SEO for WordPress: What It Does Alone, What the Plugin Adds, and What Falls to You

You installed WordPress. Then you installed the SEO plugin everyone recommends. You filled in the fields, and the little dot in the editor turned green.

Months later, the site shows up when someone types the company name. For everything else, silence.

The search results push the same answer: more plugins. On September 14, 2026, we pulled the top ten results for “seo para wordpress” (SEO for WordPress). The search ran on Google in Brazil, where we are based. Seven of the ten had the word plugin right in the URL. The first was a plugin’s page in the official WordPress directory. ¹

The question that solves your problem has three parts. What WordPress does on its own. What the plugin adds. And what neither of them does for you. This piece separates the three with each maker’s own documentation. The case I use is the site I know best: ours.

Three columns on one sheet

Think of a brand-new car. It leaves the factory with brakes, headlights, and seat belts. At the accessory shop, you add a roof rack and a dash cam. Neither one picks the destination, or knows whether the road exists.

In WordPress, the factory is the system itself. The accessory is the SEO plugin. The destination, and the map to get there, stay with whoever publishes.

Draw three columns on a sheet of paper. Everything that follows falls into one of them.

First column: what WordPress already does on its own

The title of each page

Open any page on a WordPress site and look at the browser tab. Nobody typed that text in there. The system put it together.

WordPress generates the title tag by itself when the theme declares support for it, a feature available since version 4.1. On inner pages, the rule joins the page title, a hyphen, and the site name. On the homepage, it joins the Site Title and the text in the Tagline field. ² That field lives under Settings > General, and it comes empty out of the box. ³

Google counts on this. Its starter guide says most content management systems turn the title you write into the title element in the HTML. ⁴ The work left for you is writing good titles.

Which copy is the original

Think of a contract with three copies in a drawer. Someone has to stamp one of them as the original, or each reader picks their own. On a web page, that stamp is a line of code pointing to the main address. This is called a canonical URL.

On every post and every page, WordPress writes that line by itself, and has since version 2.9. ⁵ Category, author, and date archives fall outside this rule, because the function only acts on individual content. ⁵

The list of addresses for the crawler

Think of the tenant directory in a building lobby. The delivery driver does not have to knock on every door to learn who lives there. Online, that directory is a file with the site’s addresses, made for the search engine crawler. This is called a sitemap.

WordPress has built this file by itself since version 5.5, released on August 11, 2020. The address is /wp-sitemap.xml. It includes every public content type, the categories, the author archives, and the homepage. The robots.txt that WordPress generates points to it. ⁶

Two rules from the same documentation deserve a note. The sitemap turns off when you check the box that asks search engines not to index the site. And it lists only the address, with no priority or frequency, because search engines do not usually read those fields. ⁶

Google confirms the design. If you use WordPress, Wix, or Blogger, you probably already have a sitemap available, and you do not have to do anything. ⁷ The same page points out two limits. A site with well-linked pages is usually discovered almost entirely without a sitemap. And a sitemap helps discovery without guaranteeing crawling or indexing. ⁷ How often the crawler comes back is a topic for crawling and crawl budget.

The format of your addresses

Imagine your street address printed on your business card. Change it later, and everyone who kept the card has to be told. On a website, each piece of content has a fixed address like that. This is called a permalink.

The WordPress documentation says that address should be permanent and never change. ⁸ You choose the format under Settings > Permalinks. The options range from a numbered address, like ?p=123, to the Post name format. ⁸

Here the official documentation trips over itself. One page says the default uses the day and the name, then calls the numbered format the default in its own list. Another page says the default is the numbered format, the one called Plain. ⁸ The practical rule is to open the screen and see what is selected on your site.

On words in the address, the two sources pull in different directions. One WordPress page calls addresses with words SEO-friendly. ⁸ Google is more restrained. It recommends readable words, in your audience’s language, rather than long ID numbers. ⁹ And it writes that a keyword in the domain or the path, by itself, has hardly any effect beyond appearing in breadcrumbs. ⁴

An address with words helps a person understand the link before clicking. Choose the format before you publish your first page.

