Our empirical data policy
Many commercial speed test sites publish automated rankings claiming to identify the "fastest and slowest cities in America." We deliberately do not publish synthetic municipal average speed tables or city leaderboards. Aggregating thousands of voluntary browser tests across an entire municipality produces numbers that are statistically misleading and practically useless for individual households.
A single municipal average blends high-speed commercial gigabit fiber links with aging residential DSL connections, conflates fixed wireline lines with mobile cellular tests, and suffers from severe self-selection bias: users test their connections disproportionately when their network is lagging or experiencing an outage. When a website reports that a city has an "average download speed of 185 Mbps," that figure does not guarantee you can buy 185 Mbps at your address, nor does it tell you whether your apartment building suffers from severe peak-hour latency spikes. We only publish verified measurement methodologies and address-specific verification frameworks.
Why municipal averages are misleading
Broadband infrastructure is not deployed evenly across municipal boundaries. Physical cables, optical distribution nodes, and provider franchise territories are divided at the street, block, and parcel level. Relying on a city-wide speed statistic creates three critical blind spots:
| Metric Dimension | What City Aggregations Report | Real Physical Reality at Your Address |
|---|---|---|
| Access Technology | City reported as "90% High-Speed Covered" | A fiber boundary stops one block away; your parcel is locked into legacy copper DSL or a saturated cable node. |
| Symmetry & Upload | High average download (e.g. 350 Mbps) | Cable upload remains capped at 10 to 35 Mbps, causing video call freezing whenever cloud backups trigger. |
| Peak Contention | Unweighted arithmetic mean of all tests | Shared neighborhood optical nodes choke during peak evening streaming hours (7:00 PM – 11:00 PM). |
| Building Wiring | ISP-advertised street speeds | Degraded building telephone risers or dense Wi-Fi radio crowding throttle indoor throughput to a fraction of the line rate. |
Arithmetic mean vs. median distortion
Metropolitan speed aggregations that use arithmetic means are heavily skewed by commercial connections and top-tier multi-gigabit fiber subscribers. A small percentage of 2 Gbps or 5 Gbps subscribers mathematically pulls the city average upward, even when the vast majority of residential households in older suburbs receive 50 to 100 Mbps on aging coaxial or copper lines. Even median metrics fail to capture the bifurcated reality of modern broadband: cities rarely have a uniform middle tier; they have residents with symmetrical fiber access and residents without it.
The "across the street" divide
Telecom infrastructure reflects decades of historic right-of-way agreements, utility pole attachment disputes, and capital investment boundaries. One side of an arterial avenue may have been built out by an independent fiber overbuilder, while the opposite side remains tied to legacy incumbent copper. Assuming your address receives good internet because your friend three blocks away gets gigabit fiber is one of the most common mistakes tenants and homebuyers make.
Physical layers shaping local speeds
Your actual internet performance is dictated by four distinct physical and network layers between your room and the broader internet exchange:
1. Last-mile physical medium
The wire or radio link reaching your exterior demarcation determines your hard physical ceiling for throughput, latency, and packet stability:
- Fiber to the Home (FTTH): Glass optical fiber connects directly to an Optical Network Terminal (ONT) at your building. Fiber delivers symmetrical download and upload speeds (e.g. 500/500 Mbps or 1000/1000 Mbps) with ultra-low idle latency (typically 2 to 10 ms) and zero copper electrical attenuation over suburban distances.
- Hybrid Fiber-Coaxial (HFC / Cable): Fiber runs to a neighborhood optical node, where signals convert to coaxial copper cables serving clusters of 100 to 500 homes. Cable offers high download speeds (300 to 1,000 Mbps) but severely asymmetric upload speeds (typically 10 to 35 Mbps on DOCSIS 3.1) and higher susceptibility to upstream radio-frequency ingress noise.
- 5G Fixed Wireless Access (FWA): Wireless signals transmit from a nearby cellular macro or small-cell tower directly to an indoor or window-mounted gateway. Speeds fluctuate dynamically based on cell tower line-of-sight, building wall materials (such as Low-E glass), atmospheric conditions, and real-time mobile handset traffic priority.
- Copper DSL: Signals travel over unshielded twisted-pair telephone copper from a central office or street-side cabinet (DSLAM). Bandwidth degrades rapidly with wire distance: homes within 1,000 feet may receive 50 to 100 Mbps on bonded VDSL2, while homes 5,000 to 10,000 feet away drop below 10 to 25 Mbps.
2. Neighborhood node oversubscription
Broadband networks are engineered around statistical multiplexing: providers assume not every household will download or upload maximum bandwidth simultaneously. Cable (DOCSIS) systems share upstream and downstream RF channels across an entire neighborhood node. In dense suburban subdivisions or apartment complexes, simultaneous 4K video streaming, gaming updates, and cloud backups between 7:00 PM and 11:00 PM can saturate shared node channels, causing bufferbloat latency spikes and packet loss.
3. Regional routing and internet exchange (IXP) proximity
Your physical distance from a major carrier-neutral Internet Exchange Point (IXP) sets your absolute latency baseline. If you live in a tier-one peering hub (such as Northern Virginia, Chicago, Dallas, Silicon Valley, or Atlanta), your idle round-trip time (ping) to major cloud providers, CDNs, and gaming servers may sit below 10 ms. If you live in a smaller city or rural municipality without local peering, your ISP must backhaul traffic hundreds of miles to the nearest regional hub, adding 20 to 50 ms of unavoidable physical latency.
