Who Files the Most Trademarks — and Who Abandons Them

Ask most people which firms file the most U.S. trademark applications and they'll name the big intellectual-property practices. The filing record says something more interesting: the highest-volume filers are three very different kinds of operation, and the odds that any given application is dead within a year swing from 1 in 200 to 1 in 4 depending on which kind you used.

What we measured, up front. Filing volume is the trailing twelve months — June 2025 through May 2026. Outcomes are measured on earlier cohorts old enough to have resolved: first-year abandonment on 2024 filings, and eventual registration on the 2022–2023 filings (most have finished prosecuting). A mark filed last month hasn't had time to register or die, so counting it as either would mislead — which is exactly the methodological trap this kind of ranking usually falls into.

We mapped every U.S. application to the entity that filed it, then followed each application forward on a fixed clock. Here's what the data actually shows.

Who files the most

Applications filed over the trailing twelve months (Jun 2025–May 2026):

Rank	Filer	Archetype	Apps (12 mo)
1	Swyft Legal	DIY / startup mill	10,966
2	LZ Legal Services	DIY / startup mill	10,349
3	Flatfee Corp.	DIY / startup mill	6,562
4	Alioth Law	High-volume shop	5,515
5	Grogan, Tuccillo & Vanderleedeen	High-volume shop	3,454
6	Sparring Legal	DIY / startup mill	3,215
7	LegalForce RAPC	DIY / startup mill	3,037
11	Brown Brothers Law	Seller-volume shop	2,034
17	Greenberg Traurig	Traditional brand firm	1,514

The first firm a brand manager would recognize — Greenberg Traurig — doesn't appear until #17. Above it sit operations built to file in bulk. But "bulk" isn't one thing. The client lists pull them apart cleanly.

Three kinds of high-volume filer

1. DIY / startup mills — Swyft Legal, LZ Legal Services, LegalForce RAPC. Their clients are startups, solo founders, and small LLCs — MindHYVE.ai, WigiWork, Paper Street Media, Essex Token. Marks come a handful at a time from filers testing an idea. These are flat-fee, high-throughput operations, and roughly 40–52% of their filings are use-based (the rest intent-to-use or other).

2. Seller-volume shops — Brown Brothers Law, Murray Ziel & Johnston, MDE Patents. Their books are dominated by Chinese e-commerce sellers: Dongguan Hengtai Biotechnology, Shenzhen Shangmu Fashion, Dongguan Huadi Trading, Hangzhou Tianyuan Pet Products. These are largely Amazon Brand Registry filings — 69–71% use-based, backed by a product already selling, which is why they sail through examination. It's high volume, but a completely different machine from the startup mills.

3. Traditional brand firms — Greenberg Traurig. The marquee end: Home Depot, Polo Ralph Lauren, Wynn Resorts, Margaritaville — hundreds of marks per client, deep institutional relationships. Only 32% of its filings are use-based; the rest are intent-to-use, the sophisticated practice of protecting a brand before it ships. That choice matters for how the numbers read, as we'll see.

The abandonment gap (first-year failure)

To compare early outcomes fairly we use the 2024 cohort and ask: what share of each filer's applications were abandoned within 365 days of filing? Every application gets the same one-year clock, so a younger book can't flatter the result.

Filer	Archetype	Abandoned ≤ 1 yr (2024)
Swyft Legal	DIY / startup mill	25.0%
LegalForce RAPC	DIY / startup mill	21.5%
LZ Legal Services	DIY / startup mill	16.0%
Brown Brothers Law	Seller-volume shop	6.7%
Murray, Ziel & Johnston	Seller-volume shop	4.8%
Greenberg Traurig	Traditional brand firm	4.1%
MDE Patents	Seller-volume shop	0.5%

The split is clean and it follows the archetype. The DIY mills abandon at 16–25% — one in four, at the top. The seller-volume shops and the brand firm sit at 0.5–6.7%: up to a fiftyfold difference in the share of applications that die in their first year. A startup that paid a flat fee to file on a half-formed idea often won't answer an Office Action; a seller with live Amazon inventory has a specimen ready; a brand firm has a client paying to win. Abandonment is the fingerprint of which situation you're in.

What actually registers (the multi-year view)

First-year abandonment captures early failure. But does the mark ever register? For that we need cohorts old enough to have finished, so we turn to 2023 filings — two to three years on, mostly resolved. The pattern holds and it sharpens:

Filer (2023 cohort)	Registered	Abandoned	Still pending
Flatfee Corp.	81%	17%	3%
Sparring Legal	53%	42%	5%
LZ Legal Services	49%	45%	6%
LegalForce RAPC	48%	45%	7%
Swyft Legal*	40%	47%	13%
Greenberg Traurig	45%	23%	32%

*Swyft's 2023 cohort is small (~465 marks) — read as directional. Several of today's fastest-growing filers (Brown Brothers, Murray, Alioth) barely existed in 2023, so their eventual-registration rates can't yet be measured on a stable sample.

Two things stand out. First, the big startup mills — LZ, LegalForce, Swyft — ultimately abandon roughly 45% of everything they file. Nearly half the applications never become registrations. And the segment isn't monolithic: Flatfee Corp. converts at 81%, better than most law firms, proof that "flat-fee filing service" doesn't have to mean high churn.

Second, Greenberg Traurig's 45% registered is the lowest-looking figure on the list — and the most misleading. Nearly a third of its 2023 filings are still pending, because they're intent-to-use applications for brands working their way to market; they're maturing, not failing. (Its 23% multi-year abandonment is mostly ITU marks abandoned when a planned brand never launches — a business outcome, not an examination loss, which is why its first-year abandonment was just 4.1%.) Registration counts reward simple in-use filings and penalize strategic ones — so abandonment within a fixed window, not a raw registration rate, is the cleaner read on prosecution quality.

What it means if you're filing a trademark

The volume leaderboard and the survival leaderboard are nearly inverse lists, and the gap between them is the whole point. A flat-fee mill that abandons nearly half its filings is not cheaper if your mark is one of the half — you've lost the fee, the priority date, and the time. Match the filer to the job: a use-based mark for a product already selling is low-risk almost anywhere; a strategic intent-to-use filing, or a mark likely to draw a refusal, is exactly where the gap will find you. And as Flatfee shows, the label on the door predicts less than the track record behind it — which is the one thing worth checking before you file.

Methodology

• Source: every U.S. trademark application in the USPTO record (~13.9M), mapped to its filing entity via correspondent-of-record.
• Volume window: filing date in the trailing twelve months (Jun 2025–May 2026).
• Abandonment metric: the 2024 filing cohort, measured at a fixed 365 days after each mark's own filing date — a controlled, time-to-event rate that removes the bias of younger applications not yet having had time to resolve.
• Registration / eventual-outcome metric: the 2022–2023 filing cohorts (two to three years matured), reported as current status. Filers without a stable mature sample are footnoted, not rated.
• Use-based vs. intent-to-use and client mix come from the filing-basis field and owner-of-record; archetypes are assigned from observed client base and filing basis, not self-description.
• Excluded: normalization artifacts and entities with too few mature filings to compute a stable rate.
• Figures refreshed monthly.