News
What an agency, a company, or a state office actually did — filed, voted, approved, or published. Stated in the past tense, attributed to the record, and written so the sequence of events stays intact.
Headline states the event
Editorial standards Lonnard James Broadband Review
Every story on this site — a speed test guide, a spectrum filing, a policy vote — is held to the same bar. What follows is the bar itself: where our facts come from, how we mark the difference between news, analysis and opinion, what happens when we get something wrong, and the wall we keep between the people who report and the people who pay.
Why these pages exist
A published standard is a promise a reader can check.
These standards describe how Lonnard James Broadband Review gathers information, labels what it publishes, and handles mistakes. They apply to every story on the site regardless of section — news, spectrum coverage, policy reporting, the guides on this site that explain how to run a reliable speed test, all of it.
They are written down for a practical reason. Broadband reporting sits close to money: carriers, vendors, consultants and trade groups all have an interest in how connection performance gets described. A speed test number, in particular, is easy to quote and easy to misread. Rather than promise neutrality in the abstract, this page shows the working rules we follow and the places where a reader can hold us to them.
Where a rule below conflicts with speed or convenience, the rule wins.
Every factual claim in a story traces to a document, a filing, an official statement, or a named source. Where a report relies on another outlet's work, we say so and link to it. In the current cycle, stories on the 7 GHz study, the October 29 FCC votes, and the Starlink Mobile approval all rest on published statements and filings rather than private briefings.
Traceable to a record
Whether the subject is an FCC docket, an NTIA band study, or a state broadband office publishing its own speed-check tool, the underlying document is identified so a reader can open it and read the same thing we read.
Named sources, named roles
People are described by what they do — an engineer at a state broadband office, a policy analyst at a trade group, a spokesperson for a carrier. Where a source cannot be named, the story says why and what the sourcing rests on instead.
Borrowed reporting, credited
When another newsroom breaks a story we then cover, that outlet is credited by name and linked in the piece. We add context and follow-up where we have it; we do not present another desk's work as our own.
No private briefings as a basis
Recent coverage has not relied on off-record briefings to establish the core facts. Where a briefing adds context, it is attributed as background and the load-bearing facts still trace to something public.
Numbers carry a source
Any figure that appears — including the more than 90 percent figure for Europe-to-Asia data capacity carried under the Red Sea — is attributed where it came from and qualified by what it measures. Numbers without an origin do not run.
News reports what happened. Analysis connects a decision to its likely consequences and is labeled as such. Opinion appears rarely and is marked in the headline. A reader should never have to guess which of the three they are reading, and we do not blend them inside a single story.
What an agency, a company, or a state office actually did — filed, voted, approved, or published. Stated in the past tense, attributed to the record, and written so the sequence of events stays intact.
Headline states the event
What a decision is likely to mean — for spectrum auctions, for how consumers read a speed test result, for infrastructure timelines. Reasoned from the record, labeled in the headline, and honest about where it stops being provable.
Headline marks it as analysis
Rare, written in the first person where it appears, and marked in the headline before a reader spends a sentence on it. Opinion never carries factual claims that have not been through the same sourcing check as a news story.
Headline says opinion
The test we apply is simple: if a reader skipped everything but the headline, would they still know which of the three they had opened? If the answer is no, the label gets rewritten before publication.
When we get a fact wrong, we correct the article and note the change at the foot of the story with a date. Substantive errors are flagged near the top where they cannot be missed.
We do not delete stories to remove an error, and we do not silently rewrite a headline after publication. A story that is wrong stays visible with the correction attached, so the record shows both the mistake and the fix.
If a correction changes how a reader should understand a number — a capacity share, a vote date, a filing's scope — the corrected passage is rewritten in plain language rather than annotated with a footnote.
How a correction runs
Spotted something that does not hold up? Send the article title and the disputed line to the newsroom and we will check it against the source.
Contact the newsroomCommercial independence
Coverage decisions are made by editorial staff. No advertiser, sponsor, or partner reviews a story before publication or requests a change afterward. If a commercial relationship ever exists with a company we cover, it is disclosed in the relevant article.
