Google ranks pages by combining many signals, not a fixed scorecard. The ones Google actually documents are relevance to the query, content quality and helpfulness, links and other signs of reputation, usability signals like mobile-friendliness and HTTPS, and context such as the searcher's location, language, and device. Everything else is inference.
That last sentence is where most "ranking factors" lists go wrong. There is no published weighting, no confirmed count, and no dial you can turn to buy a position. What you can do is separate the signals Google has described from the ones the industry infers, and from outright folklore — then work the short list you control and measure whether your positions respond.
What are Google's ranking factors, really?
Google describes ranking as a set of automated systems that assess the query and the pages that might answer it. The public framing (Search Essentials, the "How Search works" documentation, and Google's own posts about core updates) consistently points at a handful of themes:
- Meaning of the query — what the searcher wants, including synonyms and whether the intent is informational, commercial, local, or fresh news.
- Relevance of the content — whether the page contains the answer, in the words and concepts the query implies.
- Quality and helpfulness — whether the content reads as written for people, with evident experience and trustworthiness, rather than assembled to fill a slot in a SERP (search engine results page).
- Usability — mobile-friendliness, HTTPS, and page experience aspects Google documents as things it considers, without ever claiming they outrank relevance.
- Context and settings — location, language, device, and search history, which is why two people searching the same phrase see different results.
Notice what isn't on that list: a magic word count, a keyword density figure, a specific bounce rate, or a domain-authority score. Those are third-party metrics or SEO shorthand, not signals Google says it uses.
Which ranking factors can you actually control?
Sort them by control, because that decides where your hours go. The "confidence" column reflects how well-supported each item is, not a weighting.
| Signal | Your control | Confidence it matters | Practical move |
|---|---|---|---|
| Query–content relevance | High | Documented | Cover the actual question in the actual language searchers use |
| Intent match (page type vs SERP) | High | Widely observed | Match the format the current top 10 rewards — guide, product, list, tool |
| Title and headings | High | Documented as used | Make the title answer the query precisely; no clickbait |
| Content depth and accuracy | High | Documented (helpfulness) | Close gaps, add examples, remove filler |
| Internal links | High | Documented as crawl/context signal | Point relevant, strong pages at the target with descriptive anchors |
| Mobile-friendliness, HTTPS | High | Documented | Fix once; treat as hygiene, not leverage |
| Core Web Vitals / page speed | Medium | Documented, modest | Fix real problems; don't chase a perfect score |
| Structured data | Medium | Rich-result eligibility, not a rank boost | Use it to change how you appear, not where |
| External links to your page | Low | Documented as a signal | Earn them with things worth citing; never buy them |
| Competitor behavior | None | Real | Track it, react to it |
| Personalization, location, device | None | Documented | Track from a fixed location/device so you see one consistent reality |
The pattern is the point: nearly everything with high control lives on your page — the same list that decides whether a page-two ranking crosses onto page one, which is why our on-page improvements guide is the practical companion to this explainer.
Do ranking factors still matter if Google changes constantly?
Yes — but think in terms of what the systems are trying to reward rather than a checklist to satisfy. Core updates change how quality and relevance get assessed, not the fact that they are being assessed. A page that genuinely answers the query better than the pages around it has been the winning bet across every documented update; a page that satisfies a checklist without satisfying a reader has not.
This is also why rankings move on days you changed nothing: re-evaluation, competitor edits, and SERP-layout changes all shift positions. That's normal volatility, not a factor switching off — and telling noise from real movement is a prerequisite for judging any change you make.
How do I check the ranking factors on my own page?
Run this on one page at a time — ideally a page already sitting at #11–20, where the gap is small enough for a single round of work to plausibly show up in your rank data.
- Search the target query from your target location and device. Note what kind of pages hold the top 10. If they're all comparison tables and yours is an essay, stop here — intent is your factor.
- Read your page against the top three. List every subtopic they cover that you don't, and every question their headings answer that yours doesn't.
- Audit the title and H1. Do they carry the query's actual meaning, or a clever phrase? Precision beats personality on the results page.
- Check technical hygiene. HTTPS, mobile rendering, no accidental noindex, no layout chaos on load. Pass/fail, not competitive advantage.
- Count internal links pointing at the page. If it's one or zero, that's the cheapest lever you own — add links from your most relevant pages with descriptive anchor text.
- Confirm indexing and the ranking URL. In Search Console, check the page is indexed and see which URL Google actually ranks for the query.
- Record the date and the change. Without a log, you can't attribute movement later.
- Wait, then read the trend across several checks over weeks, not the next morning's position.
Step 8 is where most processes die. Inconsistent checks — different locations, devices, days — produce numbers that argue with each other, and improvements stay invisible. Fixed-parameter tracking is what makes a "before and after" mean anything; the discipline is in our rank tracking guide.
Which "ranking factors" should you ignore?
Some persistent claims deserve a quiet retirement:
- Keyword density targets. Relevance is coverage and meaning, not a percentage.
- Domain authority as a Google signal. A useful third-party estimate for comparing sites; Google doesn't use it.
- Publishing frequency for its own sake. Cadence isn't a signal; helpfulness is. Weekly thin posts are a cost, not a strategy.
- Exact word-count minimums. Length is a side effect of completeness.
- Anything requiring bought links or pages generated at scale. Documented spam-policy violations, and the downside is your whole site.
The time you don't spend on folklore is time available for the page sitting at #12.
FAQ
How many Google ranking factors are there?
Nobody outside Google knows, and the commonly repeated numbers are industry lore rather than confirmed counts. Google describes ranking as multiple interacting systems weighing relevance, quality, usability, and context — not a fixed list with published weights. Treat any definitive numbered list of factors as an opinion piece, not documentation.
What is the most important Google ranking factor?
There isn't a single one, and the honest answer is that importance shifts by query. For most site owners the highest-return work is relevance and intent match — making sure the page genuinely answers the query in the format that query rewards — because no amount of technical polish compensates for answering a different question than the one being asked.
Are backlinks still a ranking factor?
Links remain one of the signals Google describes as helping it assess pages, and they still matter competitively. What has changed is tolerance for manipulation: paid links and link schemes are explicit spam-policy violations. Earn citations by publishing something worth citing, and treat link building as slow, low-control work next to the on-page moves you can make this afternoon.
Does page speed affect rankings?
Google documents page experience aspects including Core Web Vitals as things it considers, but frames them as modest next to relevance and helpfulness. Fix genuine problems — a page that's painful to use loses readers regardless of position — but don't chase a perfect score on a page whose real problem is that it doesn't answer the query.
How do I know if a change improved my rankings?
Log what you changed and when, then compare positions across several checks taken with the same location, device, and language settings over the following weeks. A single check proves nothing; a consistent trend is evidence. Crossing from page two into the top 10 is the threshold worth watching, since that's where clicks change — see striking-distance keywords for how to prioritize which pages to test first.
Work the factors you control, then measure
The useful version of "ranking factors" isn't a list to memorize — it's a filter. Documented, controllable signals get your hours. Inferred signals get attention once those are handled. Folklore gets ignored. Then you check whether your positions responded, because on your site the only factors that count are the ones that move your keywords.
Add Me In Top 10 is being built to close that loop: scheduled checks from a fixed location and device, movement since your last check, and a standing report of the keywords at #11–20 where this work pays back fastest. Track your rankings with Add Me In Top 10 and see which changes actually moved you.
