Compare ISP Plans and Performance Evidence
The useful question is not which provider wins a global leaderboard. It is which available plan fits your address, equipment, budget and applications. Start with the terms offered to you, then collect comparable measurements rather than treating a brand-wide score as a prediction for your home.
Compare the plan, not just the headline rate
For US plans, the FCC's Broadband Consumer Labels guidance explains where to find information about prices, introductory rates, data allowances and broadband speeds. Keep the actual label or plan document for the offer you are considering. A label for another tier or another address is not a substitute.
Use the same headings for each offer. Record the regular price as well as the introductory price, equipment charges, the download and upload information provided, data allowances, and any commitment or cancellation terms. Where a field is missing, mark it “not supplied” and ask the provider instead of filling it with a national average.
- Address and offer
- Can you actually order this exact service at your address, and what equipment is included?
- Capacity in each direction
- Compare upload as well as download information. A household sending large files has a different constraint from one mainly downloading them.
- Cost over your intended term
- Use the offer's stated recurring and one-off charges. Do not compare one provider's introductory price with another's standard price.
- Performance evidence
- Ask what period, location and service tier a published measurement covers. Keep that separate from the advertised plan description.
One worksheet for the offers you can actually order
This is a blank comparison aid, not provider research or a measured ranking. Copy it into a private note for the exact offers available at your address. Record the source link and date beside each offer. Use not supplied for missing terms; do not fill a gap with a national average or a different service tier.
| Field to check | Offer A | Offer B |
|---|---|---|
| Provider and exact plan identifier | not supplied | not supplied |
| Availability at your address (check privately) | not supplied | not supplied |
| Regular monthly price and required fees | not supplied | not supplied |
| Promotion price, duration and conditions | not supplied | not supplied |
| Equipment and installation charges | not supplied | not supplied |
| Download and upload terms, with units | not supplied | not supplied |
| Data allowance and excess-use terms | not supplied | not supplied |
| Commitment, cancellation and renewal terms | not supplied | not supplied |
| Source URL and date saved | not supplied | not supplied |
For US offers, use the provider's actual Broadband Consumer Label alongside the FCC guidance above. Elsewhere, obtain the written offer and applicable local disclosures. An advertised rate, a typical rate and a contractual minimum are different statements: keep the provider's wording rather than silently treating them as interchangeable.
Next action: ask the provider to supply missing terms and confirm the address-specific offer before choosing. This worksheet does not calculate eligibility, predict delivered speed, or establish a cancellation right.
Keep observations separate from offer terms
Do not enter a household speed result as if it were an advertised plan field. Record the device, wired/Wi-Fi condition, date and timezone, service and endpoint, download/upload units, any timing values, and completion or failure status in a separate note. Keep unsuccessful attempts and warnings rather than replacing them with zero.
Use the controlled Wi-Fi versus Ethernet sequence and the copyable observation record. A browser test describes one observed path, not a provider-wide performance result.
Take a specific question to support
Privately provide the exact plan identifier, affected task, device and connection method, dated observations and failures, and what changed between attempts. Ask support to check the provisioned tier, connection faults and supplied equipment, then retain the case reference and agreed next action. Redact account numbers, service addresses, public IP addresses and bills before sharing anything publicly.
Make your own tests comparable
Ofcom recommends testing over several days at different times and trying Ethernet instead of Wi-Fi when investigating speed. Its practical broadband guidance also identifies device setup, other household use and time of day as possible influences. These troubleshooting principles are useful beyond the UK; contractual rights still depend on your location and agreement.
- Write down your plan and the device used. If possible, start with a computer connected to the router by Ethernet, rather than changing both the device and connection method at once.
- Pause discretionary downloads and backups for a baseline. Note unavoidable household activity instead of pretending the line was idle.
- Use the same test service and endpoint selection where possible. Record when the endpoint changes.
- Repeat at different times. Keep unsuccessful runs with an explanation; selecting only the best result hides useful evidence.
- Then test the Wi-Fi location where you actually work or play. Treat it as a separate condition, not another sample of the wired baseline.
A browser test estimates a path involving your device and a test endpoint. A poor result is a reason to investigate, not proof of an ISP-wide fault. See the advertised-versus-measured checklist for a support-ready record.
What would justify a provider ranking?
A ranking needs a defined population, collection period, sample counts, exclusions and a reproducible calculation. It also needs to explain whether it compares similar plans and conditions. The FCC's historical measurement methodology describes service-tier validation, data cleaning and limits on which providers appear in a comparison. That historical method is an example of documentation, not current performance evidence for the providers below.
VelocityVerify is not presenting verified provider scores, city medians or a measured leaderboard on this page. The directory is alphabetical and unranked. Listing a provider does not establish availability, endorsement or performance at your address.
Provider names, not separate reviews
The names below are an alphabetical, unranked reference list, not current offer or coverage research. Use one shared comparison worksheet for the exact offers available to you. We do not send you through separate copies of a generic checklist. Older provider links remain compatibility pages pointing back to this guide.
- AT&T Fiber
- BT Ultrafast
- CenturyLink DSL
- Comcast Xfinity
- Cox Communications
- Deutsche Telekom
- Frontier DSL
- Google Fiber
- NBN Co (Australia)
- NTT East (Japan)
- Optus (Australia)
- Orange France
- SK Broadband (Korea)
- SoftBank
- Spectrum
- T-Mobile Home Internet
- Verizon 5G Home
- Verizon Fios
- Virgin Media (UK)
- Ziggo (Netherlands)
Choose the next useful action
If you have not ordered yet, save comparable plan documents and ask about missing terms. If you already have service, establish a repeatable baseline before buying hardware or upgrading. If a particular game or meeting service fails while general tests look normal, record that application's time and symptoms separately; a generic speed result does not measure every destination.
Use the diagnostic tool as one observation, not a provider verdict. For an explanation of what aggregate statistics can and cannot show, read how to read broadband speed reports.