TCPA Class Action Discovery: How Defendants Keep It Narrow

Class discovery is where a TCPA case gets expensive, and it’s also where a defendant has the most room to steer. Courts can test the plaintiff’s own claim first, trim requests that aren’t proportional to the case, and put call logs under a protective order. None of that happens by default. In my view, a company that names a narrow, dispositive issue early and has its records preserved has more control over discovery than the plaintiff’s first requests suggest.

Can we make the plaintiff’s own claim go first?

Sometimes. Courts call it bifurcated or phased discovery, and the orders discussed below suggest the answer depends on whether the defense can point to a specific issue that could end the named plaintiff’s claim.

The rules allow it. Rule 26(f) requires the parties’ discovery plan to address “whether discovery should be conducted in phases or be limited to or focused on particular issues.” The 2003 Rule 23 committee note recognizes that a defendant “may prefer to win dismissal or summary judgment as to the individual plaintiffs without certification and without binding the class that might have been certified.” The same note also warns against “forcing an artificial and ultimately wasteful division” between certification and merits discovery.

District courts weigh similar considerations and reach different results. In Cameron v. CHW Group, a federal magistrate judge in Utah granted bifurcation in a putative Do Not Call class action in August 2025, while acknowledging that bifurcation is not the norm. The court pointed to issues that discovery about the plaintiff alone could decide before costly class discovery, such as consent and whether his number was on the National Do Not Call Registry (the federal list of numbers whose owners have opted out of telemarketing). The Cameron court wrote that TCPA cases, in its view, “present a unique opportunity that warrants bifurcated discovery due to those cost savings.”

In May 2024, a federal judge in Arizona denied bifurcation in Workman v. CarGuard Administration, a prerecorded-call class action (calls that played a recorded or machine-generated message). The court found that the defendant “has not pointed to a specific defense or theory that it anticipates will be supported by undisputed facts or where Defendant would be entitled to judgment as matter of law.” It also found that several elements overlapped between class and merits discovery, and it refused to hold back class members’ names and call records until after certification.

The difference between those two orders is the lesson. A motion that only predicts a summary judgment win after a first phase is likely to lose, because the Workman court agreed with the plaintiff that this hope applies to every class action. A motion that names a fact about the plaintiff, such as a dated sign-up record or a number the company never dialed, and shows that a short first phase can resolve it, has a real chance. I’d attach the proposed first-phase requests so the judge can see how small the phase is.

Losing the motion isn’t the end of it. The Workman court noted that “There is no requirement that Defendant wait until the dispositive motion deadline to file a dispositive motion,” and it recorded the plaintiff’s concession that, once discovery into the plaintiff’s claim supported an early summary judgment motion, the defendant might have a strong basis to seek a stay of further discovery.

How does proportionality limit class discovery in a TCPA case?

It caps discovery at what the case needs. In a TCPA case, I’d use it to narrow the time period and campaigns, or to offer a sample.

Rule 26(b)(1) allows discovery that is “relevant to any party’s claim or defense and proportional to the needs of the case,” weighing factors that include “whether the burden or expense of the proposed discovery outweighs its likely benefit.” The court must also limit discovery it finds “can be obtained from some other source that is more convenient, less burdensome, or less expensive,” under the same rule.

The strongest proportionality arguments tie the request to the class definition. If the complaint challenges one texting campaign, every other campaign is a stretch. Older logs are a fair target for an objection, though a plaintiff may argue older calls bear on willfulness, as our post on TCPA limitations and class defense explains.

Proportionality has limits, though. For autodialer and prerecorded-voice claims, 47 U.S.C. § 227(b)(3) allows actual monetary loss or $500 per violation, whichever is greater, and a court may, in its discretion, increase the award to as much as three times that amount if the violation was willful or knowing. Do Not Call claims under § 227(c)(5) have their own damages provision. Across a large class, Rule 26’s amount-in-controversy factor can favor the plaintiff, so I’d offer something concrete, such as a sample of call records or a stipulation to the number of messages sent, in exchange for not producing every row.

Can we keep call logs and consent records confidential?

Yes, under a protective order, but don’t count on it to keep class members’ phone numbers out of the case.

Rule 26(c) lets a court act “for good cause,” including by ordering that “a trade secret or other confidential research, development, or commercial information not be revealed or be revealed only in a specified way.” A company’s lead sources and vendor pricing can fit that description. The same rule lets the court set terms for “allocation of expenses,” which is the hook for asking the plaintiff to share the cost of an expensive data pull.

The showing has to be specific. In Beckman Industries v. International Insurance Co., the Ninth Circuit, quoting the Third Circuit, wrote that “Broad allegations of harm, unsubstantiated by specific examples or articulated reasoning, do not satisfy the Rule 26(c) test.”

