Skip to content
Ahmed Alazzawi
Business Intelligence

CFPB Consumer Complaints Analytics — 2025

An Excel-based performance, discrepancy, and decision analysis of 482,074 CFPB consumer complaints received in 2025 across checking or savings accounts, credit cards, debt collection, and mortgages.

Complaints analyzed
482,074
Received Jan 1–Dec 31, 2025 · four products
Overall untimely response rate
1.98%
9,523 complaints recorded as untimely
Untimely cases from debt collection
86.1%
vs. 58.7% of complaint volume
Modeled cases avoided
4,627
Strong portfolio scenario · conditional, not a forecast

01Executive summary

Concentration, not merely volume

The principal business finding is concentration. Across 482,074 complaints, 9,523 were recorded as untimely — an overall rate of 1.98%. Debt collection accounted for 58.7% of complaints but 86.1% of untimely complaints, with a 2.89% untimely rate materially above the other analyzed products.

Three additional findings change how performance should be interpreted. Monthly volume and response timeliness did not move together: January produced the most complaints (53,474), whereas April recorded the highest untimely rate (2.61%). A deep investigation of CCS Financial Services, Inc. found its recorded untimely rate fell from 58.24% (January–August) to 2.64% (September–December), and decomposition showed that changing issue composition did not explain the decline. Finally, apparent geographic underperformance partly reflects company mix: excluding CCS reduces Georgia's benchmark-adjusted excess untimely cases from 275.41 to 144.27, a 47.6% decrease.

A decision model translates these observations into conditional outcomes. Under a specified strong-improvement scenario, modeled untimely cases decrease from 9,523 to approximately 4,896 at unchanged complaint volumes. A 10,000-run historical-pattern simulation gives a mean of 9,541 untimely cases and a 13.18% modeled probability of exceeding a hypothetical 10,000-case threshold. These are decision-support illustrations, not forecasts or validated causal effects.

02Dataset & objectives

Dataset overview and analytical objectives

Complaint volume alone is an incomplete performance measure. A high-volume category may have a low rate of recorded untimeliness; a smaller category may have a high rate but contribute little to the total number of late responses. The analysis therefore evaluates volume, untimely-response rate, and absolute untimely cases together.

Complaints
482,074
Received in 2025
Product categories
4
Selected analytical extract
Records matched to a state
480,344
1,730 lacked state information
Workbook worksheets
15
CFPB_2025_Project.xlsx

The underlying public export was scoped upstream to checking or savings accounts, credit cards, debt collection, and mortgages, followed by enrichment and validation. The original record count was preserved throughout. 158 sub-issue fields were missing. 12,762 records had forwarding intervals exceeding 30 days; these were flagged, not declared erroneous. 64,231 records matched an earlier record on non-ID attributes, but all complaint IDs were unique and no records were removed solely for attribute similarity.

Analytical objectives

  1. 01Identify where recorded response risk concentrates across products, months, issues, companies, and states.
  2. 02Investigate unusual patterns and test competing explanations before attributing causes.
  3. 03Translate observed gaps into explicit, conditional improvement scenarios.
  4. 04Quantify how variable annual untimely cases might be under historical month-level patterns.

03Performance & KPIs

Performance overview and KPIs

Debt collection's share of untimely complaints (86.1%) is considerably larger than its share of complaint volume (58.7%). The concentration is not simply a by-product of volume; the category also has the highest rate among the four analyzed products.

Performance overview showing complaint volume, response timeliness, and product-level comparisons for 2025
Fig. 01Performance overview — complaint volume, response timeliness, and product-level comparisons.Full size

Mortgage has a higher rate than checking/savings and credit cards, but a much smaller absolute untimely burden. These distinctions matter when the business objective is to reduce the number of delayed responses rather than only to rank categories by rate.

Product performance, 2025

Financial productComplaintsUntimelyUntimely rate
Debt collection283,2048,1962.89%
Credit card89,9783450.38%
Checking or savings account84,1956090.72%
Mortgage24,6973731.51%
Total482,0749,5231.98%

04Monthly analysis

Monthly performance: peak activity is not peak untimeliness

January was the highest-volume month despite having one of the lowest untimely rates. April's 2.61% rate was the highest, even though April volume was far below January's — a caution against using raw complaint counts as a proxy for response quality.

