including Internet advertising) and database management" (Roku petition, https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1555164](/patent/1555164)/download-documents; IPR2022‑01259 Exhibit 2007, https://www.docketalarm.com/cases/[PTAB](/ptab)/IPR2022-01259/Samsung_Electronics_Co._Ltd._v._AlmondNet_Inc/06-02-2023-Patent_Owner/Exhibit-2007-Exhibit_2007/).

For a 2000 priority date, this artisan knows: HTTP cookies and third-party cookie mechanics; client-side redirects and embedded 1×1 GIF "ad calls"; ad network / "rep firm" economics and CPM pricing; and the state of the art represented by DoubleClick's DART, Engage, 24/7 Media, and AdForce.

4. The prior-art references, and what each discloses

4.1 Primary: US 5,948,061 — Merriman & O'Connor (DoubleClick), filed 1996‑10‑29, issued 1999‑09‑07

This is the single most damaging reference, and it is squarely within the critical date. Its disclosure, per the Google Patents and Espacenet records:

The one genuine gap: US 5,948,061 locates the recognition/targeting logic in a central ad server node, whereas claim 1 requires the first site's own computer (the "first computer controlled by a proprietor of a first Internet site") to place the cookie, detect the action, store the cookie-reachable information, and later identify the visitor at the second site. Also absent is the specific two-price / first-site-retains-the-spread structure of [1f]–[1g].

Those two gaps are then closed by the secondary references and by the express teachings embedded in US 5,948,061 itself.

4.2 US 5,855,008 — Goldhaber et al., Dec. 1998 ("attention brokerage")

Cited on the face of Shkedi's own US 7,428,493. An "attention brokerage" teaches the core economics claim 1 recites: a broker acquires the attention of an audience from one party and resells it to advertisers, retaining a spread. It supplies the [1f]–[1g] revenue-split concept (second party paid a fee; broker keeps the difference).

4.3 US 5,848,396 — Gerace, Dec. 1998

"Method and apparatus for determining behavioral profile of a computer user." Teaches building a behavioral profile of an individual user from observed interactions with a web site (pages viewed, time spent, responses to content) and using that profile to select advertising. This directly supplies:

• [1b] detection of a specific predetermined action on the first site;
• [1c] storage of the resulting characterization;
  and makes claims 2–5 (time-based, question-answer, section-interest, "reaction to predetermined aspects of the site") predictable design choices.

4.4 US 5,710,884 & US 5,717,923 — Dedrick, Jan./Feb. 1998

Targeted electronic promotion delivery based on a stored consumer profile in a database, updated from prior interactions/transactions. Supplies claim 7's "stored in a database linked to the cookie."

4.5 US 6,216,129 — Eldering, Apr. 2001

Targeted insertion of advertisements into media streams based on subscriber profiles (illustrated across satellite, cable TV, and wireless networks). Relevant to claim 1's web-television dependencies (claims 23/24 in the earlier characterization) and to any §103 theory that the off-site placement can occur in a "second information-media" other than a web page. (Prior art by virtue of its 1997‑era filing under pre‑AIA § 102(e); the issue date postdates 2000, so verify the 102(e) date before relying on it.)

4.6 US 6,055,573 — Gardenswartz; US 5,956,693 — Geerlings; US 6,055,513 — Katz

Internet-based interactive advertising with user profiles; targeted advertising systems; and automated negotiation of transactions (commissions/spread). Katz is particularly apt for the "retains at least part of the difference" concept as an automated negotiation between a buyer and seller of media.

4.7 Printed publications: Zeff & Aronson, Advertising on the Internet (2d ed. 1999); Engage Technologies S‑1 (July 19, 1999); Acxiom web pages (Feb.–Nov. 1999); Morgan Stanley Dean Witter e‑Marketing Report (May 2000)

These are the industry-state-of-the-art materials that appear in the family record (the '493 reference list at D. Del. 109‑16 lists Zeff's Advertising on the Internet, the Engage Technologies S‑1A filed 1999‑07‑19 at pp. 36–50, www.acxiom.com archived 1999‑02‑24 to 1999‑11‑14, and a Morgan Stanley Dean Witter e‑Marketing Report). A POSITA reading these would have known that (a) CPM is the standard pricing unit, (b) ad networks pay affiliates a share of revenue, and (c) premium ad inventory was routinely oversold, creating demand for secondary placements.