The box that hides the site

On the Reading Settings screen, under Search engine visibility, sits a box labeled “Discourage search engines from indexing this site.” Once checked, it adds a request not to index to the code of the entire site. ¹⁰ The box blocks nobody’s access, and it is up to the search engine to honor the request. ¹⁰ That case, with a step-by-step check, is in the piece on when your site isn’t showing up on Google.

What the factory leaves out

The SEO page in the WordPress documentation says meta tags, the page description among them, are not included by default. It points to two paths: add them by hand, or use a plugin from the official directory. ¹¹

This is where the second column begins.

Second column: what the SEO plugin adds

I use two examples, Yoast and Rank Math. I have no affiliation with either one, and I will not tell you which is better. Everything below comes from what the makers themselves write.

A description for each page, with a template

Under the blue title on the results page, two lines of summary appear. The text you suggest for those lines lives in the page’s code. This is called a meta description.

Google uses that suggestion sometimes, when it describes the page better than the text itself. Most of the time, the snippet comes from the page content. ¹² Google’s help asks for a unique description on every page. If you lack the time for all of them, start with the homepage and the most visited pages. ¹²

Both plugins give you one field per page and a template for the whole site. The template combines variables, like the title and the site name, and fills in any field left empty. ^{13 14} For a large site built from a database, like a product aggregator, Google accepts and encourages descriptions generated this way. For a hand-written article, it points out that one more sentence takes minimal effort. ¹²

One sitemap in place of the other

Yoast writes that turning on its sitemap turns off the default WordPress sitemap. If you turn off the Yoast sitemap, the WordPress one stays in place as a backup. ¹⁵ The decision is on record in version 14.5 of the plugin, which disables the sitemap that came with WordPress 5.5. ¹⁵ Rank Math builds its index at /sitemap_index.xml. ¹⁶

In practice, the address of your sitemap changes with the plugin you install. Search Console needs to receive the address that is actually live.

Yoast’s help adds a rule that works for any site. A sitemap at /sitemap.xml was not made by Yoast, and may come from another plugin or from WordPress itself. Before you turn on its sitemap, the maker asks you to disable other sitemap plugins and delete sitemap files from the server. ¹⁷ One sitemap generator per site.

Switches for what WordPress creates on its own

WordPress creates pages nobody wrote. The author archive, the list for each category, the archive for each date. The plugin gives you the switch for those pages.

Rank Math recommends turning off Author Archives on a single-author blog, because they repeat content that already exists. Once they are off, the author address redirects to the homepage. ¹⁴ Yoast leaves the content types set to noindex out of its sitemap. ¹⁷

Structured data

Both advertise structured data markup, the labels that tell the machine what the page and the company are. ¹⁸ In my book SEO de Entidade, published in Portuguese, I wrote that this markup is legitimate and inexpensive. It helps the machine avoid confusing you with something else. The chapter’s verdict fits in one line: mark up for the right reason, and never as a switch for citation. ¹⁹

Third column: what neither one solves

Our company site in Brazil, seogenome.com, ran on WordPress 6.9.7, with Elementor, until the switch to the new site on September 2, 2026. On August 31, we ran a full diagnosis of the site. It covered 22 published pages, the sitemaps, Search Console, and Analytics. ²⁰

The technical profile lists the page builder, languages, caching, analytics, the WhatsApp button, redirects, and Web Stories. No SEO plugin appears on that list. ²⁰ The basics were live. All 22 pages had a canonical URL, and the native sitemap responded at /wp-sitemap.xml. ²⁰

The platform change that came later stays out of this piece. The six findings below apply to any WordPress site, and that is why I share them.

Finding 1: seventeen articles published as pages

The site had zero posts. ²⁰ The WordPress documentation says a page, by default, gets no categories or tags, and stays out of the feed. ²¹ The diagnosis recorded the full consequence. No categories, no tags, no archive, no listing, no feed. ²⁰ WordPress offers both types, and the plugin works with both. The choice of type was ours, made every time we hit publish.