Monthly performance analysis showing complaint volumes, untimely rates, and month-over-month changes in 2025
Fig. 02Monthly performance — complaint volumes, untimely rates, and changes over time.Full size

Complaint volume fell 43.05% in February. It rebounded later in the year, with December reaching 47,481 complaints, 16.20% higher than November.

Within checking or savings accounts, 9,907 complaints were recorded during January 15–18. Two institutions represented 8,563 (86.43%) of those: Navy Federal Credit Union (4,417) and Capital One Financial Corporation (4,146). This is strong evidence of a concentrated reporting or complaint-volume episode; it is not enough to conclude that the cause was a service outage, operational failure, or erroneous record.

Monthly complaints and untimely rate, 2025

MonthComplaintsUntimely rate
January53,4741.39%
February30,4511.95%
March33,5182.38%
April34,2762.61%
May36,8372.10%
June36,6842.04%
July41,5531.64%
August42,0962.02%
September43,4821.55%
October41,3591.95%
November40,8632.31%
December47,4812.14%

05Company, issue & geography

Company, issue, and geographic analysis

Segment views distinguish the most common problems and the largest populations from the segments with the highest recorded late-response rates — and show where composition, rather than performance, drives an apparent difference.

Company analysis showing variation in complaint volumes and response outcomes by company
Fig. 03Company analysis — variation in complaint volumes and response outcomes by company.Full size

Among the high-volume debt-collection companies shown in the workbook, CCS Financial Services, Inc. had 6,450 debt-collection complaints, of which 2,259 (35.02%) were recorded as untimely. That annual average obscures a dramatic change over the year, investigated in the next section.

Issue-level patterns: the type of problem matters

“Attempts to collect debt not owed” was the largest listed issue by volume, but “Took or threatened to take negative or legal action” had a higher untimely rate (4.49%) and contributed 2,182 untimely cases. “Communication tactics” also had a relatively high rate (3.79%) despite a much smaller volume.

Issue analysis showing complaint topics, their frequency, and differences in response timeliness
Fig. 04Issue analysis — complaint topics, frequency, and timeliness differences.Full size

Selected issues

IssueComplaintsUntimelyUntimely rate
Attempts to collect debt not owed115,6822,9892.58%
Written notification about debt64,0491,1811.84%
Took or threatened to take negative or legal action48,5782,1824.49%
Managing an account44,9593630.81%
False statements or representation38,4611,2613.28%
Communication tactics9,5193613.79%

Geographic patterns: an impact measure, with a qualification

The state-only benchmark covers the 50 U.S. states: 477,871 complaints and 9,451 untimely cases, or 1.98%. Excess cases compare observed untimely cases with those expected at the pooled 50-state rate (observed untimely − state volume × benchmark rate), so the measure incorporates population size.

Geographic analysis showing a state-level untimely rate map, excess untimely cases, and geographic priorities
Fig. 05Geographic analysis — state-level rate map, excess cases, and geographic priorities.Full size

Of Georgia's 946 untimely cases, 296 (31.29%) were associated with CCS Financial Services. When CCS is excluded from both Georgia's observed cases and the benchmark population, Georgia's excess falls from +275.41 to +144.27 — a 47.6% reduction. Geography is therefore partly reflecting which companies are represented in each state; product composition is another potential confounder.

Highest excess untimely cases vs. 50-state benchmark

StateComplaintsUntimelyRateExcess cases
Georgia33,9079462.79%+275.41
North Carolina17,4664332.48%+87.57
Florida51,3601,0882.12%+72.24
Louisiana9,3412312.47%+46.26
Arizona10,8502552.35%+40.42

Three rankings, three different questions

  1. 01Highest volume: Texas — 75,632 complaints.
  2. 02Highest untimely rate: North Dakota — 6.35%, but only 28 untimely cases among 441 complaints.
  3. 03Highest excess untimely cases against the 50-state benchmark: Georgia — +275.41.