In Workman, the court found that withholding putative class members’ names and numbers before certification would deprive the parties of information the court needed to rule on certification, that disclosure didn’t seriously invade their privacy, and that “Any privacy concerns may be resolved through a protective order limiting the use of such information,” in the court’s words. I’d spend the effort on the order’s terms instead, such as a bar on contacting class members before certification and a return-or-destroy clause.

What can we learn from the plaintiff’s phone and the carrier?

Plenty, but message content usually has to come from the plaintiff.

Our post on defending serial TCPA plaintiffs covers what to ask a repeat filer about the number and prior suits. For the deposition itself, see our post on deposing the TCPA plaintiff. Subject to listed exceptions, such as the consent of the sender or a recipient, the Stored Communications Act says a provider of electronic communication service to the public “shall not knowingly divulge to any person or entity the contents of a communication while in electronic storage by that service.” It has no exception for civil subpoenas. The same section lets a provider disclose subscriber records, not including contents, “to any person other than a governmental entity.”

So a subpoena to the plaintiff’s wireless carrier may yield who held the number and the call and text detail records, but don’t expect it to produce what the texts said. Ask the plaintiff early to preserve and produce the messages from the phone itself. Our post on spoliation in TCPA cases covers what courts can do when either side loses its records.

Do our vendors’ records count as ours?

Sometimes, depending on the contract, and that changes who answers the discovery.

Rule 34 reaches documents “in the responding party’s possession, custody, or control.” In the Ninth Circuit, “Control is defined as the legal right to obtain documents upon demand,” as the court put it, quoting its own earlier decision, in the 1999 In re Citric Acid Litigation appeal, which applied that test to a subpoena and rejected a looser practical-ability test. If the company’s contract with a texting platform or call center gives it the right to demand the logs, a plaintiff can argue they’re the company’s to produce. Read the contract before you tell opposing counsel the records belong to someone else.

Where the vendor controls its own data, the plaintiff can serve a subpoena, and Rule 45 requires the party serving one to “take reasonable steps to avoid imposing undue burden or expense on a person subject to the subpoena.” If the vendor’s files hold the company’s confidential information, I’d ask the plaintiff to agree that the case’s protective order covers the vendor’s production. I’d also coordinate with the vendor’s counsel before it produces anything, since a lead generator’s files can help the defense on consent and help the plaintiff build a class.

What should a TCPA litigation hold cover?

Every system that touched the calls, and every vendor that keeps its own copy.

Rule 37(e) applies when electronically stored information that should have been preserved in the anticipation or conduct of litigation is lost because a party “failed to take reasonable steps to preserve it.” If the information can’t be restored or replaced through other discovery, a court that finds prejudice may order measures no greater than necessary to cure it. The harsher remedies, including an adverse-inference instruction or default judgment, require a finding under Rule 37(e)(2) that the party “acted with the intent to deprive another party” of the information.

I think TCPA data is unusually perishable. If a platform keeps message logs only for a set retention period, a company that doesn’t send its vendor a written preservation request can lose the consent and opt-out records its defense needs. Rule 26(f) also requires the parties’ discovery plan to state their views on “any issues about disclosure, discovery, or preservation of electronically stored information,” so a defendant that knows its systems can propose a specific preservation scope instead of agreeing to a preserve-everything letter.

How do early motions change the discovery fight?

They set how much discovery happens, and when.

Under Rule 26(d)(1), “A party may not seek discovery from any source before the parties have conferred as required by Rule 26(f),” with limited exceptions. No federal rule stays discovery automatically while a motion to dismiss is pending, so a defendant that wants a pause has to ask for one. I’d explain why the motion is likely to narrow the case.

In my view, a motion to compel arbitration is the strongest tool for stopping discovery. If the court denies it and the company appeals, the Supreme Court held in Coinbase, Inc. v. Bielski that a district court “must stay its proceedings while the interlocutory appeal on arbitrability is ongoing.” That’s one reason I’d check every consumer-facing agreement for an arbitration clause before the first scheduling conference.

The defense can also put certification in play early. In Vinole v. Countrywide Home Loans, an overtime class action, the Ninth Circuit held that nothing in Rule 23(c)(1)(A) prohibits a defendant from “seeking early resolution of the class certification question.” The plaintiffs there had nearly ten months to conduct discovery, and the court noted that the pleadings alone often won’t resolve certification. In Mantolete v. Bolger, an employment case, the Ninth Circuit, following its earlier Doninger decision, stated that a plaintiff seeking class discovery bears the burden of making a prima facie showing that Rule 23 is satisfied or “that discovery is likely to produce substantiation of the class allegations,” and that without it a court may refuse class discovery. A motion to deny certification works best when the company’s records already show that consent or opt-out questions vary person by person, and our post on TCPA standing explains how those records can separate injured class members from uninjured ones. Our post on consent defenses at class certification covers what the company needs in hand for that fight.

