There is no objective ranking of lead-gen counsel, and anyone selling you one is selling you a directory listing. But there is objective data on who actually does this work — every federal TCPA case is a public docket — and almost nobody uses it. This aggregates it.
This is not a ranking, and it cannot be turned into one. There is no sort-by-volume control here by design. A high docket count means a firm is busy, not that it is good — and the best compliance lawyer you could hire may barely appear, because the work that succeeds is the advice that keeps you out of court entirely.
Use it to build a shortlist and to write better questions for a call. Do not use it to pick a firm.
152 firms that appear in at least 10 federal TCPA dockets filed since April 2021, listed alphabetically. Nothing you type is stored or sent anywhere.
5,938
TCPA dockets in the window
filed since Apr 1, 2021
88
Federal courts involved
143 days
Median case duration
all closed cases in the window
All 152 firms, A–Z.
The three limits that matter
1. Every count is a floor, never a total
These dockets come from RECAP, an archive of federal court records that people have paid to pull from PACER. It holds what somebody bought, not everything that exists. So a firm showing 40 dockets has at least 40 — it may have many more that nobody happened to purchase. A thin result tells you the archive is thin. It never tells you a firm is inexperienced.
2. You cannot see which side they were on
A docket record lists the firms and the parties as two separate lists, with nothing connecting them. The link between an attorney and the party they represented lives behind an authenticated endpoint. Until that is wired in, every count here mixes defense work and plaintiff work — and for a lead-gen operator that distinction is the whole point. Ask on the call.
3. Win rate is deliberately absent
We could compute one. We will not publish one. Cases are not randomly assigned to firms, so an outcome rate measures which cases a firm accepts at least as much as how well it handles them. A firm that takes only clean cases and a firm that takes the hard ones can post identical rates while being nothing alike. A number that confounding does not get more useful with a caveat under it — it gets screenshotted without one.
Source and method
Source: CourtListener RECAP search API (Free Law Project), queried cause:(47:227) — the federal cause code for 47 U.S.C. § 227. Filtering on the cause returns cases actually brought under the TCPA; a keyword search would return anything that merely mentions it.
Window: dockets filed on or after April 1, 2021, the date Facebook, Inc. v. Duguid, 592 U.S. 395 (2021) narrowed the autodialer definition and reshaped TCPA practice. A doctrinal cutoff is easier to defend than an arbitrary “last five years.” That window holds 5,938 dockets.
Snapshot taken 2026-08-06. Firms are listed if they appear in at least 10 dockets in the window; 152 qualify.
Name grouping. Firm names in PACER are free text typed into an address block, not a controlled vocabulary — one large firm appears under eleven different strings, several carrying a street address, and the numeric firm ID tracks the office rather than the firm. Names are reduced to a leading stem and grouped, so that “Shamis and Gentile” and “Shamis & Gentile, P.A.” count once rather than twice.
The check that guards against merging two firms into one. The real hazard in grouping by name is that different firms share a leading surname — Kaufman P.A., Kaufman Dolowich and Kaufman Borgeest are three businesses. So for any single-word stem we look at the word that follows it across every spelling on file: one firm yields one dominant continuation, several firms yield several. A stem that looks like a merge is dropped rather than published, because a combined count attached to a real named business would be the worst error this page could make.
Firms that renamed appear under both names. PACER records the name used at the time of filing, so a firm that rebranded mid-window is split across two entries. We do not merge them, because the safe-looking merge is sometimes wrong: Steptoe LLP and Steptoe & Johnson PLLC look like a rename and are in fact different firms.
Median case duration is measured from filing to termination across dockets that have closed. Open dockets are excluded from that figure and reported separately, so a firm with many live cases is not made to look fast.
Not every entry is a law firm. PACER’s firm field occasionally carries the name of an individual litigating without counsel — TCPA attracts a handful of prolific pro-se plaintiffs, and they surface here looking like small firms. We do not filter them out, because guessing which names are individuals would introduce a worse error than disclosing the limitation. Click through to the dockets if a name is unfamiliar.
Counts come from the harvest, not from a follow-up phrase query: all 65 monthly slices paginated to exhaustion with zero failed pages, so the window is complete. An “all time” figure is deliberately not shown — it would have to come from a phrase query, and a phrase query cannot faithfully represent a firm name containing “&”. Better to publish one number that is right than two where one is quietly wrong.
How to actually use this
The companion article walks the full method, including the free regulatory layer this tool cannot show — FCC comment filings in CG Docket 02-278, which reveal who is in the room when the rules get written. How to vet lead-gen counsel with public court data covers what to ask once you have a shortlist.
This is an educational tool, not legal advice. It offers general, best-practice guidance from an operator's perspective and does not account for your specific facts or jurisdiction. Using it does not create an attorney–client relationship. For advice on your situation, consult qualified counsel.