My website does not appear on Google: the checks, in order

«My website does not appear on Google» describes two different faults, and it is the confusion between them that wastes money.

The first is that the site is not in the index. Google does not hold those pages, so it cannot show them to anyone, not even to someone searching the exact name of the business. Often, a technical fix of a few minutes settles it.

The second is that the site is in the index and is not chosen for the searches that matter. That one has no button, and anyone who promises you a position is promising something Google itself does not guarantee.

Two questions come before everything else: is it the website that is missing, or the business listing? And is the site in the index at all? After those, the eight checks are numbered in order, and the order is the useful part: each one has to be resolved for the next to make sense. They are all free, none needs a supplier, and several can be done from a phone.

The Maps listing is not fixed on the website

These are two different systems. The pages of the site come from the search index. The side panel with the address, opening hours, phone number and reviews, and the pin on the map, come from the Google Business Profile, which is created and verified separately.

A clinic that corrected its opening hours on the website and still sees the old hours on Maps does not have a website problem. If what is missing is the listing, no change to the site will make it appear.

Google’s Portuguese documentation is clear on two points that save weeks of worry: only verified businesses show their details on Maps, and «Ratings for new businesses can take up to a month to appear in search results» (Google Business Profile help, in Portuguese).

If it really is the website that is not showing, read on.

The search that separates the two faults

Search for site: followed by the address, with no space anywhere in between: site:aminhaempresa.pt.

If nothing comes back, the problem is indexing and the next seven sections are for you. If pages come back, the site is in the index and you can skip straight to «Indexed and still not showing: the second fault».

One caveat almost nobody makes, and it comes from Google itself: this operator does not necessarily return every indexed address of a site (Google Search Central, «Get your website on Google»). It is good for a first reading. It is not good for concluding that a specific page is missing from the index.

1 and 2. Only Search Console answers without guesswork

Search Console is Google’s free tool for the owner of a website. Its URL inspection tool tells you, address by address, whether that page is or is not on Google. It is the only answer that is not deduction.

Check 1: the right property. There is a trap in the setup, and it catches a lot of people on day one. A domain property covers every subdomain and several protocols (http, https, ftp); a URL-prefix property covers only the addresses that start exactly with that prefix, protocol included (Search Console property types). Anyone who verifies http:// and visits https:// is reading an empty report about a place where nobody lives.

Check 2: two addresses, not one. Inspect the home page and one inner page. A site whose home page is in the index while its inner pages are not is a different case from a site that is not there at all, and the inspection tells you which of the two you have in front of you.

Google’s second sentence about this tool is worth keeping as well: being on Google does not guarantee that the page will appear in results (URL inspection). Keep it, because it is the sentence that splits this article into two halves.

A pause: a two-week-old site is not broken

Google says it can take a few weeks to notice a new site, or any change to an existing one (Google Search Central, «Get your website on Google»), and that crawling a submitted address takes anywhere from a few days to a few weeks.

Two warnings in the same documentation head off the next mistake: asking for indexing again, and again, does not make Google crawl any faster, and asking for indexing does not guarantee that the page will end up appearing in results (asking for a recrawl).

If the site is two weeks old, there is nothing wrong with it. It is two weeks old. This is the exact moment when the calls selling urgency arrive, and it is the worst possible moment to hire anyone for anything.

3. A forgotten noindex removes the site from results

On a newly built or newly migrated site, this is the first thing to check, and it costs nothing to fix. A noindex instruction tells Google to drop that page from results, even if other sites link to it (the noindex documentation). It sits there during the build, so that the site under construction is not found, and on handover day somebody has to take it out.

Checking takes half a minute. Open the page, view the source (Ctrl+U in most browsers) and look for the word noindex. In WordPress there is also a box in the reading settings that asks search engines not to index the site: if it is ticked, that is your answer.

There is a variant that does not show up in the source, because it travels in the server’s response header. That one is only visible in Search Console’s URL inspection, which shows the site owner exactly what Google received.

4. A closed robots.txt hides the noindex inside it

Type the site address followed by /robots.txt. If Disallow: / is sitting under User-agent: *, you are asking crawlers not to go anywhere. It is another leftover from the build.

What this file does is manage crawler access, and Google warns that it is not a mechanism for keeping a page out of search: a blocked page can still appear if other sites link to it, only with nothing useful to show (introduction to robots.txt).

And here is the reason these two checks come in this order: if the page is blocked in robots.txt, Google never gets to read the noindex inside it. Blocking and forbidding at the same time is not doubly safe. It is contradictory, and the result is unpredictable.

5. To Google, four addresses are not one site

To a person, aminhaempresa.pt and www.aminhaempresa.pt are the same thing, and so are http and https. To a machine they are four addresses. What should exist is one main address and redirects from all the others to it.

