Search Terms and Their Evidentiary Significance

What Google records about a site, in the party's own account

The search terms on this page are the ones that appear in Search Console exports, analytics reports, crawl files and expert reports, and each is defined here from the documentation that defines it, with a note on who holds the underlying record and which rule the term usually reaches. Search terms are not interchangeable: an impression and a click are counted by different systems, a backlink and a disavow file are controlled by different parties, and a search volume figure is an estimate while a server log is a record. Treating these search terms as a single category is how an opinion ends up resting on the wrong kind of data.

This glossary sorts the vocabulary of a search dispute by where the data comes from, because that decides how it is authenticated, whether it is a record or an estimate, and how long it survives. One phrase recurs: a “sufficiency target” is the data source an opposing expert attacks under Rule 702(b), the requirement that an opinion rest on sufficient facts or data. Google “doesn’t guarantee that it will crawl, index, or serve your page,” and no documentation promises a ranking outcome from any technique below.

Google Search Console and the Performance report

“The Performance report shows important metrics about how your site performs in Google Search results.” The figures are Google’s but land in the party’s account, the nearest thing to a party’s record of engine-side behavior. Google’s Analytics help states the retention: “Search Console keeps data for the last 16 months.” Data not exported before the window rolls is lost, the Rule 37(e) exposure; every export pulled is “facts or data considered” under Rule 26(a)(2)(B)(ii).

Impression

“An impression means that a user has seen (or potentially seen) a link to your site in Search, Discover, or News.” It is a engine-counted event, not a user action, and aggregation changes the count: by property, a site appearing twice on one page “counts as a single impression,” while by page “Each unique URL in a Search result is counted separately.” A comparison across periods is valid only at the same aggregation.

Click

Google counts “any click that sends the user to a page outside of Google Search, Discover, or News.” That is the engine-side count of a visit; the GA4 session is the site-side count of the same visit, and the two never reconcile exactly, so the expert has to state which one is “the traffic.”

Average position

“A relative ranking of the position of your link on Google, where 1 is the topmost position,” by a method Google says “might change in the future.” Because it is averaged across queries, new low-position rankings worsen the average while traffic grows. It is the first-party alternative to a rank tracker.

Crawl Stats report

“The Crawl Stats report shows you statistics about the search engine’s crawling history on your website,” with host status covering “the past 90 days.” Google warns of “slight differences” from the site’s own logs; it is a short-lived cross-check, not a substitute.

Core Web Vitals

“A set of metrics that measure real-world user experience for loading performance, interactivity, and visual stability of the page,” drawn from the Chrome User Experience Report, gathered “from actual users visiting your URL.” The owner neither records nor controls it, and the documentation says only that good scores align “with what our core ranking systems seek to reward.”

What Google decides about a site and tells the owner in writing

Crawling and indexing

“Google downloads text, images, and videos from pages it found on the internet with automated programs called crawlers,” then stores them in the index, “a large database hosted on thousands of computers.” Both are Google’s conduct and neither is guaranteed: “Indexing isn’t guaranteed; not every page that Google processes will be indexed.” No party can produce the index; indexing on a past date is inferred from the Page indexing report or from captures made at the time.

Manual action

Issued when “a human reviewer at Google has determined” that pages are “not compliant with Google’s spam policies”; Google will “notify you in the Manual actions report,” and the owner may “Request Review.” It is the one ranking outcome Google documents in writing, with a date, a reason, and a reconsideration history, inside the party’s account: the strongest evidence that Google penalized a site. Its absence does not show the absence of an algorithmic effect.

Spam policies and link spam

“Link spam is the practice of creating links to or from a site primarily for the purpose of manipulating search rankings,” detected through “automated systems and, as needed, human review that can result in a manual action.” The policies are the published standard against which an agency’s or competitor’s conduct is measured; the version in force on the conduct date governs.

Core update

“Several times a year, the search engine makes significant, broad changes to our search algorithms and systems,” announced on its list of ranking updates; these “don’t target specific sites or individual web pages.” The dated list is a public timeline for testing whether a drop coincides with the search engine rather than with the defendant’s conduct, the Rule 702(d) confound question.

