Why your website is not ranking: a diagnostic checklist

Work through this in order. The checks at the top take seconds and rule out the most common causes; the ones at the bottom take longer and are less often the culprit. Stop at the first one that fails — fixing a lower-order problem while a higher-order one is still live tells you nothing.

You need two things open: your site in a browser, and Google Search Console for the property. If you have never verified the property in Search Console, do that first. Almost every check below is faster and more reliable there than anywhere else, and it is free.

Step 1 — Is the page in the index at all?

Search Google for site:yourdomain.com/the-exact-page-path. Then, in Search Console, paste the full URL into the URL Inspection tool.

If the page is not there. Nothing else on this list matters yet. A page that is not indexed cannot rank for anything. URL Inspection will tell you why in plain language: "Discovered – currently not indexed", "Crawled – currently not indexed", "Excluded by 'noindex' tag", "Blocked by robots.txt", "Duplicate, Google chose a different canonical". Each of those points at a different fix, and the tool names the one you have.

"Crawled – currently not indexed" is the most common and the most misread. Google fetched the page, looked at it, and chose not to keep it. That is a quality and duplication judgement, not a bug. Adding more internal links or resubmitting will not usually change it. Look at whether the page says anything that is not already said, better, elsewhere — including elsewhere on your own site.

If the page is there, move on — but check, rather than assuming.

Step 2 — Is something on the site telling Google to stay out?

Three files and tags cause the bulk of accidental invisibility. Check all three even if the page is indexed, because a partial block can suppress a page without removing it.

Open yourdomain.com/robots.txt in a browser. A line reading Disallow: / under User-agent: * blocks the whole site. This happens most often when a staging site is copied to production with its settings intact.

View the page source (Ctrl+U) and search for noindex. In WordPress, check Settings → Reading for the "Discourage search engines" box, which writes exactly that.

Search the source for canonical. If the canonical tag on your page points at a different URL, you have told Google that the other URL is the real one. That is correct behaviour for duplicates and a disaster when it is set site-wide to the homepage, which some themes and plugins do by default.

What to do. Fix, then request indexing through URL Inspection. Recovery is usually days, not months, for a site Google already knows.

Step 3 — Does the page target a query real people type?

This one is quiet, because nothing looks broken. The page is indexed, the site is fine, and the phrase you are checking simply has almost no searches — or has plenty of searches but in wording nobody would guess.

Type your target phrase into Google and look at what autocomplete offers instead. Look at the "People also ask" box and the related searches at the foot of the results. Those are drawn from real query data and they are free. If your phrase produces no autocomplete at all and the results page is a thin mix of unrelated things, that is a strong hint the phrase is not one people use.

Then check Search Console → Performance → Queries and sort by impressions. This is what you are actually being shown for. Frequently it is not the phrase you optimised for.

What to do. Rewrite the page around the wording that appears in your impressions data, or pick a different target. There is a fuller method in keyword research without a paid subscription, including how to read autocomplete and related searches as a data source rather than a curiosity.

Step 4 — Does the query want what your page is?

Search your target phrase and read the top ten results as a set. Ask one question: what *type* of thing are they?

If they are all product category pages and you have written a 2,000-word guide, you will not rank, however good the guide is. If they are all guides and you have a product page, the same applies in reverse. Google is showing what people click on. A page of a different type is competing in the wrong event.

An illustrative case. A two-person accountancy practice writes a long explainer titled "self assessment deadlines". The results for that phrase are dominated by official government pages and calculator tools, because people searching it want a date and a form, not an essay. The page is fine. The target is wrong. The same firm's page on "what happens if you miss the self assessment deadline" — a question with genuine explanatory intent — has a real chance.

What to do. Match the format of what is already winning, or pick a query whose winning format is the one you can produce. This is the single highest-value edit on most underperforming pages, and it costs nothing but a rewrite.

Step 5 — It is ranking, and you are looking at the wrong thing

Before you spend money on this problem, rule out the possibility that there is no problem. This is where a surprising share of "my site does not rank" reports end.

You are logged in. Google personalises results using your history. A site you built, visit daily and click on constantly will sit higher for you than for anyone else. Checking your own rankings while signed in is close to useless.

You are somewhere else. Local results change by physical location, not just by the town you type. A plumber in Leeds checking "emergency plumber" from an office in Manchester sees the Manchester set. Localisation is not limited to obviously local queries either — many commercial searches are quietly geo-adjusted.

How to check properly. Use a private window, signed out. Better, ignore manual checking altogether and use Search Console → Performance, filtered to the query and to the country you care about. Average position there is a real aggregate across real users, not one sample from one chair.

You are confusing impressions with position. A page can gain thousands of impressions while sitting at position 40, which is invisible in practice. Rising impressions with a flat position means Google is showing you to more queries, not ranking you better. Rising position with flat clicks usually means your title and description are not earning the click — a separate and much cheaper problem to fix.

Average position is an average. A page at "position 12" may be at 6 for one phrase and 30 for another. Always filter to a single query before drawing a conclusion. If the mechanics of how any of this is decided are hazy, how Google ranking works, explained plainly covers the parts that matter for diagnosis.

