Understanding the Complete SOHO SEO Structure

What an “SEO plugin” actually handles: metadata, structured data, sitemaps and robots—SOHO SEO’s complete feature map

In the previous two articles, I explained how the search landscape has changed and how Google processes a piece of content, from crawling through to ranking.

There may be only one name—”SEO plugin”—but in practice, that label covers a collection of functions with entirely different purposes.

I think bundling all of them together under one vague label ends up confusing users rather than helping them.

So in this article, I’m going to organise SOHO SEO’s features into several clear groups.

Why SOHO SEO is necessary

Anyone opening Yoast or Rank Math for the first time is likely to feel overwhelmed.

There are more than a dozen menus and hundreds of settings.

What is often unclear, however, is whether those settings actually affect crawling, influence indexing, or merely control how things appear on screen.

I believe this is the fundamental problem with SEO plugins. They offer many features, but do not explain where those features are used.

That confusion is precisely what led me to start working as a beta tester myself.

Meanwhile, my site was sinking deeper into a maze, and its index was becoming a complete mess.

That is why I mapped SOHO SEO’s features onto the four stages discussed in the previous article—crawling, rendering, indexing and ranking.

I then added two more criteria.

The first is whether a task is handled directly on the site or through a cloud server. The second is whether it is free or consumes credits.

Viewed through these three lenses, the overall feature set becomes much easier to understand.

Group One — Metadata: The Identity Card for Every Page

The most fundamental group is metadata. It includes the SEO title, meta description, canonical URL, Open Graph data and Twitter Cards.

I think of this as a page’s “identity card”. It explicitly tells both Google and social platforms what the page is about.

There is one design principle here that I consider especially important.

If a custom SEO title or description has been entered in the editor, SOHO SEO uses it first. If not, it generates one from the excerpt, then the opening section of the body copy, and finally the site tagline.

The reason is simple: we should not assume that users will fill in a meta description for every post.

Most people do not. But if a post is published without a meta description, Google may pull any sentence it chooses from the body for the preview, and that is not always very compelling.

So I designed the automatic generation logic to serve as a basic safety net.

There is one more safeguard. If Yoast or Rank Math is already active, SOHO SEO’s metadata output quietly does nothing.

In my view, outputting the same metadata twice does not help SEO.

In fact, duplicate metadata can leave Google unsure which version to trust. That is why this group is designed as a default that works only when no other plugin is present.

Group Two — Structured Data and the Robots Meta Tag: Explicit Signals Sent to Google

The second group is structured data.

The core function is to output headline, featured image, publication date, modification date, author and publisher information through Article JSON-LD.

This may look similar to metadata, but it serves a different purpose.

I have also placed the robots meta tag in this group.

In particular, guaranteeing max-image-preview:large is essential. Without it, a page is not even eligible to appear as a large image card in Discover.

There is a reason I grouped this with structured data. Both are invisible to readers, but they provide explicit signals that influence how Google classifies a page and how it displays it.

I made one more practical decision within this group.

I also set an upper limit on how many content images from the body can be included in the schema and sitemap. This prevents the output from becoming excessively large, even in extreme cases where a very long article contains hundreds of images.

These details are not glamorous, but on a live site, this is exactly where things tend to go wrong.

Group Three — Sitemaps and Index Control: Managing the Crawl Scope and Search Database

The third group covers sitemaps and index control, which I discussed in the previous article.

It includes the standard sitemap, including image tags; the news sitemap, which follows the 48-hour rule; and noindex management.

I grouped them together because all three functions govern what is added to or removed from Google’s search database.

I also gave this group its own management screen.

At first, I simply placed the sitemap URL at the top of the dashboard. I eventually decided that was not the right approach, so I moved it into a dedicated menu.

The sitemap-generation logic remains untouched; only the management screen displaying its URL and status has been separated out.

Group Four — Content-Quality Diagnostics: Trustworthiness as Seen by Readers

The fourth group is fundamentally different from the first three.

Metadata, structured data and sitemaps are all technical tools for communicating accurate information to Google.

This group, by contrast, assesses whether the content itself is trustworthy.

The Discover checklist is a good example. It checks the featured image size, E-E-A-T requirements, whether the headline is clickbait, mobile display settings and publisher information.

I have also placed the paid emotional-style post check in the same group.

It does not rewrite sentences or interfere with publication. It provides only a comprehensive assessment—a score, an overall evaluation and three supporting reasons—so the writer can see the bigger picture.

An earlier version suggested “more human” alternative sentences at the sentence level and could apply them automatically. I removed that approach entirely.

The quality of the alternatives was inconsistent, and applying the wrong suggestion could make the body copy feel even more unnatural.

Now the system provides only the evidence a person needs to make a decision, leaving the actual revisions to the writer.

Group Five — Links and Internal Structure: The Trustworthiness of the Entire Site

The fifth group is internal and external link management.

It extracts potential key terms from the body copy and automatically links them to related articles on the site or to relevant Wikipedia pages.

The network-based option is further divided according to whether a person reviews the suggestions before applying them or lets the system publish them automatically without review.

These three levels are designed to give users a genuine choice between caution and convenience.

This is not a new recommendation method. It simply runs the automated network-based extraction process every hour, processing a set number of items without human intervention.

I initially created a separate menu for it, but realised that it duplicated an existing function and merged the two.

That process taught me that splitting features into multiple items does not necessarily make a product better.

Group Six — Site Management: Redirects and Tags

The sixth group consists of the less glamorous but essential site-management functions.

This data exists in the site’s own database in the first place, so there was no reason to route it through an external server.

These features have been continuously refined to remove unnecessary steps.

Even at this point, it is hard not to be pleased with the result: the system eliminates the hassle of installing and managing so many separate plugins.

And there is one more thing.

Final Group — Automated Monitoring: Site Error Checks

The final group is a paid monitoring feature that checks robots.txt and four types of sitemap every day, sending an email only when it detects a problem. I designed it so that you do not need to think about it during normal operation; it simply alerts you when something goes wrong.

Why did I build this?

As a result, indexing kept declining while the site’s overall trustworthiness continued to deteriorate.

Can a supposedly reputable plugin really create errors and leave users in such a bewildering situation?

You are free to trust the plugins you use—or not. Looking back at my own experience, however, I do not trust those plugins at all.

In researching various cases, I found numerous reports from different sources describing these kinds of recurring errors.

In the end, I learned that the safest approach to managing WordPress was to write and use code I could trust myself. I also concluded that building a system capable of detecting problems immediately—whether they originate with the server or the hosting provider—adds another important layer of protection.

SOHO SEO relies on a cloud server for the actual checks and email delivery, but the server cannot know which sitemaps this site is really using because that information exists on the WordPress side.

The conclusion: Why SOHO SEO?

Here is the short version.

The articles that follow will explore each point on this map in greater depth.

Next, I will look at the first group—and the feature users encounter first in the editor: the real-time Discover checklist.