Discovery fights in a TCPA class action, and the tools that answer them

Plaintiff’s request	Defense tool	Authority	Where it stands
Class-wide logs before the plaintiff’s claim is tested	Phased discovery	Rule 26(f)(3)(B) and the Rule 23 committee note	Granted in Cameron on plaintiff-specific issues, denied in Workman without them
Every campaign, every vendor, every year	Proportionality	Rule 26(b)(1) and (b)(2)(C)	Strongest when tied to the class definition and class period
Lead sources, pricing, dialer settings	Protective order	Rule 26(c)(1)(G); Beckman	Requires a specific showing of harm, not broad allegations
Class members’ names and numbers	Protective order terms, not refusal	Rule 26(c); Workman	Workman denied bifurcation, finding the numbers needed for certification and privacy manageable by protective order
Vendor and platform data	Contract review and Rule 45 burden limits	Rule 34, Citric Acid, Rule 45(d)(1)	Depends on whether the company can demand the records
Discovery after a denied motion to compel arbitration is appealed	Mandatory stay	Coinbase	Stay required while the appeal is pending

Before the first discovery conference

1. Send a written litigation hold covering the call and text logs, dialer and platform data, and the consent records for every number at issue.
2. Pull the opt-out and internal do-not-call records, with the date each request came in and was honored.
3. Pull the National Do Not Call Registry scrub records and the version dates each campaign used.
4. Gather vendor and lead-source contracts, including their data retention and audit-rights clauses, along with scripts and any recorded or AI-voice message files.
5. Talk to a lawyer before responding to the plaintiff’s preservation letter, agreeing to a discovery plan, or offering anything to settle.

The firm’s page on TCPA litigation defense explains how it approaches TCPA class actions, including discovery.

Questions about TCPA class discovery

Will a court stay discovery while our motion to dismiss is pending?

Not automatically. A federal court can stay discovery but isn’t required to, and it may rule on the motion first, mooting the request. I think a request tied to a narrow issue has a better chance than a general plea about cost. An appeal from a denied motion to compel arbitration is different, because Coinbase makes that stay mandatory.

Do we have to produce class members’ phone numbers before certification?

Courts may require it. In Workman, the court treated class members’ contact details and call records as needed to show that class members received calls the same way the plaintiff did, and it found that disclosure didn’t seriously invade their privacy. The realistic goal is a protective order that limits how the list is used, bars contacting class members before certification, and requires the list’s return or destruction at the end of the case.

Can the plaintiff subpoena our texting platform directly?

Yes, under Rule 45, and if your contract lets you demand the vendor’s records, the plaintiff can also ask you to produce them. Either way, ask that the case’s protective order cover your confidential information in the vendor’s files, and coordinate with the vendor before it produces anything.

What happens if our vendor deleted the logs before we were sued?

It depends on when the duty to preserve arose, whether the company controlled the vendor’s records, and what steps it took. Rule 37(e) allows curative measures, on a finding of prejudice, when information is lost through a failure to take reasonable steps and can’t be replaced, and harsher sanctions only on a finding of intent to deprive. A prompt written hold to every vendor is the best protection, and a purge schedule documented before the claim arrived can help explain a gap.

When a TCPA class action reaches discovery

If your company is defending a TCPA class action, the first discovery plan can decide how expensive the case becomes. Rosenberger + Kawabata represents businesses defending TCPA lawsuits and class actions through its business litigation practice, including discovery fights over call logs and consent records. Contact Rosenberger + Kawabata online for a free and confidential consultation, or call (310) 894-6921.

Sources

The statute is the Telephone Consumer Protection Act at 47 U.S.C. § 227. The Stored Communications Act provision is 18 U.S.C. § 2702. The procedural rules are Fed. R. Civ. P. 23 and its 2003 committee note, Fed. R. Civ. P. 26, Fed. R. Civ. P. 34, Fed. R. Civ. P. 37, and Fed. R. Civ. P. 45.

The appellate decisions are Coinbase, Inc. v. Bielski, 599 U.S. 736 (2023); Vinole v. Countrywide Home Loans, Inc., 571 F.3d 935 (9th Cir. 2009); Mantolete v. Bolger, 767 F.2d 1416 (9th Cir. 1985); Beckman Industries, Inc. v. International Insurance Co., 966 F.2d 470 (9th Cir. 1992); and In re Citric Acid Litigation, 191 F.3d 1090 (9th Cir. 1999).

The district court orders are Cameron v. CHW Group, Inc., No. 2:23-cv-00320, ECF No. 71 (D. Utah Aug. 13, 2025), and Workman v. CarGuard Administration, Inc., No. CV-23-00961, ECF No. 53 (D. Ariz. May 15, 2024).

This post is general information about the Telephone Consumer Protection Act and related laws as of April 28, 2026. It is not legal advice, and reading it does not create an attorney-client relationship. Courts and the FCC change the rules in this area often, and the outcome of any claim depends on its facts. Past results do not guarantee future outcomes.