Hundreds of signals still not a dossier

How Oernoe Search works opens the ranking story with a sentence that is easy to misread. At the core of Oernoe Search is a ranking algorithm, a complex system that analyzes hundreds of signals to determine which pages are most relevant for your search query. A careless reader hears “hundreds of signals” and imagines a personal advertising dossier of the searcher. The rest of the guide refuses that reading. The algorithm treats every search equally. It does not incorporate your search history, location, device, or personal profile as ranking inputs. When you search, it analyzes the query, searches the index for pages that look relevant, evaluates each page using hundreds of ranking factors, ranks results from most to least relevant, and returns results with no personal data attached.

This essay is about that refusal. Hundreds of signals still means page and content quality work, not a dossier of you. It is not a rewrite of ranking-signals-are-about-pages-not-people, though it shares a boundary with that claim. It is not authority-signals-are-not-your-history. It is not same-algorithm-for-every-person restated with different nouns. It is not technical-logs-are-not-an-ad-graph. Those topics have their own homes. Here the subject is the scale of ranking factors itself: why a large signal set does not become a personal advertising file.

## What “hundreds of signals” is allowed to mean

In the guide’s How Ranking Works list, signals are ranking factors applied to pages after the query is understood and the index is searched. The Privacy-Preserving Ranking Factors section names the primary ones the algorithm considers: content relevance, authority and expertise, link analysis, page freshness, aggregated anonymized user satisfaction signals, mobile-friendliness, and page speed. The guide then names what is not on the list: your search history, your location, your device type, your age, your interests, or any other personal data.

That contrast is the whole essay in miniature. A large set of page-quality measurements can still be a large set of page-quality measurements. Scale does not convert a page score into a person score. If it did, the published sentence about treating every search equally would be false.

Aggregated anonymized satisfaction signals deserve a careful note. The guide says bounce rate, click-through rate, and similar measures are processed at scale without tying data to individuals. That is still a page-level quality input, not a named dossier of the person who clicked. Search-and-ads and How Search works both refuse the move that turns click or dwell into an advertising profile of you. Oernoe Search does not use click or dwell signals to build an advertising profile of you. Aggregated page satisfaction is not a back door into personal advertising memory.

## Equal treatment is the second half of the sentence

The key innovation sentence is blunt. The algorithm works the same way for every person. If you and a friend search for the same phrase, you get the same results unless your account has custom preferences you saved. That is fundamentally different from personalized search, where results change based on past behavior, location, and profile.

Hundreds of signals live inside that equal-treatment rule. They are the machinery that scores pages for a query. They are not a per-person interest graph that rearranges the same query into different worlds. Custom preferences you saved—language, region, filters, bookmarks, saved searches—are the documented exception. They are explicit, deletable, and not shared with advertisers. They are not the ranking-factor list by another name.

If “hundreds of signals” quietly included a personal advertising dossier, the equal-treatment claim would collapse. Journal writing on this hostname exists to keep that collapse from being papered over with complexity theater. Complexity about pages is allowed. Complexity about secret people files is not.

## How signals sit next to logs and ads

Search still sees ordinary request data needed to return a page: the query, the fact that a browser asked, technical logs for abuse and uptime. Search-and-ads says that is not the same as an ad graph. The Privacy Policy describes technical and security logs needed to run services. How Search works says those logs are not a license to sell query history to advertisers. Technical logs and ranking signals are different objects. Logs keep the service running and safe. Ranking signals score pages. Neither is supposed to become a personal advertising dossier of the searcher.

Publisher ads are a third object. Selected finished pages on www.oernoe.com may load Google ads. Ads are requested as non-personalized by default. Google may still process page context and advertising measurement data under its own policies. That is disclosed in Search-and-ads and in the Privacy Policy. It is not Search ranking. Mash the three objects together and you get the zero-tracking slogans this company already had to correct. Keep them apart and you can say something checkable: ranking uses hundreds of page signals; Search does not sell queries as an advertising file; the publisher discloses Google ads when they appear.

## What a dossier would look like, and why this is not that

A personal advertising dossier of a searcher would look like a durable file of interests, age bands, and past queries used to change what ads or results you see without an explicit control you can find and delete. How Search works rejects that pattern for ranking. Search queries are not used to build an advertising profile. Search queries are not sold or used to build advertising profiles. Oernoe does not sell account or search data. Preferences you save are never shared with advertisers or used to manipulate results as a hidden profile.