What the site owner's own systems record

Server access log

Google does not define the term but assumes the file exists, telling owners to “Run a reverse DNS lookup on the accessing IP address from your logs” to catch “troublemakers” claiming to be Google. It is the only record of crawler and visitor requests the party itself generates: a business record under Rule 803(6) and a candidate for Rule 902(13) certification. Retention is host-set and short; user-agent strings can be spoofed, so counting “Googlebot” hits without verification invites a Rule 702(c) attack.

GA4 data retention

Google Analytics 4 lets a property set user-level and event data retention to 2 or 14 months, longer on Analytics 360. Google states that “The data retention setting does not affect standard aggregated reports” and “only affects explorations and funnel reports.” The page as read states no default, so none is stated here. Aggregate counts survive; explorations by landing page or user path lose history on schedule, a Rule 37(e) problem once litigation is foreseeable.

Session

“A session is a period of time during which a user interacts with your website or app,” and “By default, a session ends or times out after 30 minutes of user inactivity.” Sessions are the party’s own measure of visits, produced by code it installed and dependent on timeout, consent, and filters; a drop claim built on them has to disclose configuration changes.

Event

“An event allows you to measure a specific interaction or occurrence on your website or app.” Custom events are party-defined, so their definitions are discoverable and their consistency is a foundation question.

Engagement rate and bounce rate

An engaged session “Lasts longer than 10 seconds,” “Has a key event,” or “Has 2 or more screen or page views”; “The bounce rate is the percentage of sessions that were not engaged.” That is not the Universal Analytics bounce rate, so a chart spanning both compares different metrics.

Consent mode and behavioral modeling

“Consent mode lets you communicate your users’ cookie or app identifier consent status to Google”; when visitors decline, Google fills the gaps with modeling that “uses machine learning to model the behavior of users who decline analytics cookies.” Part of the count is then a Google estimate inside a first-party record, and the expert must say so.

Sampling and the “(other)” row

“Data sampling is the data-analysis practice of analyzing a subset of data,” applied above 10 million events, and Google “condenses less common values under the (other) row.” An exhibit built from a sampled report is at best a Rule 1006 summary of an estimate; the BigQuery event table, if enabled, is the original.

Organic, referral, and direct channels

Organic Search matches when the source is on Google’s list of search sites or “Medium exactly matches organic”; Direct is arrival “via a saved link or by entering your URL.” “Organic traffic” is thus a Google classification applied to the party’s tagging; mis-tagged campaigns and missing referrers move sessions between channels with no change in behavior, so an organic-loss figure must be reconciled against Direct and Referral.

What the page and the server configuration contain

HTTP status code

A 2xx response “doesn’t guarantee indexing,” Google “doesn’t use the content from URLs that return 4xx status codes,” and persistent 5xx responses get a URL removed from the index. It is a server response the owner controls and the access log records; historic codes are provable only from preserved logs or captures.

301 and 302 redirects

“Redirecting URLs is the practice of resolving an existing URL to a different one.” Google treats 301 as permanent and 302 as temporary, and for permanent redirects “the indexing pipeline uses the redirect as a signal that the redirect target should be canonical.” The page does not say redirects pass ranking value; redirecting disputed URLs after a demand letter is the Rule 37(e)(2) intent pattern.

Canonical tag

A rel="canonical" annotation is “A strong signal that the specified URL should become canonical,” and of the canonicalization methods “none of them are required.” Google calls it a signal, never a directive. Search Console assigns all variations’ clicks, impressions, and position to the canonical URL, so a canonicalization change moves reported traffic between rows with no change in user behavior.

noindex

“noindex is a rule set with either a <meta> tag or HTTP response header,” and “For the noindex rule to be effective, the page or resource must not be blocked by a robots.txt file.” An accidental noindex is a recurring cause of traffic loss attributable to the owner or its agency rather than a defendant, provable only from captures.

robots.txt

“A robots.txt file tells search engine crawlers which URLs the crawler can access on your site”; it “is not a mechanism for keeping a web page out of Google.” One owner-controlled text file, among the most frequently archived on the web, and the classic exhibit for what the owner told crawlers on a date.

