Markus in practice

Real sessions. Real USPTO data. Nothing invented.

Five working sessions with Markus, condensed — TTAB strategy, clearance, office actions, and filing craft. Each one links to the raw, unedited conversation, and every serial number is checkable against the register.

See plans & pricing Download as PDF

01TTAB strategy · adversary intelligence

It scouted the opposer's litigation record before the opposition was even filed.

An attorney's client — a startup with a pending application for BEASTFUEL (energy drinks, Class 32) — just received an extension of time to oppose from Monster Energy. Is this a bluff, or incoming fire?

My client's BEASTFUEL application (Class 32, energy drinks) just drew an extension of time to oppose from Monster Energy. How seriously should we take it — do they follow through, and what do they actually target?

Attorney

Take it seriously

Short answer: take it seriously. Monster Energy Company is one of the most active trademark enforcers on the TTAB docket, and its mark is broadly famous — including specifically in Class 32 energy drinks, which is your client's exact category.

Do they follow through? Since 2002 the enforcement record shows 240 extensions of time to oppose, 184 actual oppositions, and 20 cancellations — extensions are frequently just the first procedural step before Monster actually opposes, not a bluff they let lapse. Recent oppositions they've pursued to filing:

Target markSerialProceeding status
FEAST LIKE A BEAST98396388Opposition suspended
VIVE100% ULTRA98784666Opposition pending
SEA MONSTROSITY99618279Opposed · terminated
UNLITIGATED SEA MONSTER99188638Opposed · terminated

Notably, FEAST LIKE A BEAST shows Monster has actively opposed marks built around aggressive/predatory-animal branding — exactly the naming territory "BEASTFUEL" sits in.

Show full response

Why it's especially relevant here: MONSTER ENERGY reads as broadly (dilution-tier) famous, and is confirmed famous specifically in Class 32 — the same class as BEASTFUEL. That fame both (a) broadens Monster's §2(d) scope of protection well beyond literal word overlap, and (b) opens a potential §43(c) dilution angle even if a likelihood-of-confusion argument is thin.

Recommendation: don't treat this as a low-probability placeholder. I'd run a full DuPont-style confusion analysis on BEASTFUEL against Monster's core marks now — before the extension converts to an opposition. Want me to pull Monster's core registrations and run that comparison?

Markus · live run Jul 16, 2026 get_owner_ttab_enforcement, is_mark_famous 61 secondsRead the unedited session

Why this needs the TTAB docket

A general model has no TTAB docket. It cannot pull an opposer's real extension-to-opposition record, cite the specific marks they've actually dragged to the Board, or read a fame signal in your client's class. Markus did it in 61 seconds — with serial numbers you can check.

Enforcement counts as returned by Markus's Class 32–scoped enforcement tool; Monster's full all-class TTAB docket is larger still (2,300+ proceedings as plaintiff). All four opposed marks independently verified against TTAB records, July 2026. Client scenario fictionalized.

02TTAB defense · crowded-field evidence

The opposer claims exclusivity. The register says thirty-seven owners.

Defending an opposition, the attorney needs to show the opposer's three-letter mark APS is weak and entitled to only narrow protection. Markus turns a hunch into an evidence table.

Opposer claims exclusive rights in APS. Run a crowded-field search — how many third parties actually hold live APS registrations, and in what classes?

Attorney

50 live marks · 37 owners

APS is about as crowded as a three-letter mark gets — an exclusivity claim is very hard to sustain on this record. Standalone APS: 50 live marks held by 37 distinct owners. That blows past the standard crowded-field threshold (roughly 6+ independent owners). A sample of the genuinely unrelated third parties:

OwnerClass(es)Goods/Services
Arizona Public Service Company35, 39, 40, 42Electric utility services
Biomet Manufacturing, LLC10Blood-collection kits
APS Limited28Airsoft / toy guns
"APS Trading" OOD7, 9, 16, 37Thermoprinters, circuit boards
Akron Polymer Systems Inc.42Polymer R&D
The American Pediatric Society35, 41Pediatric education

Distinct owners hold standalone APS registrations across 18 Nice classes — with Classes 35, 9, 7, and 42 the most contested.

Show full response

Bottom line for the opposition: no single party dominates. Arizona Public Service has the largest cluster (11 marks), but that is one owner in a field of ~37 — the opposite of exclusivity. The register shows APS coexisting peacefully across essentially unrelated goods and services, which supports arguing the term is weak/diluted and entitled only to a narrow scope of protection.