4.8 Trap to avoid: the patent's own "US Patent References" list

The FPO "US Patent References" section on the '100 patent is dominated by Shkedi's own patents (7,979,307; 8,244,586; 8,566,164; 8,671,398; 8,600,815; etc.). Under pre‑AIA § 102 these are not prior art — they are not "by another," and they share inventorship/ownership. Any obviousness theory built on them is defective. The only uses of that list are (i) to establish the field's conventionality and (ii) as § 102(e) art only if a Shkedi document has a different inventive entity, which is not the case here.

5. Grounds of rejection

Ground 1 — Claim 1 obvious over Merriman (US 5,948,061) in view of Gerace (US 5,848,396) and Goldhaber (US 5,855,008)

Mapping. Merriman supplies [1P] (off-site targeted Internet advertising, cookie-based, cross-site), [1a] (cookie placement), [1d] (recognition of the previously-seen visitor when the browser contacts the ad server from the second, affiliate site), [1e] (causing the selected ad to be served into the affiliate site's ad space), and the fee-to-the-affiliate mechanism for [1f]. Gerace supplies [1b]–[1c] (a specific predetermined action on the first site detected and stored as profile data). Goldhaber supplies the brokerage economics for [1f]–[1g] (a party acquires audience access and resells at a margin, retaining the spread).

Motivation to combine.

1. Same field, same problem. All three address delivering targeted advertising to identified Internet users. Merriman expressly frames affiliates as "entities that generally for a fee contract with the entity providing the advertisement server," creating both the technical and the commercial architecture claim 1 recites.
2. Express teaching toward the modification. Merriman itself teaches that the affiliate supplies "a link message to said user node identifying the IP address of said advertisement server node" (RE44,899 claim 13). Who holds and acts on the cookie-reachable state is therefore a known, arbitrary and relocatable implementation choice — moving the recognition logic to the first site's server is the classic KSR "simple substitution of one known element for another" and "predictable variation."
3. Business-method rationale. The patent's own specification states the motivation in terms indistinguishable from the art: "a web site with 20% of sold ad space and 80% unsold ad space might prefer to let its advertisers reach its audience outside of its web site in return for a lower price... the number of pages the visitor will read outside the site are bigger than the number of the reader will read within the site" (US10026100 Description). A motivation stated in the applicant's own specification cannot simultaneously be the thing that makes the claim nonobvious.
4. No new technology required. The specification repeatedly says the invention uses "generally existing, known modules and technology," and that it makes "the number of pages the visitor will read outside the site" monetizable. KSR at 417 ("if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious").

Ground 2 — Claim 1 and claim 7 obvious over Merriman in view of Dedrick (US 5,710,884 / 5,717,923) and Goldhaber

Dedrick teaches a maintained consumer-profile database updated from prior interactions and used for later targeting — the [1c] "stores information reachable through the cookie" and claim 7's "in a database linked to the cookie." Merriman already teaches a persistent user database queried on each ad call. Motivation: database lookup on cookie retrieval is the very mechanism Merriman uses; Dedrick merely confirms the conventionality of storing profile state in a database keyed to a persistent identifier.

Ground 3 — Claim 1's commercial limitations obvious over Merriman in view of the Engage S‑1, Acxiom, and Zeff

For [1f] "causes the second site's proprietor to receive revenue ... less than the first price" and [1g] "keeps at least part of the difference":

• Merriman: affiliates receive a fee; the ad server is the monetizing intermediary.
• Zeff: establishes CPM as the standard unit and that ad inventory is bought and resold across sites.
• Engage S‑1 / Acxiom: documents that the market priced audience access differently across contexts, and that a network or rep firm could sell the same audience at a different price depending on whether the placement was on the audience's "home" site or elsewhere.
• KSR market-forces rationale: price discrimination between channels for the same underlying audience access is a classic market-driven variation. "[W]here a defendant merely combines two prior art elements according to known methods to yield predictable results," the claim is obvious. The incremental "difference" (the spread) is arithmetic, not inventive.