That applies to the practical things a reader would want to know. Sponsored material, if it ever runs, is labeled as sponsored and kept out of the news feed. Advertising is sold against audience, not against coverage topics. Nobody on the commercial side sees a draft.
The rule also governs what we publish about performance. Rankings are not for sale, test results are not produced on behalf of a carrier, and we do not accept placement in exchange for a mention in a story about speed test methodology or provider performance.
These are the places where broadband coverage most often goes wrong, so they are worth naming directly.
Provider rankings for pay
We do not publish provider rankings based on affiliate arrangements. An ordering that follows a commission is advertising, not reporting.
Carrier-commissioned tests
We do not run tests on behalf of a carrier. A measurement produced to a provider's brief answers the provider's question, not a reader's.
Fastest and best claims
We do not state that a service is the fastest or the best, because performance depends on address, wiring, and time of day. Where a number appears, it comes from a named source, including the more than 90 percent figure for Europe-to-Asia data capacity carried under the Red Sea.
Single readings as verdicts
One speed test on one evening is a snapshot, not a judgement. Our guides explain what to hold constant and how to repeat a measurement before treating it as evidence.
Poll coverage reports the sample, the pollster, and the question wording where available. The Reuters/Ipsos findings on AI risk are reported that way, including both the dissatisfaction with how risks are handled and the support recorded for new government rules.
We do not present a single poll as a settled public consensus. Where two surveys on the same subject disagree, that disagreement is part of the story rather than something averaged away.
We accept tips at the newsroom contact address. Sources are told in advance how their identity will be handled, and we do not publish identifying details without agreement.
Documents sent to us are reviewed before any excerpt appears in print. If a document cannot be used without exposing the person who sent it, we say so rather than publishing it and sorting the consequences later.
Anyone named critically in a story is offered a reasonable opportunity to respond before publication, and their response is included or noted. If they decline, the story says so.
This applies to companies as well as individuals, including the parties in the Disney license dispute and the certification bodies affected by the proposed testing rules. A company is offered the same window as a person and the same expectation of a straight answer.
Where a response arrives after publication, it is added to the story with a note rather than folded silently into the existing text.
We use software for transcription and research, and editors verify every fact before publication. We do not publish AI-generated text as reporting, and we do not generate quotes.
Where a story concerns AI policy, such as the Cantwell framework, that coverage follows the same sourcing rules as any other beat. The subject being AI does not change the standard applied to the facts about it.
Transcription tools speed up interviews. They do not decide what gets published, and no sentence attributed to a person is reconstructed by a model.
How this reaches you
Standards only matter if they change what you can do with the reporting. Three practical consequences follow from the rules above.
If a story says a region carries a share of a data route or an agency finished a band study, the filing or statement behind that line is identified. You can read the same source and decide for yourself.
The headline tells you whether something is a report, an analysis, or an opinion piece. That matters most on a subject like spectrum policy, where the difference between a ruling and a prediction is the whole point.
Dated correction notes stay attached to the story they belong to. A reader arriving a week later sees the same fix as someone who was there when it ran.
The details people tend to want after reading the standards above.
Yes. Send the readings, the address type, the router and the time of day. We are more likely to write about a pattern across several homes than about a single household, and we will not turn your message into a story without agreeing the ground rules with you first.
Editorial decisions are made by editorial staff, and no advertiser or partner reviews a story before publication. If a commercial relationship ever exists with a company we cover, the relevant article says so.
The article is corrected at the point of error and a dated note is added at the foot of the story. Substantive errors are flagged near the top. Stories are not deleted to hide a mistake.
Because a single speed test result says little on its own. Performance depends on the address, the wiring inside the home, the router, and the time of day. Guides on how to run a reliable speed test and how to read the results do more for a reader than a leaderboard.
Use the contact page or write to the newsroom address. Tips, corrections and story suggestions all come through the same desk, and we reply to anything that carries a question we can answer.
The rules above are the operating instructions for everything else on this site — the spectrum and policy coverage, the infrastructure reporting, and the guides on measuring a connection properly. If something you read here does not match these standards, tell the desk.
Newsroom address: 3003 13th Ave W, Seattle, WA 98119-2021.
Phone +1 206 438 9529 — Monday to Friday, 9:00 AM to 6:00 PM.