Wrong-Party Contact: The Data Problem Behind the Most Complained-About Issue in Collections

TL;DR

• Attempts to collect a debt not owed has been the most common debt collection complaint issue every year since the CFPB began accepting them in 2013, and its monthly average rose 115% in 2025 against the prior two-year average.
• Inside that issue, the CFPB's FDCPA report for 2024 found 60% of consumers said the debt was not theirs, 28% attributed it to identity theft, 10% said it was already paid and 3% said it was discharged in bankruptcy. The first two are about identity; only the last two are about the account itself — payment posting for the 10% and account state for the 3%.
• The fastest-growing debt category in 2025 was "I do not know" — debts consumers did not recognize — up 240% on the prior two-year monthly average.
• Regulation F limits the fix. §1006.10 lets a collector seek location information from a third party but not state that the consumer owes a debt, and generally allows contacting that person only once.
• Wrong-party contact is not a conduct failure you train away. It is a data failure you instrument: identity confirmation before substance, phone provenance on the record, and a disposition code that survives the call.

Why this matters

As of September 2026.

This is for collections managers, servicing directors and dealer principals who look at a right-party contact rate every week and at a complaint every few months, and have never quite connected the two. It is also for anyone about to buy a skip-tracing product on the theory that more numbers is the same thing as better numbers.

The post covers what the federal complaint record actually says about wrong-party contact, why the largest share of the problem is identity rather than payment posting, what Regulation F permits when you go looking for a borrower, and which controls close the gap. It is orientation on published rules and data, not legal advice; how a specific engagement is classified under the FDCPA is determined with your counsel.

The complaint record is remarkably consistent

Start with volume. The CFPB's 2025 Consumer Response Annual Report, published in March 2026, reports approximately 387,400 debt collection complaints in 2025, of which about 304,700 (79%) were sent to companies for review and response. Companies responded to 97%, closing 66% with an explanation and 23% with non-monetary relief.

Now the composition, which is the interesting part. The Bureau writes that the most common issue was "Attempts to collect debt not owed" and adds that this "has been the predominant issue selected by consumers since the CFPB began accepting debt collection complaints in 2013." In 2025 the monthly average for that issue "increased 115% compared to the monthly average for the prior two years." The debt type growing fastest was the one labeled "I do not know," up 240% on the same basis, with "Other debt" up 111%.

The Bureau's own explanation is worth carrying: "A principal reason for the observed increase in debt collection complaints is credit reporting. Compared to prior years, in 2025, consumers increasingly complained about collections they did not recognize appearing (or reappearing) on their reports."

The finer breakdown comes from the prior year. The CFPB's Fair Debt Collection Practices Act Annual Report 2025, covering calendar 2024, found that within the attempts-to-collect-debt-not-owed issue, 60% of consumers said the debt was not theirs, 28% attributed it to identity theft, 10% said it was already paid and 3% said it had been discharged in bankruptcy.

What the consumer said	Share of the issue, 2024	What it is really about
Not my debt	60%	Identity — the account was matched to the wrong person
Identity theft	28%	Identity — the account was opened in the wrong person's name
Already paid	10%	Payment posting and account state
Discharged in bankruptcy	3%	Account state and suppression

Source: Consumer Financial Protection Bureau, FDCPA Annual Report 2025 (calendar 2024 data).

Read the right-hand column. Roughly seven-eighths of the most-complained-about issue in the industry is an identity and matching problem, not a ledger problem. A servicing floor that responds to it by tightening payment research is fixing the smaller half of the smaller half.

Why an auto portfolio generates wrong-party contact

Auto servicing has structural exposure that a card portfolio does not.

The contact information is captured once, at the deal desk, at the least reliable moment in the relationship: a long signing, a borrower focused on the payment, a number transcribed by hand. It then ages across a loan term measured in years while mobile numbers are disconnected and reassigned. Co-signers, references and previous owners sit in the same record as the borrower, and a poorly designed screen makes all of them look like the same kind of number.

Scale sharpens it. The Federal Reserve Bank of New York's Quarterly Report on Household Debt and Credit records 4.9% of consumers carrying a third-party collection account on their credit report in the second quarter of 2026, and the report's data workbook puts outstanding auto loan balances for the same quarter at $1.713 trillion (sheet "Page 3 Data"), with 5.49% of those balances 90 or more days delinquent (sheet "Page 12 Data"). Small percentages of a portfolio that size are large populations of people being called, and every wrong number dialed is both a compliance exposure and a wasted attempt. Official comment 14(b)(2)(i)–3 to §1006.14(b)(2) says calls to a number the collector learns is not the consumer's are not treated as calls placed to the consumer for the seven-calls-in-seven-days presumption, but a dialer can only leave them out of the count, and an audit can only see why, if the wrong-number finding is recorded on the number.