Ground 4 — Alternative: Merriman + Gardenswartz + Katz

Gardenswartz supplies Internet-based interactive advertising with user profiles and real-time selection; Katz supplies automated negotiation of a transaction spread, useful if Patent Owner argues the "difference retained" element requires an automated accounting step.

Ground 5 — Dependent claims 2–14 and 16–28, 30–42

Under In re Keller / In re Merck, a combination is obvious where the references teach or suggest all claimed elements and the combination is motivated. Nothing in the '100 specification recites any unexpected result specific to any dependent claim.

6. Secondary considerations

• Nexus failure. Any commercial success (the Google license referenced in the IAM interview, https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1558243](/patent/1558243)/download-documents; the June 2024 Amazon verdict) attaches to other family members — '639 and '139 — not to the '100 patent's specific claims. Under In re GPAC / Wm. Wrigley, a presumption of nexus requires the commercial product/patent to embody the asserted claims; here the asserted claims' distinctive feature (the two-price/spread structure) is a business arrangement, not a technical improvement, which weakens the nexus.
• Copying / industry praise would be the strongest counterweight, but the record I located does not supply it for the '100 patent specifically.
• Terminal disclaimer on the '100 patent means its claims are not patentably distinct from the earlier-expiring family members — a fact that cuts against independent patentable significance of the '100 claims relative to the family's earlier disclosure.

7. What is not available, and why this matters

A reader of the modern AlmondNet litigation record will be tempted to transplant the Board's grounds into the '100 patent. They do not fit on the date. The IPR art litigated against the family is post‑2000 or barely pre‑2000:

• Baig (US 2008/0113674 A1, filed 2007‑10‑31, published 2008‑05‑15) — the backbone of the Meta and Samsung grounds against the '878, '398 and '260 patents (IPR2023‑01281, https://www.docketalarm.com/cases/PTAB/IPR2023-01281/Meta_Platforms_Inc/docs/08-04-2023-Petitioner/Petition_as_filed-2-Petition__as_filed.pdf; IPR2025‑00129 against the '260 patent). Not prior art to a 2000 priority.
• Beyda (US 9,928,522, filed Dec. 17, 2004) and Narin (US 7,039,699, filed May 2, 2000) — the grounds in IPR2022‑01319 against the '139 patent, denied under § 325(d) (https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1558069](/patent/1558069)/download-documents). Beyda is 102(e)-ineligible against a 2000 priority; Narin's May 2000 filing date and any pre‑2000-11‑28 priority filing would need separate verification.
• Rosenberg ("tracking a web browser across distinct domains") — Roku's IPR ground, said to be uncited during prosecution (https://ptacts.uspto.gov/ptacts/public-informations/petitions/1555164/download-documents). I could not verify its filing date; if it postdates 2000‑11‑28, it is unavailable against the '100 patent.
• Madhavan, Baum, Damick — Roku's grounds against the '398 patent, which the Board refused to institute (bannerwitcoff.com PDF, IPR2023‑00227, denial on § 315(b) time bar; https://bannerwitcoff.com/wp-content/uploads/2023/06/IPR2023-00227.pdf). Also date-sensitive.
• Costa, Zwicky, Howcroft, Middeljans, Koningstein, Banga690, Appelman, Satapati, Hahn — all post‑2000 in the family IPRs.

Practical consequence: the '100 patent is harder to invalidate than its siblings because the pre‑2000 art base is narrower, and because the Board has already declined the family's modern art on § 325(d) and time-bar grounds. The strongest § 103 case against the '100 patent runs through Merriman/DoubleClick (US 5,948,061) — the same reference the applicant amended around during the '639 prosecution — rather than through Baig.

Aligning evidence: Amazon's invalidity case against the family relied on the Engage system as an anticipatory/obviousness reference (AlmondNet's own motion in limine "TO EXCLUDE ARGUMENT AND EVIDENCE THAT THE ENGAGE SYSTEM ALONE RENDERS OBVIOUS CLAIM 37 OF THE 139 PATENT," AlmondNet v. Amazon.com, 6:21‑cv‑00898, https://www.courtlistener.com/docket/60319431/almondnet-inc-v-amazoncom-inc/). And the Board did invalidate four challenged claims of an AlmondNet ad patent as obvious in the Meta/Amazon proceeding (Nov. 30, 2023, https://www.law360.co.uk/ptab_cases/62973745aaca57019e2045a3) — as to the '398 patent, not the '100 patent.