Want me to run a class-scoped crowded-field count for the specific class your applicant filed in, or pull the opposer's own TTAB enforcement history to see whether they've actually policed APS before?

Markus · live run Jul 16, 2026 list_marks_containing_term, get_mark_owner_landscape 45 secondsRead the unedited session

Why this needs the live register

Generic AI will draft the shape of a crowded-field defense — with invented registrations. Markus counted the field (exactly 50 live marks, 37 owners — verified to the digit) and returned real owners with real classes, ready to become an exhibit. This isn’t an argument; it’s evidence.

Field counts and every owner row independently re-verified against the live USPTO register, July 2026. Opposition scenario fictionalized.

03Clearance · knockout search

Clearance that hears the mark the way consumers will.

A client wants LUVLOKK for a new dating app. The literal string is clear — no application has ever been filed for it. A lawyer's real question is what it sounds like.

Run a knockout on LUVLOKK for a dating/relationship app — Classes 9 and 45. Include phonetic equivalents, not just the literal spelling.

Attorney

CRITICAL CONFLICT

Verdict: the engine flags this as CRITICAL — do not treat LUVLOKK as clear for a dating app. The phonetic net pulled 513 sound-alike hits (METAPHONE "LFLK"), and several land squarely in your market — 17 high and 20 medium conflicts overall. The ones that actually matter:

MarkSerialStatusClassesGoods/Services
LOVELOCK99359483Registered45Marriage partner introduction / dating services
LUVLO99514615Registered9, 41Downloadable audio-visual & music recordings
LUVLINK79318043Registered9Downloadable software to share content
LUVLY98007877Registered9Downloadable mobile app software

LOVELOCK is the standout obstacle. It's a live registration in Class 45 for marriage partner introduction and dating services — essentially identical services — and "LOVELOCK" vs. "LUVLOKK" is a phonetic near-twin. This is the one most likely to block you outright under §2(d).

Show full response

The broader "LUV-" field is crowded (LUVLY, LUVLOO, LUVLOOP, LUVLOOK…), which cuts both ways: "LUV" is a weak, diluted prefix — but the second element ("LOKK"/"LOCK") is what ties you dangerously close to LOVELOCK specifically.

Bottom line: LUVLOKK carries critical §2(d) risk for a dating app. I wouldn't recommend proceeding without a plan to distinguish or clear LOVELOCK. Want me to run a full 13-factor DuPont analysis of LUVLOKK vs. LOVELOCK, pull crowded-field evidence on the "LUV" prefix, or check domain availability?

Markus · live run Jul 16, 2026 run_knockout_search (phonetic + sound-alike expansion) 40 secondsRead the unedited session

Why this needs a phonetic index

A general model can’t run a phonetic search of the register — it doesn’t have one. It would call LUVLOKK "likely available" because the literal string is unregistered. Markus expanded to 513 sound-alikes and surfaced a registered, phonetically identical mark covering the identical services — the refusal you’d otherwise discover ten months and one filing fee later.

All statuses and registration numbers independently re-verified (LOVELOCK Reg. 8202856; LUVLO Reg. 8303965; LUVLINK Reg. 6854616), July 2026. Client mark and scenario fictionalized.

04Office actions · evidence gathering

The office action response is an argument. Markus brings the exhibits.

An examiner demanded a disclaimer of ".AI" in a client’s Class 42 software application. The attorney wants third-party evidence that the Office routinely registers "[TERM].AI" marks without one.

Find me third-party "[TERM].AI" marks on the Principal Register in Class 42 where .AI was not disclaimed — I want to argue examiner consistency.

Attorney

11 of 19 registered without disclaimer

Good news for an examiner-consistency argument: among live, registered Class 42 marks containing "AI," a majority registered without a disclaimer — of 19 sampled, 11 registered without disclaiming it and only 8 disclaimed it. The pattern actually favors non-disclaimer.

".AI"-style marks that registered on the Principal Register without disclaiming "AI":

MarkOwnerReg. No.Classes
CARE.AIStryker Corporation71608119, 10, 37, 42
IDOX.AIFoxit Software Incorporated752387642
VUE.AIM2P US Corporation539249542
DOGAN AIDogan, Ufuk612088942
RETINSIGHT AI TO EYERetInSight GmbH71234689, 35, 38, 42

For contrast, marks in the same class that did disclaim "AI": COOP AI (Reg. 7642218), TERAFLOP AI (Reg. 8171542), and AI MASCOT (Reg. 7682709).