What Regulation F lets you do when you go looking

This is where operators get surprised, because the tool everyone reaches for — calling someone who might know where the borrower is — is specifically regulated.

Section 1006.10 governs the acquisition of location information, which it defines as a consumer's "place of abode and telephone number at such place" or "place of employment." When communicating with a third party for that purpose, a debt collector identifies himself or herself individually by name, states that he or she is "confirming or correcting the consumer's location information," and identifies his or her employer only if expressly requested. The collector must not state that the consumer owes any debt, must not communicate by postcard, and must not use any language or symbol on an envelope or in the contents of a mailed communication indicating that the debt collector is in the debt collection business or that the communication relates to the collection of a debt. Once the collector knows the consumer is represented by an attorney on the debt and knows or can readily ascertain that attorney's name and address, the rule directs communication to that attorney, unless the attorney fails to respond within a reasonable period of time.

Then the frequency limit that surprises people most. Under §1006.10(c), a collector "must not communicate more than once with such person unless requested to do so by such person, or unless the debt collector reasonably believes that the earlier response of such person is erroneous or incomplete and that such person now has correct or complete location information." Once. Not once a week — once, absent one of those two exceptions.

Two related provisions complete the picture. Section 1006.6(d)(1) lists the small set of people a collector may communicate with about the debt — the consumer, the consumer's attorney, a consumer reporting agency where otherwise permitted by law, the creditor, the creditor's attorney and the collector's own attorney — with exceptions that include the location-information purpose and the consumer's prior consent. And §1006.14(b)(3)(iii) keeps calls placed to the parties listed in §1006.6(d)(1)(ii) through (vi) — the consumer's attorney, a consumer reporting agency, the creditor, the creditor's attorney and the collector's own attorney — out of the call-frequency count. The neighbor, relative or employer called for location information is not on that list, so calls to that person count toward the presumption like any other call: §1006.10(c) opens "In addition to complying with § 1006.14(b)(1)," and the Bureau's official interpretation at comment 14(b)(2)(i)–1.iii applies the seven-calls-in-seven-days presumption to unanswered calls placed to a particular third party for location information. On top of that count, §1006.10(c) limits the collector to communicating with that person once, absent the exceptions it sets out.

The operating consequence is simple to state and easy to get wrong: a location call is counted like any other call and also carries a once-only communication limit that a collection call does not, so a system that cannot tell the two apart cannot show that limit was respected.

The controls that actually reduce wrong-party contact

Failure mode	What it looks like on the floor	The control
Matching the wrong person	Similar name, same city, purchased or appended data	Identity confirmed before any account substance is discussed; two-factor identity standard written into the script
Aged or reassigned mobile number	Repeated no-contact, then a stranger answers	Provenance stored on every number — source, date added, last verified — and an age threshold that forces re-verification
Reference or co-signer dialed as the borrower	Attempts logged against the account either way	Contact roles typed on the record; location calls on a separate track with a §1006.10 script
A wrong-party answer that never sticks	The same number dialed again next week	A disposition code that permanently suppresses the number, applied by the system rather than a note in free text
Already-paid or discharged accounts still worked	A borrower calls in angry with a receipt	Payment posting and bankruptcy suppression checked at queue build, not at call time

None of these is expensive. All of them are configuration, script and data-model decisions that hold whether the person on the phone is having a good day or a bad one. Our guide to collections KPIs for auto lenders covers how right-party contact should be measured so that these fixes actually show up in the number, and the first-party versus third-party guide explains why the classification of a program changes which rules apply to it.

An illustrative operating example

The example is illustrative and uses round numbers. It is not a client result.

A lender works a queue of 3,000 delinquent accounts with a right-party contact rate reported at 22%. The floor treats that as a productivity number and buys an appended-phone product to lift it. Contact rate rises to 26%, and the complaint count rises with it, because a share of the new connections are strangers.