Hundreds of page signals do not become that dossier by counting higher. A page’s freshness score is about the page. A page’s mobile-friendliness is about the page. Link analysis is about how public pages cite each other. Authority and expertise are about the source as a public publisher, not about your private history. Content relevance is about how well page text matches the query you typed right now. None of those become a person file because there are many of them.

## How this claim is checked against public pages

Read How Oernoe Search works from The Search Algorithm through Privacy-Preserving Ranking Factors. Confirm “hundreds of signals” and “hundreds of ranking factors.” Confirm equal treatment without incorporating search history, location, device, or personal profile. Confirm the primary factor list and the explicit not-list of personal data. Confirm that results return with no personal data attached. Confirm that click or dwell is not used to build an advertising profile of you. Confirm that Search does not sell queries or account data to advertisers.

Read Search queries and Google ads for the Search versus publisher-ads split, the refusal to sell account or search data, and the statement that technical request data is not an ad graph. Read the Privacy Policy for the statements that Oernoe does not sell account information or search history and does not use search history to create advertising profiles. Read Using an Oernoe account for the claim that an account saves preferences you chose rather than a hidden click profile. Read About and the homepage for the operator story: Anoepal runs Oernoe, www is the publisher, Search lives on its own hostname, and finished original journal articles are the public record.

If those pages disagree with this essay, trust the pages and fix this draft. Journal articles on this hostname are supposed to add a method or a sharpened reading of published behavior, not invent a softer privacy story.

## Distinct neighboring claims

Ranking signals are about pages not people is the broader framing of signal targets. Authority signals are not your history isolates one factor family. Same algorithm for every person is the equal-treatment rule. Technical logs are not an ad graph is about request logs rather than ranking factors. This essay’s job is the scale sentence itself: hundreds of signals can still be page work, still not a dossier of the searcher.

## Practical reading for operators and reviewers

Ask whether any named ranking factor requires a personal advertising profile of the searcher. Ask whether two unsigned people who type the same query get the same results. Ask whether aggregated satisfaction signals are tied to individuals. Ask whether Search exports query history to advertisers. Ask where Google ads appear on www and whether that appearance is being mashed into ranking-signal talk.

Do not ask us to pretend a large signal set is proof of personalization. Do not ask for invented statistics about how many signals exist beyond what the guide already says. Do not treat a complex ranking stack as evidence of a secret people file when the published factor list is about pages. Check the not-list first.

## Why scale rhetoric needs a boring operational meaning

Scale language dies when it is only impressiveness. In this product story, hundreds of signals means many measurements of pages and content quality applied after a query is understood. It does not mean many measurements of you. That is why the guide pairs the scale sentence with equal treatment, with a page-factor list, and with an explicit refusal to use search history, location, device, age, interests, or other personal data as ranking inputs. Remove any one of those legs and “hundreds of signals” becomes complexity theater.

Operators sometimes want a warmer sentence: we use hundreds of signals about you so Search gets better for you. That warmer sentence is exactly the personalized-search path the guide rejects. Signals about you without an explicit, deletable preference are profiling. Signals about pages are ranking. Keep the colder sentence. It is the one we can defend when a reviewer hears “hundreds” and reaches for a dossier metaphor.

People also confuse ranking signals with account preferences and with publisher ads. Preferences are authored settings you can delete. Ads on selected www pages are a disclosed publisher system. Ranking signals score public pages for a query. Confusing the three makes support harder and makes privacy claims harder to test. Ask which object moved before you accuse the algorithm of building a personal advertising file.

## Closing

Hundreds of signals are still not a dossier when three conditions hold. The signals named for ranking are about pages and content quality. The algorithm treats every search equally unless you saved explicit preferences. Search does not turn those signals, or your queries, into a personal advertising profile sold or shared as a people file. How Oernoe Search works already publishes those conditions in the ranking and privacy-preserving factors sections. This journal note exists to keep the scale sentence from dissolving into a personalized-search apology or a fake-complexity claim about the searcher. If the factor list starts requiring a person file, say so and fix the product. If it remains page work, say that plainly. Scale is allowed. A secret dossier of the searcher is not. That is the whole deal for this URL.