06Discrepancy investigation

Discrepancy investigation: explaining the CCS shift

Within debt collection, CCS Financial Services' recorded untimely rate fell from 58.24% in January–August to 2.64% in September–December — a 55.60 percentage-point drop and a 95.47% relative reduction.

Untimely rate, Jan–Aug
58.24%
2,188 of 3,757 complaints
Untimely rate, Sep–Dec
2.64%
71 of 2,693 complaints
Observed change
−55.60 pp
95.47% relative reduction
Two-proportion test
Z = 46.16
p < 0.001, under its assumptions
CCS Financial Services response timeliness investigation showing the monthly untimely rate, issue-mix decomposition, response classification comparison, findings, and recommended actions
Fig. 06CCS discrepancy investigation — response-rate break, evidence, and decomposition.Full size

The mix-only counterfactual is approximately 58.69%, slightly above the early-period actual rate of 58.24%. A changed distribution of complaint issues therefore cannot explain the decline. Recorded untimeliness fell within all seven investigated issue categories, and within both recorded company response classifications.

Rate-change decomposition

ComponentContribution
Change attributable to issue mix+0.45 pp
Change attributable to within-issue rates−56.05 pp
Net observed change−55.60 pp

Screening beyond one company

A curated discrepancy screen extends the investigation beyond CCS. These cases were identified through segmentation and then verified against the data; they are not an exhaustive algorithmic scan of every company.

Screened patternVerified evidence
Servicer under contract with Federal Student Aid651 of 651 complaints marked untimely; 59.29% occurred in Q4
Rent Recovery SolutionsJan–May 0% untimely; Oct–Dec 100%; 575 complaints in the annual screening cohort
January checking/savings concentration8,563 of 9,907 (86.43%) complaints on Jan 15–18 linked to Navy Federal and Capital One

Investigative workflow

  1. 01Detect a disproportionate or abrupt pattern.
  2. 02Verify the numerator and denominator.
  3. 03Test competing explanations.
  4. 04Decide what additional evidence is needed.

07Baseline & What-If

Baseline and What-If scenario modeling

The workbook compares the actual 2025 baseline with hypothetical reductions in untimely response rates at the observed 2025 complaint volumes, converting observed gaps into explicit, conditional scenarios.

Baseline and what-if scenario model comparing observed and modeled annual untimely complaints
Fig. 07Baseline and What-If model — modeled changes in annual untimely complaints.Full size

The strong case corresponds to a 48.6% modeled reduction in untimely cases. Under that scenario, debt collection contributes approximately 3,948 of the 4,627 modeled avoided cases (85.3%), again reinforcing its numerical importance to the portfolio.

Portfolio-wide scenario comparison

ScenarioModeled untimely casesCases avoided
Observed baseline9,523—
Moderate portfolio improvement≈6,602≈2,921
Strong portfolio improvement≈4,896≈4,627

08Sensitivity & Goal Seek

Sensitivity analysis and Goal Seek

A two-variable data table compares scenario combinations, and Goal Seek frames the reverse question: at the same annual debt-collection volume, what untimely rate corresponds to a target of 5,000 untimely cases rather than 8,196?

Debt-collection complaints
283,204
Held constant
Required untimely rate
2.89% → 1.77%
Approximate; −1.13 pp
Target untimely cases
5,000
3,196 fewer than observed 8,196
Sensitivity analysis two-variable data table comparing modeled outcomes across scenario inputs
Fig. 08Sensitivity analysis — two-variable scenario comparison.Full size
Goal Seek analysis showing the debt-collection untimely rate required to meet a 5,000-case target
Fig. 09Goal Seek — required improvement to meet the specified target.Full size

09Monte Carlo simulation

Monte Carlo simulation and scenario distributions

A deterministic scenario reports a single number under a fixed assumption. The Monte Carlo analysis asks how annual untimely cases could vary if historical month-level patterns recurred in different combinations.