Step 6 — Is the competition simply far stronger?

Now look at who holds the positions you want, with the honest question: could this page plausibly be beaten by us, this year?

Open the top five results and note, for each: who published it, how old the site appears to be, whether it is a national brand or a directory, and whether the page looks maintained or abandoned. You do not need a paid tool for a first pass. A results page owned entirely by national publishers, government sites and large marketplaces is telling you something.

| What you see in the top ten | What it usually means | | --- | --- | | Mostly national brands and directories | Very hard; consider a narrower query | | A mix, including small sites and old pages | Genuinely winnable | | Several thin or outdated pages | Winnable quickly with something better | | Almost all forum and community threads | Google reads this as a discussion query | | Heavy ads plus shopping results above the fold | Organic click share is small even at #1 |

That last row matters and gets ignored. Ranking first on a query whose entire visible area is paid units and shopping tiles is worth much less than ranking first on a clean informational page. Check what the results page actually looks like before deciding a position is worth chasing. A structured way to run this comparison is in competitor analysis for search.

What to do. Move down the ladder. Narrower, longer, more specific queries are less contested and convert better. Winning a small query today beats losing a large one for two years.

Step 7 — Is the site simply too new?

New sites take time. Google has no history with the domain, few or no links point to it, and there is no behavioural evidence about whether people find it useful. None of that can be argued with — it can only be waited out or worked around.

If your site is a few months old, ranking for anything competitive is unlikely regardless of what you do. That is not a fault to be diagnosed. How to rank a brand new website: a realistic timeline sets out what the early months usually look like, and how long SEO takes, and what the evidence supports covers why the honest answer is a range rather than a number.

The practical test: are you ranking for your own brand name and for very long, very specific phrases? If yes, the site is working and the constraint is authority and time. If you cannot rank for your own business name, go back to Step 1 — something is broken.

Step 8 — Is something technically broken in a way you cannot see?

Only now is it worth checking technicals, because technical faults are a less common cause of "not ranking" than the four above and take longer to investigate.

Run the page through Search Console's URL Inspection "Test live URL" and view the rendered HTML. If your main content is missing from the rendered output, JavaScript rendering is failing and Google is seeing a near-empty page. Check the page on a phone, on mobile data. Check that the mobile version contains the same content as the desktop version — content hidden or dropped on mobile is content Google may not count, since it indexes the mobile version.

Check Search Console → Page Indexing for a spike in errors, and → Manual Actions. Manual actions are rare, but if you have one, everything else here is irrelevant until it is lifted. If you have bought links, read Google's spam policies on link schemes before doing anything else; the policies describe link buying as a violation, and a site carrying that risk needs it resolved, not decorated. Our own honest treatment is in do you need to buy backlinks, and what makes a backlink high quality covers how to judge the ones you already have. Deeper checks live in the technical SEO guide.

Step 9 — Is the content thin?

Thin does not mean short. Google's documentation is explicit that there is no preferred word count. Thin means the page adds nothing that is not already available, usually because it was written to fill a slot rather than to answer something.

Read your page against the top three results and ask what a reader gets from yours that they cannot get from those. If the honest answer is "nothing", that is your diagnosis, and no amount of technical work will change it. Rewriting around a real question, a real process or real first-hand detail is the fix. Content strategy for small sites covers how to choose what is worth writing at all.

When nothing is broken

Sometimes you reach the end of the list and everything passes. The page is indexed, nothing blocks it, the query is real, the intent matches, the technicals are clean, the content is genuinely useful — and it still sits on page four behind sites that have been publishing for a decade.

That is a real outcome and most articles will not tell you so, because most of them are sold by people whose product is more SEO work. The honest position is that organic search is a slow, compounding channel where age and accumulated links carry enormous weight, and a good page on a young site can be correct in every respect and still lose for a long time. Your options are patience, a narrower target, or a different channel entirely.

If the reason you want rankings is customers rather than rankings themselves, it is worth comparing what visibility costs across channels before committing another year: paid versus organic, an honest cost comparison lays out the trade-offs, and our public rankings page shows what every position on this marketplace costs and who holds it, since our own model is paid placement rather than search. That model suits some businesses and not others — what pay-to-rank cannot do is the honest limit case, and if your problem is Step 1 or Step 2, no paid placement anywhere fixes it. Fix the site first.

Quick answers

I fixed the problem. How long until it changes? For indexing and blocking faults, days. For content and intent changes, weeks. For authority and age, months at minimum.

Should I resubmit the sitemap? It does no harm and rarely helps. Sitemaps assist discovery, not ranking.

My rankings dropped suddenly. Check Manual Actions first, then whether the drop coincides with a Google update, then whether anything changed on your site that week. Sudden and site-wide points at a site-wide cause.

Search Console shows nothing at all. Either the property is misconfigured, or the site genuinely has no impressions. Verify the property covers the right protocol and subdomain — https://www. and https:// are different properties.