Three of the non-disclaimed examples — CARE.AI, IDOX.AI, and VUE.AI — are the closest structural analogs to a "[TERM].AI" formative, since they follow the same word-plus-".AI" construction rather than "AI" as a standalone second word. Those three are your strongest citations for an examiner-consistency argument.

Show full response

Want me to pull the full disclaimer statement language or prosecution history on any of these to strengthen the response?

Markus · live run Jul 16, 2026 get_disclaimer_evidence 19 secondsRead the unedited session

Why this needs the case files

Ask a general model for supporting registrations and you get plausible fabrications — mark names and reg numbers that don’t exist, which is malpractice-grade risk in a filed response. Markus reads the actual USPTO disclaimer statements on each file — it found both the marks that disclaimed and the ones that didn’t, and named the three structural analogs worth citing.

Every registration re-verified against USPTO case-file statement records: no "AI" disclaimer on CARE.AI, IDOX.AI, or VUE.AI; COOP AI’s "AI" disclaimer present. July 2026. Client scenario fictionalized.

05Filing · classification & identifications

One messy client description in. A five-class filing strategy out — with Term IDs.

The everyday work, done right: an attorney pastes a client’s plain-English business description and gets back classification, pre-approved ID Manual identifications, and the fee strategy — in 53 seconds.

My client provides design, construction, and installation of timber framed structures, lighting installation services, and the sale of timber frame and outdoor lighting products. What classes and identifications do we need?

Attorney

5 classes · pre-approved Term IDs

Your client's six activity lines map to five Nice classes. Here's the breakdown, using pre-approved USPTO ID Manual wording — each entry's Term ID lets you select it in the filing system and avoid the $200/class free-form surcharge:

ClassCoversSuggested identificationTerm ID
42Design of timber structuresBuilding design services042-3142
37Construction of timber structuresBuilding construction services037-633
37Lighting installationLighting apparatus installation037-1226
19Timber frame products (goods)Structural timber; building timber019-511 / 019-720
11Outdoor lighting products (goods)Lighting fixtures011-785
35Selling both product linesRetail store services featuring… (fill-in template)035-3004
Show full response

Notes worth flagging: Class 37 carries two of your lines — building construction and lighting installation combine into one class fee. Class 19 is for timber materials; if any framing is actually metal, those shift to Class 6 (e.g., "Steel frames for building," 006-549). And the retail template needs its placeholders completed, plus confirmation whether they sell their own goods (retail may be redundant — the goods registrations cover that) or third-party goods too.

One reminder: these must be selected in the USPTO ID Manual picker at filing — pasting the same wording into a free-form box still triggers the surcharge.

Cost consideration: that's potentially 5 classes at ~$350–$550 each. If budget is tight, the goods classes (11, 19) plus the service classes (37, 42) may make Class 35 redundant. Want me to run a clearance/knockout search on the client's mark across these classes?

Markus · live run Jul 16, 2026 suggest_gs_descriptions ×6 (USPTO ID Manual) 53 secondsRead the unedited session

Why this needs the ID Manual

A general model gets you 80% of this — and the missing 20% is the expensive part: the actual ID Manual Term IDs (every one shown here verified active in the current Manual), the select-don’t-paste surcharge trap, and the manufacture-vs-resell fork that decides whether the client files three classes or five.

All six Term IDs independently verified active in the USPTO ID Manual with matching class and wording, July 2026. Client description contains no identifying details.

About these examples

Every Markus response shown is genuine, unscripted product output from a live run on July 16, 2026 — condensed for space, never rewritten. Attorney-side scenarios are fictional composites drawn from real working sessions; no customer or client matter is depicted. Every data point was independently re-verified against the USPTO record. All third-party marks belong to their respective owners and appear solely as factual references to public USPTO records. Markus produces first drafts and research for review by a qualified practitioner — it is a research tool, not legal advice.

The best demo is your own docket.

Every session above started with one plain-English question. Ask Markus about your marks, your clients, your deadlines, your opposer — it's included with Professional and Max plans.

Meet Markus