8. Where Patent Owner likely survives, and my confidence

Relatively strong PO positions:

1. No single reference discloses the first site's own computer performing post-cookie, off-site visitor identification while the first site's proprietor retains the price spread. That is a real, if narrow, gap; the combination must be supported by a specific articulation of why the artisan would relocate Merriman's central-server logic to the affiliate's own server. The best support is Merriman's own "link message"/affiliate-fee teaching plus Zeff's repping model — but a district court could find that insufficient absent a strong declaration.
2. § 325(d) risk. Both Beyda and Narin were before the Examiner, and much of the DoubleClick/Engage/Acxiom material appears in the family file histories (the '493 IDS, the '639 file-history excerpts). A petition built on art already of record risks discretionary denial — a procedural, not substantive, shield.
3. Written description / enablement of the broad claim, given the 2000 specification is largely business prose with no algorithms. That cuts both ways: it is an invalidity theory under § 112, but it also shows the claimed subject matter is at the level of arrangement of known components, which supports § 103.

Verbatim-text caveat. Until claim 1's actual language (and claim 15's) is quoted from the patent, do not assume that "(a) places a cookie on each of many visitor computers" is literally recited; if the claim instead recites "causes a cookie to be placed," some § 103 mappings change.

9. Explicit uncertainties

1. Claim 15 and claim 29 verbatim text — not retrieved; the earlier section flagged this too. Any § 103 mapping to claim 15/29 is provisional.
2. "Roth" reference identity — I could not locate a distinct US "Roth" reference in this family. The verified primary reference is US 5,948,061, Merriman & O'Connor, DoubleClick (filed 1996‑10‑29, issued 1999‑09‑07). This contradicts the earlier section's "Roth/Engage" attribution and its speculation that Merriman = US 5,855,008. Flagged per the operating rules; the earlier section should be corrected.
3. US 5,855,008 bibliographic details — corroborated only via the '493 front-page reference list in the record (5,855,008 A 12/1998 Goldhaber et al.), plus training data. Verify independently before filing.
4. US 6,216,129 (Eldering) — its 2001 issue date means reliance requires its pre‑2000‑11‑28 § 102(e) filing date to be established.
5. No confirmed IPR, PGR, or FWD on the '100 patent. IPR2022‑01236 (Roku v. Intent IQ), cited in the earlier sections, remains unverified as to whether it named the '100 patent; my searches did not confirm it.
6. Date discrepancy, restated. The task header says "Current Date: April 26, 2026," while the system context says 2026‑09‑28 and the earlier section also flagged the mismatch. This does not change any pre‑2000 prior-art date analysis.
7. Litigation status is as of the June 2025 Roku dismissals; the W.D. Tex. 6:21‑cv‑00731 dismissal was with prejudice as to AlmondNet's claims and without prejudice as to Roku's counterclaims, i.e., no validity adjudication on the '100 patent was reached.

10. Bottom line

The strongest § 103 case against US 10,026,100 is:

Merriman & O'Connor, US 5,948,061 (DoubleClick, 1996/1999), in view of Gerace, US 5,848,396 (1998), in further view of Goldhaber et al., US 5,855,008 (1998) — with Dedrick (US 5,710,884/5,717,923) for the database-dependent claims and Zeff/Engage S‑1/Acxiom for the CPM and revenue-split claims.

Merriman teaches nearly every technical element of claim 1 — cross-site cookie identification, behavioral profiling, off-site targeted ad delivery into an affiliate's ad space, and payment of a fee to the affiliate. Gerace and Goldhaber close the two remaining gaps: (a) detection and storage of a specific predetermined action as a targeting condition, and (b) the two-price/spread economics. The motivation is supplied by (i) Merriman's own affiliate/link-message architecture, (ii) the well-documented 1998–2000 industry practice of reselling audience access at differential CPMs, and — decisively for KSR purposes — (iii) the specification's own admissions, which describe the invention as an arrangement of "generally existing, known modules and technology" motivated by ordinary advertising economics. The most likely survival points for Patent Owner are the "first-site-performs-the-off-site-identification" relocation and the § 325(d)/procedural posture, not the technical content of the claims.