The instructive part is what the reporting could not show. "Right-party contact" was being calculated as connections divided by attempts, with the numerator counting any human who answered. The floor had no field distinguishing a confirmed borrower from an unconfirmed answer, no provenance on the appended numbers, and no permanent suppression when someone said "you have the wrong person" — so the same numbers returned to the queue the following cycle.

The remedy was not a different data vendor. It was a confirmed-identity flag as the numerator, provenance on every number, and a system-applied wrong-number suppression. Reported contact rate fell to 19%, which looked like a step backwards and was in fact the first accurate reading the floor had ever had.

What operators say

Rather than paraphrase, here is how the CFPB records the two sides of this dispute in its 2025 Consumer Response Annual Report.

On what consumers asked for: "In many instances, consumers requested that collectors validate the debts and provide the original 'wet ink' contracts or other account documentation to substantiate the debt."

On how companies answered: "Some collectors reported closing the account and updating the reported information, whereas other collectors provided verification information (e.g., the name of the original creditor, account statements) and stated that the consumer remained liable for the debt."

Both responses are available to an operation that kept its documentation. Only one is available to an operation that did not, and it is not the favorable one.

Frequently asked questions

What is the most common debt collection complaint?

Attempts to collect a debt the consumer says is not owed. The CFPB reports it has been the predominant issue selected by consumers since the Bureau began accepting debt collection complaints in 2013, and that its monthly average in 2025 rose 115 percent compared with the monthly average for the prior two years.

Is wrong-party contact a compliance problem or a data problem?

Both, but it originates in data. In the CFPB's 2024 figures, 60 percent of consumers raising the attempts-to-collect-debt-not-owed issue said the debt was not theirs and 28 percent attributed it to identity theft, against 10 percent saying it was already paid. Roughly seven-eighths of the issue is identity and matching rather than payment posting.

How often can a collector contact a third party to find a borrower?

Under 12 CFR §1006.10(c), a debt collector must not communicate more than once with a person for location information unless that person requests further contact, or unless the collector reasonably believes the person's earlier response was erroneous or incomplete and that the person now has correct or complete location information.

What can a collector say when asking someone where a borrower is?

Section 1006.10(b) limits the content: the collector identifies himself or herself individually by name, states that he or she is confirming or correcting the consumer's location information, and identifies his or her employer only if expressly requested. The collector must not state that the consumer owes any debt, must not communicate by postcard, and must not use any language or symbol on an envelope or in the contents of a mailed communication indicating that the debt collector is in the debt collection business or that the communication relates to the collection of a debt.

Do location calls count toward the seven-calls-in-seven-days presumption?

Yes. Regulation F has no exclusion for location calls. Section 1006.14(b)(3) removes only three kinds of call from the frequencies: calls placed with the called person's prior consent given directly to the collector, within seven consecutive days of receiving it; calls not connected to the dialed number; and calls to the parties in §1006.6(d)(1)(ii) through (vi), which are the consumer's attorney, a consumer reporting agency, the creditor, the creditor's attorney and the collector's attorney. A neighbor, relative or employer called for location information is not one of those parties. The CFPB's official interpretation, comment 14(b)(2)(i)–1.iii, works through up to seven unanswered location-information calls to one third party over seven consecutive days and concludes the collector is presumed to comply with §1006.14(b)(1). The frequencies are counted per person and per debt, so these calls count for the person called. Section 1006.10(c) then applies in addition: once the collector has communicated with that person, the rule allows no further communication with them unless that person requests it, or the collector reasonably believes the earlier response was erroneous or incomplete and that the person now has correct or complete location information. That is why location attempts are worth logging per person, with a record of whether a conversation took place.

How should a wrong number be handled so it does not come back?

With a disposition code applied by the system that permanently suppresses the number on the account, not a note in a free-text field. A free-text note does not stop the next queue build from dialing the same number, which is how the same stranger gets called repeatedly and how a single data error becomes a pattern.

The bottom line

The most-complained-about issue in debt collection has not changed in more than a decade, and the federal record says plainly what it is made of: not people refusing to pay, but people being contacted about accounts that are not theirs. That is a matching, provenance and suppression problem, and Regulation F deliberately narrows the tools available to solve it by phone.

Operations that confirm identity before substance, store where every number came from and when it was last verified, keep location outreach on its own track with its own script and its own once-only limit, and let the system rather than the agent suppress a bad number will see a lower reported contact rate and a truer one. If you want a dedicated team working your accounts inside your systems with those controls in place from the first day, tell us about your portfolio.