XML sitemap

“A sitemap is a file where you provide information about the pages, videos, and other files on your site,” and submitting one “doesn’t guarantee that all the items in your sitemap will be crawled and indexed.” As evidence, a party-authored summary of its own URL inventory on a date.

hreflang

“If you have multiple versions of a page for different languages or regions, tell Google about these different variations.” Errors here are a common confound in international traffic-loss claims.

Structured data

“A standardized format for providing information about a page and classifying the page content”; Google does not say it guarantees rich results. It is how review ratings and prices are asserted to Google, which puts it at the center of false-advertising and review-manipulation matters.

Site migration

Google’s page on site moves recommends “a URL mapping” and “HTTP permanent redirects if possible,” and says to “Expect temporary fluctuation in site ranking during the move.” A migration is the paradigm concurrent event in a traffic-loss claim and the paradigm case of evidence that cannot be re-examined once the old site is gone; the URL mapping, if prepared, is a discoverable record, and its absence is itself evidence.

What third parties estimate

Backlinks and links

“Google uses links as a signal when determining the relevancy of pages and to find new pages to crawl.” A link lives on someone else’s page; Google’s complete link graph is available to no one, Search Console’s Links report is a partial view, and third-party link indexes are estimates from each vendor’s crawler, sufficiency targets rather than a party’s record.

Anchor text

“Anchor text (also known as link text) is the visible text of a link,” set by whoever controls the linking page. In trademark and defamation matters it is often the statement at issue; in link-spam matters, changes after a demand are the Rule 37(e)(2) pattern.

Disavow file

Google’s disavow tool is “an advanced feature” for links that “have caused a manual action, or likely will cause a manual action.” It is a dated, party-authored file: an admission under Rule 801(d)(2) of which links the party regarded as harmful and when, and simultaneously remediation and alteration of the link evidence if uploaded after a demand.

Keyword

Google does not formally define the term. In keyword-advertising cases it is the bid term in the advertiser’s Google Ads account, a party record; in organic cases it is the Search Console query, the search engine’s record of what users typed.

Search volume estimate

Keyword Planner’s figure is “The average number of times people have searched for a keyword and its close variants,” averaged “over a 12-month period,” and “Your search volume statistics are rounded.” Third-party volumes are models built on clickstream panels, a step further removed. It is the usual input to a yardstick damages model, so the error band has to be stated.

Rank tracker

Google documents no such product. A rank tracker is a third-party service that issues automated queries from chosen locations and devices and records where a URL appears. Its output is a sampled, location-specific third-party record, retained only while the subscription lasts, and the single most common sufficiency target in a search case; average position is the first-party alternative.

Zero-click search

Undefined by Google: a search that ends without a click to any destination outside Google. The published percentages are third-party panel estimates, not Google figures.

What the archive holds

Wayback Machine capture

The Internet Archive explains that an archived URL embeds a fourteen-digit timestamp that “translates as yyyymmddhhmmss,” that “Not every date for every site archived is 100% complete,” and that for legal use “Our affidavit request procedure can be found here.” A third-party record, neither Google’s nor a party’s; the timestamp is per resource, so a page’s HTML and its images may carry different capture times.

Google cache

Google’s current search-operator help documents site:, before:, after:, and filetype: and does not mention a cache: operator. Trade press reported in 2024 that the cache link was removed; confirm the dates from the search engine’s changelog. Either way, the cached copy no longer exists as a source for what a page said on a date; archive or forensic captures have to substitute.

What the results page and AI search add

SERP

A search engine results page is Google’s output at one moment for one user, device, location, and personalization state, “a set of different types of search result visual elements” whose look “can change over time” and varies between desktop and phone. it is not stored for litigants; a screenshot of one is Rule 901(b)(1) evidence of what that witness saw, nothing more.

AI Overview and AI Mode

“AI Overviews help people get to the gist of a complicated topic or question more quickly,” and “AI Mode is particularly helpful for queries where further exploration, reasoning, or complex comparisons are needed.” Sites appearing in these features are “reported on in the Performance report, within the ‘Web’ search type,” with no separate filter, so an AI-Overview-attributed traffic loss cannot be isolated from Google’s own data.

The entries reduce to one habit: before accepting any number in a search dispute, ask who generated it, who can delete it, and whether it is an observation or a model. That answer decides the rule.

Frequently Asked Questions

Published 2026-11-04