Dashboards
Capability proof
U.S. PPP Loan Data
Search and filter nearly one million SBA Paycheck Protection Program loans by company, geography, industry, size, and jobs.
Explore the PPP loan dataThe PPP release is an unusually detailed snapshot of American business activity during the COVID lending window. Each record represents an organization that received program funding and includes useful signals such as borrower name, geography, loan amount, reported jobs, business type, lender, and—in most cases—an industry classification.
A PPP record does not prove that a company is operating today, but it does show that the borrower existed and received support during the program period. When the data is cross-referenced with NAICS descriptions and made searchable, it becomes a powerful starting point for B2B prospecting, industry research, local market mapping, and acquisition-target discovery.
The outcome looks simple.
The work behind it wasn’t.
Make a massive flat dataset useful
Why it was difficult
Nearly one million records are technically accessible in a CSV, but not practically explorable for most business users.
The response
Normalize the source fields, join NAICS codes to readable industries, preserve import integrity, and index the data for fast narrowing from a national view to a focused company list.
Estimate scale without false precision
Why it was difficult
Loan amounts and reported jobs provide useful company-size signals, but they are not reported revenue.
The response
Translate the available signals into broad estimated revenue brackets, label them clearly as estimates, and let users combine them with jobs, industry, and geography.
Make every dimension work together
Why it was difficult
The real value comes from combining state, business type, industry, company, size, and jobs—not browsing one column at a time.
The response
Use linked filters, live summary metrics, sortable results, validated requests, and bounded pagination to keep complex exploration quick and understandable.
Move from discovery to action
Why it was difficult
Finding a promising company is only the beginning of a sales or acquisition-research workflow.
The response
Link every borrower name to a prefilled company-and-state Google search, creating an immediate handoff into current research and contact discovery.
Built for people trying to find the signal inside a very large market.
B2B growth teams
Sales and business-development teams targeting established organizations.
Reduce a broad territory or industry to a much more relevant prospect universe.
Combine geography, industry, jobs, and estimated revenue bands, then move directly into company research.
Acquisition teams
Acquisition entrepreneurs and corporate-development teams.
Surface smaller but still substantial companies inside a familiar market.
Reveal traditional operating businesses that may never appear in startup databases or conventional deal feeds.
Market intelligence
Researchers, advisors, and economic-development teams.
Understand an industry's footprint within one state or across the country.
Move between market-level totals and the underlying organizations using consistent public signals.
A one-time public release,
made genuinely useful.
One searchable national snapshot
Across the imported records
Mapped through NAICS descriptions
National geographic coverage
Another signal of company scale
From a national dataset
to a focused company search.
A national business dataset becomes approachable through clear market totals and one focused search surface.
Geography, estimated company scale, jobs, and industry work together to reduce a broad market to a useful target set.
The resulting companies stay easy to scan and compare across location, loan amount, jobs, industry, type, and lender.
Every borrower link opens a ready-made company-and-state search so discovery can move directly into current research.
Start with a city, state, industry, employee range, or estimated revenue band; narrow almost one million records to a focused company list; compare the market totals; and continue directly into current web research.
For prospecting, that can replace hours of assembling an initial company universe from scattered sources. For market and acquisition research, it can surface established-industry companies that are easy to miss in datasets dominated by startups, public companies, or self-reported directories.
Large-scale public-data product design: acquiring and normalizing a difficult government dataset, enriching it with industry context, translating imperfect business signals into clearly labeled estimates, and delivering a fast, public research interface.
See the work