When that is not done, or is done the wrong way round, quiet things happen. Google picks for itself which version it treats as the main one when nobody tells it, and the canonical tag you place on your pages is a strong signal, not an order (canonicalisation). A badly configured theme that points the canonical of every page at the home page can make the inner pages disappear without anyone deleting anything.

Then there is the temporary address: the site lived on the builder’s test domain, that domain is still up, and it is that copy Google knows. Search Console has a report that sorts exactly these cases, separating pages with a redirect from duplicates where Google chose a version other than the one you declared (page indexing report).

6. The site can open for you and fail for everyone else

Test from somewhere else: a private window, mobile data instead of the office network, and preferably on a device that has never opened the site. Many «disappearances» are the site failing for everyone except the owner, who has the address in memory and a session open.

The list of states that prevent indexing is short and well known, and the page indexing report names them one by one: server error, page not found, access forbidden, unauthorised access, redirect error.

Behind them there is usually one of these causes:

• hosting left unpaid;
• a domain left unrenewed, because the renewal notice went to a mailbox nobody opens;
• a «site under construction» plugin still active;
• a forgotten password on a test copy;
• a firewall blocking the crawler.

None of them is a content problem, and none of them is solved by writing more pages.

7. If the text only arrives with JavaScript, there is a blind window

Google handles pages in three phases: crawling, rendering and indexing. Rendering, the phase where JavaScript runs, can sit waiting in a queue (JavaScript SEO basics). If the text of the site is only born after that step, there is a window in which Google has the page and does not have the words.

How to see this without being technical: in Search Console’s URL inspection, ask for the code Google got and look in it for a sentence you know by heart from your own page. If the sentence is not there, Google does not have it.

This turns up on sites built in modern builders and on pages where the content only loads after a click or after scrolling: the list of services that only opens when you tap the tab, for instance.

Indexed and still not showing: the second fault

It means you have moved from the first fault to the second, and that the honest answer changes tone. Google writes, in its own essential documentation, that meeting all the requirements and best practices does not mean it will crawl, index or show that content (Google Search Essentials). There is no tag to fix here. There is a page that, for that search, was not judged the best available answer.

There is an in-between state that confuses a lot of people and that the indexing report names: pages crawled and not yet indexed, which Google says may or may not end up indexed, with no point in submitting them again. It is not a punishment. It is the index having no reason to store one more page just like the others.

And this is where the page count usually betrays the site owner. Thirty pages generated to say the same thing in different words do not add up to thirty opportunities, and we have written about why we do not build sites like that.

Check 8, and the word «penalty». There is a formal thing with that name, it is called a manual action, it happens when a human reviewer at Google decides that the pages break the spam policies, and the effect is the site dropping or being left out of results. It is announced in a report of its own in Search Console (manual actions). If that report has no message in it, there is no penalty, whatever the person who called you offering the solution says on the phone.

AI answers changed what you see, not the diagnosis

AI Mode arrived in Portugal, in Portuguese, on 7 October 2025, and appears as a tab on the results page and in the app (Google’s blog in Portuguese). A site that is not in the index is not in these answers either.

The eligibility rule Google published is useful precisely because it is short: to be able to appear in AI features, a page has to be indexed and able to be shown with a snippet. And the same page says the rest: no special markup is needed, no structured data for AI is needed, and files such as llms.txt are not used by Google Search (guide to optimising for AI features).

There is a side of this rule you control from the site, and it is worth knowing it exists: controls such as nosnippet, data-nosnippet or max-snippet limit the information that can be shown from the page in Search (AI features and eligibility requirements). And a page that cannot be shown with a snippet stops meeting the eligibility rule above.

So let it be said plainly: nobody controls whether a model cites you. What you control is the site being reachable, being indexed, and answering the question in readable form. That is a whole article of its own, and we have written one on what Google says you do not need to do.

The list in order, to do yourself and pay nothing

First, the two questions that come before everything else: is it the website that is missing, or the business listing? And what does the search site:aminhaempresa.pt return? If nothing comes back, work through the eight checks in this order:

1. Open Search Console on the domain property, the one that covers everything.
2. Inspect the home page address and one inner page address.
3. View the source and look for noindex.
4. Open /robots.txt and look for Disallow: /.
5. Confirm there is one main address and that the others lead to it.
6. Open the site in a private window, on mobile data, on another device.
7. Confirm that the text of the page exists in the code Google received.
8. Confirm there is no message in the manual actions report.

If at the end of all this the site is indexed and still does not appear for the searches that would bring you clients, the problem has stopped being technical and has become one of content, structure and competition. It is work, it is slow, and it never comes with a promised position. Before hiring anyone who promises one, it is worth knowing what a website quote has to name.

And if you would rather have someone run this list with you, with your site open beside you, book a 15-minute call. Most of these checks are no work at all for anybody, and saying so is more useful than selling an audit.

Every Google page linked above was consulted on 22 August 2026. The only sentence in quotation marks is our translation of the Portuguese-language Business Profile help; the remaining statements are our reading of the pages linked beside each of them.