How This Reference Is Researched
Every figure on this site traces to a named primary document. This page describes how that is done, and what happens when the sources disagree.
Who writes it
Every entry and guide on this site is written and reviewed by Bill Hartzer, who has worked in search since 1996 and has been retained as an expert witness in search engine optimization matters in state, federal, and international proceedings. The site is a reference written for litigators. It is not a marketing page, it gives no legal advice, and it takes no engagements.
Where the facts come from
Each page starts from a research file assembled before any prose is written. The research draws on primary documents wherever they exist: the text of the Federal Rules of Evidence and the Federal Rules of Civil Procedure, the Advisory Committee notes that explain them, court opinions read in full rather than through summaries, trial exhibits and findings of fact, decisions and consultations published by competition regulators, peer-reviewed and preprint research, and federal standards such as NIST and Office of Management and Budget guidance. Articles, law-firm alerts, and blog posts are used only to locate a primary document, never as the source of a figure.
Each figure is recorded with its publisher, its year, the file it came from, the page it appears on, and a note on what it does not establish. A statistic that cannot be traced to a document is not published, however often it is repeated elsewhere.
What is deliberately left out
Research produces more claims than it confirms. Items that could not be verified against a primary document are recorded as unverified in the project files and kept off the site. Two examples show why this matters. Proposed Federal Rule of Evidence 707 is often described as approved; the record shows it was published for comment and then returned for revision. And no reported decision has been found applying Rule 37(e) to analytics data, Search Console data, or server logs, so the page on spoliation of search data is written from the rule text and says so, rather than implying that courts have ruled.
Review before publication
Before a page is published, each statement of law is checked against the rule or opinion it relies on, each number is checked against the page of the document it came from, and each case citation is checked for its court, year, and holding. Pages that make time-sensitive claims, such as pending rule amendments, appeals, and platform retention settings, are kept on a review register and rechecked on a schedule.
How dates work
Every content page shows a published date and, where the content has changed, a last-reviewed date. Both come from a record of each page's authored content. A page's published date never moves. The last-reviewed date advances only when the substance of the page changes, not when the site's design or navigation changes, so a date on this site means the content was actually reviewed on that day.
When sources disagree, and corrections
Where two reliable sources disagree, the page says so, names both, and explains which one it follows and why. A court's finding of fact is preferred over a secondary account of it, and a document's own figures are preferred over a summary that rounds them. Where a figure is an estimate rather than a measurement, the page says what it estimates and from what.
If you find an error, the contact page explains how to report it. Corrections are made to the page itself and advance its last-reviewed date.