Finding 2: no menu

Each page had three or four internal links, always the same ones: the logo and the language switcher. Across the 17 articles, there was not a single editorial link. ²⁰ Google writes that it finds pages mainly through links from other pages. ⁴ Some plugins suggest internal links, in their paid version. ¹⁸ Deciding which page supports which one still falls to whoever designs the site.

Finding 3: a meta description on 1 of 22 pages

The other 21, the homepages included, had none. For those, Google built the snippet on its own, from the page text. ^{20 12} A plugin template would have filled the empty field. The sentence that makes someone click is still written by a person. To count this on your own site in minutes, run Screaming Frog and open the Meta Description tab.

Finding 4: an English title on a Portuguese homepage

The title served was the site name, a hyphen, and the phrase “Aligned with the future: Semantic SEO, GEO, and contextual intelligence.” ²⁰ The format matches the WordPress rule for the homepage: Site Title plus Tagline. ² The reading here is mine: the English phrase sat in the Tagline field and ended up in the title of the Portuguese page. Google asks for a title in the same language as the page. When they do not match, it may choose a different text to show. ²² WordPress did what it was programmed to do, with the text it found.

Finding 5: the sitemap listed what it shouldn’t

Anyone who typed /sitemap.xml got a success response with an empty file. The robots.txt pointed to no sitemap at all. ²⁰ According to the documentation, the robots.txt WordPress generates does point to one. ⁶ The diagnosis recorded the fact, without the cause.

The native sitemap worked, and it listed 19 /go/ addresses created by a redirect plugin. ²⁰ They existed only to send visitors somewhere else. They made the list because the native sitemap includes every public content type, and the plugin had created one. ⁶ Google asks for a sitemap with the addresses you want to see in search results. ²³ Nobody decided to put those 19 on the list. Nobody decided to take them off.

Finding 6: two people where there was one

The users sitemap exposed /author/daniel/ and /author/daniels/. Both showed up in Search Console, with 142 and 96 impressions over 90 days. ²⁰ The native sitemap includes author archives. ⁶ Rank Math suggests turning that archive off on a single-author site. ¹⁴ Neither of them knows that daniel and daniels are the same person. Pulling the pieces of one entity back together is a job with a piece of its own, on entity fragmentation.

The pattern behind the six findings

In none of the six findings was a feature missing. WordPress generated what it was programmed to generate. What was missing was someone deciding what should exist, and checking what came out.

The half-hour check in your dashboard

Print the three-column sheet and take this tour of the dashboard, in order:

1. Settings > Reading. Make sure the “Discourage search engines from indexing this site” box is unchecked. ¹⁰
2. Settings > General. Read the Site Title and the Tagline. Both go into the homepage title, and both need to be in the site’s language. ^{2 22}
3. Settings > Permalinks. Write down the format that is selected. If the site already has pages on Google, do not switch on impulse, because every address changes with it. ⁸
4. The sitemap. With an SEO plugin, open the address the plugin’s screen gives you. Without a plugin, open /wp-sitemap.xml. For each type listed, ask whether it should appear on Google. ²³
5. The authors. In the users list of the sitemap, each name should be one person. And each person, a single name.
6. The robots.txt. Open yourcompany.com/robots.txt and look for the line that starts with Sitemap. ²³
7. Posts and pages. An article with a date and a topic goes in as a post. Pages are for what does not change, like the page that introduces the company. ²¹
8. The crawl. Run a crawl and count the pages with no meta description and with duplicate titles.

With the sheet filled in, the next step is the full order of an SEO audit. It starts with your measurement tools and ends with the business.

From now on, when someone says the site’s SEO is done because the plugin is installed, you will ask about the third column. That is where all six findings of our diagnosis lived.

Open the dashboard today, go to Settings, and start with Reading. The other technical fronts are in the technical SEO section.