One provider, many services: the category overlap study
The catalogue held 46,049 service assignments for 13,594 provider records. This study shows why category totals cannot be treated as separate companies or customer access.
91.6% of provider records crossed category lines.
There were 46,049 service assignments for 13,594 provider records, or 3.39 assignments per record.
What established research measured, and what this study adds.
The FDIC household survey and Federal Reserve household report measure account use, financial wellbeing and access barriers from the consumer side.
FDIC household banking survey ↗ServeAssess measured the provider side and found that 91.6 percent of records crossed service categories.
The most common service-category pairs
| Category pair | Records | Share of providers |
|---|---|---|
| Banking + consumer lending | 7,744 | 57.0% |
| Banking + business lending | 5,260 | 38.7% |
| Business + consumer lending | 5,089 | 37.4% |
| Investing + wealth management | 4,714 | 34.7% |
| Banking + mortgages | 4,148 | 30.5% |
| Consumer lending + mortgages | 3,998 | 29.4% |
| Business lending + mortgages | 3,828 | 28.2% |
| Investing + retirement | 3,613 | 26.6% |
Every percentage and rate belongs to the sample, threshold and period shown beside it. A descriptive relationship is not presented as causation.
The bottom line
Some 91.6 percent of provider records appeared in more than one service category. The average record carried 3.39 assignments, while only 1,144 records appeared in a single category.
Adding category counts would therefore overstate the number of distinct providers. It would also hide the legal and operational links between deposit taking, lending, payments, advice and other services.
Banking and lending formed the largest cluster
Banking and consumer lending appeared together on 7,744 records. Banking and business lending overlapped on 5,260, while business and consumer lending overlapped on 5,089.
Those pair counts are service assignments inside the catalogue. They do not prove that every product was open to every customer or state on the snapshot date.
Advice formed a second cluster
Investing and wealth management appeared together on 4,714 records. Investing and retirement overlapped on 3,613, while retirement and wealth management overlapped on 3,334.
The pattern reflects the way advisory firms describe connected services and the way registry derived classifications are applied. A consumer still needs the exact legal entity, account type and permission relevant to the proposed service.
Provider supply is not financial access
The FDIC household survey and the Federal Reserve household report measure whether people use accounts, obtain credit and experience financial strain. A provider catalogue measures organizations and recorded services.
A state with many headquarters records can still have access barriers. A national online provider can serve residents without being headquartered in their state. The measures answer different questions.
Why entity resolution matters
A single brand may rely on a bank, lender, broker dealer, investment adviser or programme manager for different products. Conversely, one legal institution can operate several consumer brands.
Category overlap should therefore lead to more precise comparisons, not a universal provider score. The contract party and responsible regulator must be resolved for the service being assessed.
Method
The study split the pipe delimited categories field in the 25 July provider export, removed blank entries and counted assignments per provider. Pair counts used each unique pair once per record.
Canonical combinations sorted category names before counting so a different source order did not create a false new combination. Percentages use 13,594 provider records as the denominator.
Limits
Assignments can be inherited from registry rules or provider research and do not establish current product availability. Broad categories also contain products with different contracts, risks and customer groups.
The study measures catalogue structure. It is not a market share estimate and it does not identify financial service deserts.
Open the underlying records
Every result on this page belongs to the dated sample above. Later catalogue updates do not silently change its denominator.