Planeta SEO
The case study continues · 24 August 2026

A list of feature names tells you nothing? Here is what each one found on a real site.

The home page carries the finding from a client site already running this plugin: 56 published pages, an average score of 96 out of 100, first place in Google for most of the terms in its industry. This page carries it further. Each feature comes with what it found on that site, and where it found nothing, that is written down too.

Deep scan

Fifty six pages took about a minute, and the result is a list of fifteen lowest scoring pages with the exact reason for each

The plugin opens your page the same way a visitor does, then looks at what the server answered: the number of H1 headings, images without alt text, how much text there is, the canonical and the structured data. On the site in the case study, page scores ran from 92 to 100, and the reasons were short and specific: title is short, title is long, no share image, nothing to report.

Accidental noindex

It catches the kind that comes from a server header rather than the page itself. This is the most expensive SEO mistake there is, because the page simply vanishes from search while everything else looks fine.

How it looks

A site moves off staging and the configuration still carries X-Robots-Tag: noindex. Inside WordPress everything is set correctly, and no tool reading the database has any way to see it.

Descriptions that never arrive

It compares the description you wrote with what is in the delivered page. If they differ or it is missing, it says where to look.

How it looks

A caching plugin serves a snapshot taken before the description was written. In the admin everything is correct, and Google keeps picking its own sentence off the page.

Titles the theme rewrites

If the theme appends the site name, the finding says how long the title is afterwards and where that option lives.

On the site in the case study

Title length was the most common finding. Among the fifteen lowest scoring pages, six were flagged for a short title and one for a long one. Neither shows in the admin, because what gets measured is the delivered title, together with whatever the theme appends to it.

Heading structure

Page builders break it constantly, because the heading level gets chosen by font size rather than meaning. The finding says where a level was skipped.

On the site in the case study

Here the finding was clean. The H1 column reads one on every page in the overview, even though the site is built with Elementor. That is a result too, and we write it down: a check that always finds something is not a check, it is advertising.

Answers below subheadings

It checks whether the first sixty words below each subheading actually address that heading. That is what both Google snippets and AI answers look for.

How it looks

Below the heading „How much does it cost" sits a paragraph about company tradition. The question is asked, and the answer turns up in the third paragraph.

Whether the text arrives at all

If the content comes from JavaScript, the page is empty for most AI crawlers. Google can run scripts, they cannot.

How it looks

The database holds the whole text, and the HTML the server sends holds a handful of words. The rest is added by the builder after loading, inside the browser.

Search Console

One page sat fourth in Google and lost 34 clicks every 28 days, and no ranking report will ever show you that

Everything else in the plugin is advice about what you ought to do. This is the only part that says what actually happened. Raw Search Console numbers mean little to most people, so here they are joined into two findings that say directly what to do.

Clicked less than they should be

CTR on its own means nothing until you compare it with the position in the results. Three percent is poor for third place and excellent for tenth. Every tool shows that number, and nobody compares it with what is normal for that position.

What is shown is the count of missed clicks, not a percentage. Percentages move nobody, lost visitors do.

Clicked less than they should be28 days
position 4.2   impressions 522   clicks 8
CTR 1.5 per cent, where fourth place normally gets 8.0
second row a PDF at position 8.4, 382 impressions, 4 clicks, 7 missed
34 clicks missed over the last 28 days, on a single page. The content is good, since Google put it fourth on its own and offered it 522 times. What is left is for the title and description in the results to give someone a reason to click.

Dropped after being edited, and on the site in the case study that list was empty

Individually both pieces of data are available everywhere. Search Console shows the position drop, WordPress shows the modified date. Nobody joins them, so an SEO specialist does it by hand with two windows open.

Only a drop that coincides with an edit in time is reported. Pages drop constantly and for a hundred reasons, so a drop on its own is noise. Only a drop that began after an edit has a cause you can actually check.

On the site in our case study this list was empty, because the score had not moved by a single point since the previous scan. Good news for that site, no use as an example, so the panel beside this shows what the screen looks like when something really does drop.

Dropped after being editedscreen layout
page A service page
edited 8 August 2026
position was 4.1  →  now 11.3
clicks 175  →  38
Open the revisions and see what changed that day.

Connecting requires you to create your own Google access, and the plugin walks you through it in seven steps. The reason is that an embedded key in an open source plugin would be public on day one. The side benefit is that your data travels straight from Google to your site, with no intermediary. Other plugins route it through their own server.

Daily work

On the site in the case study, 294 images had no description and 171 addresses returned a 404, waiting in two tables instead of 465 separate visits

Bulk editor

A table of every page with fields for the title and description. Saved the moment you leave a field, with no submit button.

On the site in the case study

There was no work of that kind here, because all 56 pages already carry a meta description. The table was used for title work instead, on the seven pages where length was the finding.

Image alt text

Every image without alt text in one table. You type, press Enter, the cursor jumps to the next one. Each row also says which page uses the image.

On the site in the case study

294 images with no alt text, on a site otherwise scoring 96. Alt text is never filled in automatically. It is also what a screen reader reads out loud to a blind visitor, and a raw file name read aloud is noise rather than a description.

Pages with no incoming links

Google reaches a page through a link. A page with no links pointing to it exists only in the sitemap, which is the weakest signal a site can send.

Why it is useful

Each such page comes with a list of pages that already mention its topic in their text. You open one, find the phrase, turn it into a link. It takes a minute.

Broken links

It opens every link from your site and checks whether the destination still exists.

The error nobody notices

Site owners do not click their own links, and a visitor who hits an error does not report it, they leave. The site you linked to three years ago has changed its address since.

Pages competing for one keyword

When two of your pages target the same term, Google has to choose, often picks the wrong one, and changes its mind from week to week.

Why you cannot see it yourself

Because each page looks correct on its own. The problem exists only in relation to the other one.

Redirects and 404s

An automatic redirect when you change the address of a published page, CSV import and export, and bulk actions on logged errors.

On the site in the case study

171 addresses returning a 404, mostly old documents and pages renamed along the way. Not one of them is written down anywhere in WordPress, so until they are listed they do not exist as a task.

If you work for others

A client who sees nothing but an invoice each month cancels the first month money gets tight

A report with your logo

Every month the client sees an invoice. Without a report they see nothing else, and the retainer gets cancelled the first month money is tight.

The report carries your logo and contact details, four headline numbers with the change since last month, a graph, a list of improved pages, and fifteen to work on next. It deliberately does not list everything, because a report with fifty rows does not get read.

Permissions in three levels

A client writing their own titles does not need an administrator account. There are three levels: editing SEO fields, tools and reports, and settings.

Alongside that, a change log records who changed what and when, with the old and new value. When a page vanishes from search, that is the difference between half an hour of arguing and one glance.

Settings transfer

Export every setting to one file and import it on the next site. The tenth site does not get configured from scratch.

Self-check

Eleven checks that look at reality rather than settings. Whether the sitemap opens, whether the server serves the robots.txt you configured, whether the site can call itself, and whether protection in front of the site refuses AI crawlers even though you allowed them.

Plainly

Six things this plugin does not do, said before you buy rather than after

Nobody normally writes this on a sales page. Better to read it now than to be disappointed later.

You can run the same scan on your own site tonight

The basic version has no time limit. The tools unlock with a key that is paid once and never expires.

See the plans