How Google decides which page ranks — and why a good website can still be invisible

We built the machine below to show that process as something you can actually see. It is one bench, six stations, and one page travelling along it. Hover over it, use the arrow keys, or press a station below to send the page there. Watch the readout in the bottom corner: it tells you what is happening at that moment.

Tap a station above to send the page there. The figure follows, and the readout in its bottom-right corner tells you what is happening at that moment.

Why this matters more than any single tactic

Most SEO advice is a list of tactics with no order to it. The order is the point. If Google cannot fetch your page, no amount of keyword research will help. If it fetches the page and finds nothing to read because the copy is inside an image, the content you paid for does not exist as far as search is concerned. If it files four copies of the same page under four addresses, each copy is competing with the others.

The stations also explain something business owners find baffling: a site can rank for one search and be invisible for another that means the same thing. Nothing is broken. The page was filed once, scored against its rivals for each query, and placed. Different query, different competition, different position.

Where we usually find the problem

Across the sites we audit, the same five faults come up again and again. None of them are exotic:

• The door is shut. A leftover noindex tag or a robots.txt rule from the staging build, still blocking the live site months after launch.
• Nobody links in. A brand-new domain with no inbound links and no internal links from its own pages, so the crawler has no reason to come back.
• The page is empty to a crawler. Copy set as images, or a page that only fills in after a script that fails on Google's renderer.
• Duplicates everywhere. The same page at www and non-www, with and without a trailing slash, plus a canonical tag pointing somewhere else entirely.
• Written for the business, not the search. A page that says who you are rather than answering the question the customer typed.

What actually moves the needle

In practice, for a Leeds small business, the work looks like this: make sure Google can find and fetch every page you want ranked, keep one clean address per page, put the answer in plain text at the top of the page, and then earn the links and reviews that make the page worth ranking. That is unglamorous and it is most of the job. You can read more about how we approach it on our SEO page, and if you are starting from scratch, our guide to why websites stop getting enquiries covers the commercial side.

An honest note about the machine

The five sliders on the ranking console are a simplified stand-in with equal weights, and the field of rival scores is illustrative. Google describes ranking as using hundreds of signals and has never published their weights. We would rather show you a model that is clearly a model than pretend to a formula nobody outside Google has. What is accurate is the order of the stations and the rule the machine enforces: a page that is never fetched can never rank.

If you would like to know which station your own site is stuck at, that is exactly what we check first.

Common questions

Why is my website not showing on Google?

Usually one of three things: Google has not found it, is not allowed to fetch it, or has decided it is a duplicate of another page. Start with a Search Console URL inspection - it tells you which of the three you are dealing with.

How long does it take to rank?

Indexing a new page can take days to weeks. Moving up the results for a competitive search takes months of consistent work. Anyone who gives you a fixed date for a ranking is guessing.

Do I have to pay Google to appear in the results?

No. The unpaid results are decided by the process in this article. Google Ads is a separate, paid placement that stops the moment you stop paying.