Monte Carlo distribution of simulated annual untimely complaints with an illustrative 10,000-case threshold
Fig. 10Monte Carlo outcomes — modeled distribution and illustrative 10,000-case threshold.Full size

The model uses 10,000 seeded runs. It pairs monthly complaint volume and response rates during historical resampling, while translating outcomes to the fixed 2025 annual product volumes — preserving structure that would be lost if volumes and rates were drawn independently. It is a historical-pattern risk illustration, not a time-series forecast.

Current-condition distribution

MeasureSimulated result
Observed 2025 untimely cases9,523
Simulated mean9,541
Standard deviation408
5th percentile8,883
95th percentile10,230
Probability of exceeding 10,000 cases13.18%

Why debt collection dominates simulated variability

Under a covariance-based contribution calculation, debt collection contributes approximately 90.94% of total simulated variance. Checking/savings contributes 5.52%, credit cards 3.71%, and mortgages −0.17% — an offsetting covariance in these simulations, not a claim that mortgages pose no risk.

Distribution comparison of simulated annual untimely complaints under current conditions and 15% and 30% debt-collection improvements
Fig. 11Scenario distribution comparison — current conditions vs. 15% and 30% debt-collection improvements.Full size

Targeted debt-collection scenarios (only debt-collection rates reduced)

AssumptionSimulated meanMean avoidedP(> 10,000)
Current conditions9,541—13.18%
15% relative reduction8,3101,2300 of 10,000 runs
30% relative reduction7,0802,4610 of 10,000 runs

10Dashboard & recommendations

Final decision dashboard and business recommendations

The decision dashboard combines the observed KPIs, the portfolio scenarios, and the modeled risk distribution into a single executive view.

Executive decision analytics dashboard combining actual untimely cases, untimely rate, cases avoided, threshold risk, scenario comparison, Monte Carlo distribution, and decision insights
Fig. 12Executive decision dashboard — combined KPIs, scenarios, and modeled risk.Full size

Recommendations

  1. 01Prioritize a debt-collection diagnostic program, not a generic portfolio-wide intervention. Break the category down by company, issue, response classification, and month, then identify which operational hypotheses can be tested with additional internal evidence.
  2. 02Treat sudden performance shifts as questions before treating them as successes or failures. The CCS change is large and statistically distinguishable, but the available records do not identify the cause.
  3. 03Separate operational workload from response performance. Monitor complaint inflow, absolute untimely cases, and untimely rates together — January's volume concentration and April's rate peak show the danger of a single-metric dashboard.
  4. 04Make benchmarking conditional on composition. Geographic analysis should guide which company–product segments to review in which markets, not assign performance responsibility to states.
  5. 05Use scenarios to organize decisions and trade-offs. Before funding an intervention, add operational feasibility, cost, accountability, and internal validation — none of which can be inferred from this public dataset.
01

Data

CFPB Consumer Complaint Database

Consumer Financial Protection Bureau · Received Jan 1–Dec 31, 2025

  • Checking or savings accounts, credit cards, debt collection, and mortgages
  • Product, issue, company, state, and receipt date
  • Company response and Timely response? status
02

Limitations

Limitations

  • Selection and reporting: complaint submissions are not a random sample of customers or transactions; volume is not a measure of market share, prevalence, or total customer dissatisfaction.
  • Recorded timeliness: the analysis uses the CFPB's public response-status field; reasons for delay and operational accountability require additional evidence.
  • Confounding: differences across states, companies, issues, and months may reflect case mix, submission practices, and unobserved processes.
  • Causality: the CCS investigation documents a substantial association across periods, not an intervention effect; the split was identified after observing the time series.
  • Scenario limits: improvements are hypothetical, and the historical-pattern simulation cannot account for all future structural changes.
  • Geographic benchmark: the 50-state comparison excludes DC, territories, and missing-state records and differs from the all-record overall rate.
03

Technologies

  • Microsoft Excel
  • PivotTables
  • What-If Analysis
  • Data Tables
  • Goal Seek
  • Monte Carlo Simulation