What actually moves rankings
There is no official table of 200 weighted factors. Google ranks a mix of relevance, honesty, and technical fitness — and punishes bought visibility.
Owners are sold a “secret list of 200 factors,” a month of bot traffic, or a pack of links. There is no weighted table. There are groups of signals, and a direct ban on tricking the index.
The working question is simple: does the page do the job the person typed. Google does not publish a closed formula (“content 40%, links 30%”). Neither does anyone else who is telling the truth.
Search looks at a bundle: match to the query, usefulness and originality, a business that looks honest, technical fitness. Trying to buy visibility — fake visits, SEO links, hundreds of empty URLs — earns a restriction, not a climb.
Where “200 factors” came from
The two-hundred story is more than fifteen years old. On 11 November 2010 Danny Sullivan unpacked how Google and Bing competed on the number. At Pubcon, Matt Cutts said Google had more than 200 signals, many with more than 50 variations. Sullivan’s point was the opposite of a sales deck: this is not a contest for who has more knobs, and more signals do not mean a better engine.
On 23 September 2016, when Penguin became part of the core and started running in something close to real time, Google again wrote “more than 200 unique signals.” Examples in that era: words on the page, freshness, locale, PageRank.
Today, How Search Works says relevance is decided by hundreds of factors — including location, language, device. There is still no weights table. Google does not take payment for a higher slot, and it does not guarantee that a URL will be shown at all.
Search Console will show queries, coverage, and experience. It will not hand you a score you can grind. Inside Google, rater guidelines and live experiments decide whether the engine got better. That is not your site’s grade, and it is not something a bot farm can raise.
Content: the person’s job, not a keyword sheet
Text did not go away. What went away is the hope that phrase density and watery length would carry a URL.
Google’s helpful content guidance asks for material for people, not for manipulating results. E-E-A-T (experience, expertise, authoritativeness, trust) is not a separate ranking factor. It is a way to look at a mix of signals; trust sits at the centre. For money, health, and safety, that mix is heavier.
The same document: Google has no favourite word count. “No, we don’t.” Writing “to 8,000 because SEO” is folklore.
Core updates keep retuning what “quality” looks like in the wild. The March 2024 core update and the scaled-content rules that landed with it are the current shape of an old idea: pages that exist to occupy a slot, with nothing a person needed, are not the job. Original work and a service for a human are. Copied text, unedited machine translation, keyword sheets, cloaking, and visit impersonation are how you get cut.
AI: they look at usefulness, not “model or person”
Two mirror myths: “AI will rank us by itself” and “AI is an automatic ban.”
Google, 8 February 2023: quality, not method of creation. Automation to game results is spam. Not every automatic string: match scores and weather reports still exist. AI is not a ranking bonus.
5 March 2024 — scaled content abuse: many pages for manipulation, no help — it does not matter whether a person, a model, or both typed them.
A model that produces a checked, expert-reviewed page that solves the query is not a negative. A factory of empty URLs on throwaway domains is.
Technique: a census, not a ticket to position one
If the site fails on a phone and falls over on HTTPS, a conversation about “factors” is early. Green speed scores also do not purchase the top.
Google has no single Page Experience signal. Ranking includes Core Web Vitals: LCP at or under 2.5 s, INP under 200 ms (from 12 March 2024, replacing FID), CLS under 0.1 — plus HTTPS and a usable mobile layout. Good CWV do not guarantee first place. Relevance is first. An “uncomfortable” page can still win if it answers better.
If the question is whether a person can even send a request, that is a usability problem, not a CWV checklist.
Does traffic raise organic rank
Short answer: inflating a counter — no. Buying bots that search your name and click through is a spam problem.
Installing GA4 or Plausible does not raise position. Those are your instruments, not a signal you submit to the robot. Paid traffic does not, by itself, lift organic.
Traffic spam and fake engagement sit under Google’s spam policies. Clicker programs, “type the query and land,” selling that as a service — restrictions can cover the whole host or a section. They lift when the abuse stops and the systems believe it has not resumed.
The engine cares whether the person finished the job for the query. Not how many sessions your property recorded.
Links, the URL, and an honest firm
Links are used. Bought SEO links are punished: Penguin has been in the core, near real time, since 2016. Google’s spam policies still treat paid links that pass PageRank as a problem. The folklore of “they switched links off, then on” is a different article, and it is not this one.
A keyword in the domain or the path barely moves the needle. Google’s SEO Starter Guide: effect is minimal, aside from breadcrumb display in the snippet. Meta keywords are not read. Stuffing the slug is useless and close to spam.
There is no public “quality index” with a formula you can farm. Search Console’s experience and trust-adjacent views are orientation, not a KPI to grow as the strategy.
For businesses that take money: a firm that takes full prepayment and never shows up is a trust problem, not a “put the company number in the footer and rank” checklist. A one-off delay is not the same as a pattern. There is no official “forty review cards and a VAT number = top.”
Which “factors” are folklore
| You are sold | What the documents say |
|---|---|
| Official list of 200 weighted factors | A 2010 talking number. Now: “hundreds.” No table. |
| Traffic / installing analytics raises the site | Fake search engagement is spam. Raw traffic is not stated as a plus. |
| A hard word-count target | Google: no favourite length. |
| Keyword in the domain = top | Almost no effect, except crumbs in the snippet. |
| Install GA4 and you will climb | Not stated. |
| Buy 100 links a month | Paid SEO links are punished. |
| Any AI text = ban | They look at usefulness. The ban is mass empty spam. |
| A fast site guarantees position one | CWV participate. Relevance is first. Top is not promised. |
If there is a site and no requests
Pages for a person’s job, not a keyword cloud. Your text, not copies and not hundreds of shells. HTTPS, a usable mobile layout, a country or region that matches the offer, contacts, honest terms. Do not buy bots or link packs.
Pointless: a “200 factors” checklist, a two-week #1 guarantee, “grow the quality score,” a keyword domain as strategy.
If demand exists and requests do not, split three causes first: the SERP, an interface that cannot finish the job, or the offer itself. That split does not start with links.
The studio has built sites and search work since 2006. A ranking conversation is a written brief on named queries, not a promise of positions.
- Does each money page answer a named job for a person, or occupy space for a keyword cloud?
- Are Core Web Vitals treated as a hygiene gate, not as a promise of position one?
- Are paid links, bot traffic, and scaled empty pages off the plan — as policy, not as a grey tactic?
- If requests are missing, is the first split rank vs interface vs offer — not a link package?