Apartments, risers, and Wi-Fi crowding
Multi-Dwelling Units (MDUs)—including apartment buildings, condominiums, and student housing complexes—face unique architectural bottlenecks that frequently decouple in-unit performance from the city's broader broadband quality.
| MDU Bottleneck | Physical Cause | Diagnostic Symptom | Mitigation Strategy |
|---|---|---|---|
| Bulk Provider Lock-In | Landlord signed exclusive marketing or bulk billing contract with a single legacy cable ISP. | Zero competitive fiber choices available despite fiber conduits existing under the street sidewalk. | Review lease terms before signing; check whether 5G FWA or fixed wireless has an unobstructed window line-of-sight. |
| Aged Riser Cables | Basement demarcation receives fiber, but internal vertical risers use unshielded Cat 3 copper or split coax. | Gigabit plan purchased, but wired wall jacks negotiate at 100 Mbps or suffer high packet loss. | Have ISP run a dedicated Ethernet run from the basement or install the modem/ONT at the primary entryway demarcation. |
| 2.4 GHz Congestion | Dozens of tenant access points broadcast on only three non-overlapping channels (1, 6, 11). | Severe Wi-Fi latency spikes, dropped video calls, and speeds collapsing below 20 Mbps on wireless. | Disable 2.4 GHz for primary work devices; bind laptops and phones exclusively to 5 GHz (80 MHz) or Wi-Fi 6E/7 (6 GHz). |
| Co-Channel Interference | Neighboring 5 GHz routers use overlapping 80 MHz channel widths, forcing channel contention. | Periodic jitter bursts during evening hours despite strong four-bar wireless signal icons. | Use a Wi-Fi spectrum analyzer app to locate clear Dynamic Frequency Selection (DFS) channels or reduce channel width to 40 MHz. |
Address-level verification playbook
Instead of searching for city averages, use official federal databases and mandatory transparency disclosures to verify exact parcel availability before committing to a provider:
1. Search the FCC National Broadband Map
The Federal Communications Commission maintains a parcel-level database at broadbandmap.fcc.gov. Enter your exact street address, apartment unit, and postal code:
- Filter by Technology: Toggle the view to "Fixed Broadband" and filter out Satellite and Unlicensed Wireless to inspect genuine wireline infrastructure (Fiber, Cable, Copper).
- Examine Advertised Rates: Review the maximum advertised download and upload speeds reported by each listed provider.
- File an Availability Challenge: If the FCC map claims a provider offers fiber or gigabit service at your location, but the provider's sales representative states service cannot be installed without thousands of dollars in construction fees, submit an official "Availability Challenge" directly through the FCC portal to correct the public record.
2. Inspect the FCC Broadband Consumer Label
Under federal regulations, all major wireline and wireless internet service providers in the United States must publish standardized Broadband Consumer Labels for every active consumer plan. Always request this label before ordering service. It discloses:
- Typical Download & Upload Speeds: The actual median throughput delivered during peak usage hours (7:00 PM – 11:00 PM), rather than theoretical burst maximums.
- Typical Latency: The median round-trip ping time measured across the provider's network in milliseconds.
- Monthly Data Allowances: Clear notification of whether the plan imposes a hard monthly data cap (e.g. 1.2 TB) and what overage charges apply ($10 per 50 GB block).
- All-In Monthly Pricing: Itemized listings of promotional expiration windows, mandatory gateway equipment rental fees ($10–$15/month), and government regulatory recovery charges.
Pre-move internet due diligence checklist
If you are relocating to a new apartment, condominium, or house, complete this 6-step verification process before finalizing your lease or purchase agreement:
- Inspect physical demarcation hardware during the tour: Look inside the utility closet, media panel, or exterior wall. Identify whether the unit contains an Optical Network Terminal (ONT) indicating FTTH fiber, a coaxial threaded barrel connector indicating cable, or a legacy RJ11 telephone jack.
- Request the current occupant's provider and plan: Ask the current tenant or homeowner which ISP they use, what real-world speeds they experience during evening hours, and whether they experience connection drops during remote work hours.
- Test indoor cellular reception: Walk through interior rooms, basements, and home office spaces while running a mobile speed test on your smartphone. Verify whether cellular coverage is robust enough to serve as a mobile hotspot backup during wireline outages.
- Perform a Wi-Fi spectrum audit: Use a free Wi-Fi analyzer mobile app to count how many neighboring access points broadcast within range. If you detect 30+ overlapping networks, plan to invest in a Wi-Fi 6E/7 router capable of using the uncongested 6 GHz band.
- Audit internal Ethernet wall jacks: Check whether room wall plates feature true Cat 5e or Cat 6 RJ45 jacks connected to a central patch panel, or if they are daisy-chained telephone voice lines that cannot carry gigabit data.
- Run a day-one wired baseline test: Immediately following service activation, connect a laptop directly to your gateway using a Cat 6 Ethernet cable and execute an initial broadband speed test. Compare your measured idle ping, download, upload, and loaded bufferbloat against the figures disclosed on the provider's FCC Broadband Label.