Our Methodology

BroadbandGrade scores internet service for every U.S. ZIP code using publicly available federal data. We combine four broadband metrics into a single Broadband Grade (A through F) so consumers, journalists, and researchers can quickly assess internet quality in any area without parsing raw FCC filings.

Data Sources

Our primary data source is the FCC Broadband Data Collection (BDC), the federal database of residential and small-business broadband availability. The BDC replaced the older Form 477 framework in 2022 after a multi-year FCC modernization effort and now produces address- and location-level data with twice-yearly updates. We pull the BDC release directly from broadbandmap.fcc.gov/data-download/nationwide-data, aggregate the BSL (Broadband Serviceable Location) records up to ZIP-code-tabulation areas using the Census Bureau ZCTA-to-county and ZCTA-to-place crosswalks, and feed the standardized result into the grading pipeline.

Two supplemental sources inform the methodology and qualitative context. The National Telecommunications and Information Administration (NTIA) administers the $42.5 billion Broadband, Equity, Access, and Deployment (BEAD) Program, the largest federal broadband-funding initiative in U.S. history. NTIA's state-by-state allocation tables and unserved/underserved location counts give us a forward-looking view of where coverage is expected to improve. Measurement Lab (M-Lab) publishes anonymized open-data speed-test microdata that we use to sanity-check FCC advertised speeds against real-world throughput.

How We Calculate the Broadband Grade

Every ZIP code receives a 0-100 Broadband Score, mapped to letter grades A through F. The score is a weighted composite of four factors, each normalized to a 0-100 sub-score before being combined:

• Download Speed — 40% weight. Calculated from the share of locations in the ZIP with access to 100 Mbps and 1,000 Mbps (gigabit) download tiers. ZIPs where every location has gigabit available score 100; ZIPs with no provider above 25 Mbps score zero. The 100 Mbps threshold matches the FCC's March 2024 broadband benchmark of 100/20 Mbps.
• Provider Competition — 30% weight. Calculated from the share of locations with three or more providers offering at least the FCC broadband benchmark (100/20 Mbps). Multi-provider markets correlate strongly with lower prices and higher service quality. ZIPs with universal three-plus competition score 100; effective monopolies score zero.
• Fiber Access — 20% weight. Calculated from the share of locations with fiber-to-the-home (FTTH) deployment. Fiber offers symmetric speeds, lower latency, and higher upload caps than any competing technology, and is the architecture targeted by the bulk of NTIA BEAD funding.
• Upload Speed — 10% weight. A proxy score derived from gigabit and fiber availability, reflecting upstream capacity that matters most for video calls, cloud computing, and content creation. Cable internet has historically been highly asymmetric (1,000 Mbps down / 35 Mbps up); fiber is symmetric, which dominates the upload sub-score in most ZIPs.

Grade Scale

Provider-Level Grades

Each provider's headline grade is computed as the average Broadband Grade across every ZIP code in the provider's reported coverage footprint. A provider with a wide but speed-poor footprint (e.g. legacy DSL across rural counties) will earn a lower aggregate grade than a provider with a narrow but high-quality footprint (e.g. a fiber overbuilder in a single metro). The same ZIP-level methodology rolls up to county, state, and national rankings.

Limitations

Three caveats apply to any FCC-derived broadband ranking, including this one:

• Advertised vs. measured. FCC data reflects advertised maximum speeds, not actual experienced throughput. Real-world performance depends on network congestion, in-home Wi-Fi, distance from the central office or node, and the specific tier the customer subscribes to. M-Lab measurements typically run 60-80% of advertised peaks at evening hours; this gap is roughly consistent across major providers, so it does not materially distort relative rankings.
• Availability vs. subscription. Provider counts reflect availability — “a provider serves this location” — not affordability or actual subscription rates. A ZIP with three providers all priced above $90/month may technically be highly competitive but functionally inaccessible for low-income households. The NTIA Affordable Connectivity Program historical eligibility data (now sunset) and Lifeline subscription numbers are useful supplementary lenses on this question.
• Reporting lag. Each BDC release reflects providers' filings as of a cutoff date six to nine months prior. Rapid fiber build-outs (especially BEAD-funded ones) and late-stage decommissioning of copper plant will be visible in subsequent releases, not the current one.

Update Cadence

Data was last refreshed on April 13, 2026. We re-process all 34K ZIP codes each time the FCC publishes a new BDC release, typically every six months. New ZIPs and new providers are added automatically; deprecated ZIPs (consolidated by the Census Bureau) and exited providers (mergers, exits, asset sales) are removed.

Frequently Asked Questions