SEO Audit. Part 2. Interpretations of 🕸️Accessibility Metrics

These answer the question “can search engines can access the pages on your site easily and successfully”?

As discussed in SEO Audit Template, SEO reporting uses two main groups of metrics: SEO performance metrics and Website Analytics metrics. 

SEO performance metrics measure how well your site is doing vs. SEO through direct success metrics. Website Analytics metrics come from crawlers and we look at them to understand and explain the Performance metrics we see in the Funnel view.

Website analysis metrics can be divided into 2 broad categories: 

• Accessibility metrics and 

• Content metrics. 

Below is a more in-depth discussion of Accessibility metrics. These answer the question “can search engines can access the pages on your site easily and successfully”?

Accessibility metrics reporting in SEO Audit include the following section:

1. Distribution of URLs

2. HTTP codes

3. Performance

4. Inlinks

5. Outlinks

6. Sitemaps

Distribution of URLs

Things analysed:

1. Active / Not Active URLs Distribution

2. Indexable / Not Indexable URLs Comparison

3. Page depths

4. Internal PageRank by depth

The report shows the stats of active URLs versus not active ones. Active pages are those who generated SEO traffic (whatever the number). Active pages went all the way through the Organic Search Traffic funnel. The main opportunity lies in transferring inactive pages to active.  

Distribution section of the site audit report indicates whether a page is indexable or not. But what does it mean indexable in the first place?

An indexable URL:

1. Returns 200 Status Code.

2. Does not contain a “noindex” meta tag.

3. A canonical tag points to itself or is not set (page is not declared as duplicate).

4. Has a text (HTML content).

Page depth

Page depth indicates who deep the page is within the site’s structure. It is the number of clicks needed to reach the page using the shortest path available, starting from the Homepage. 

• Homepage: 0

• Pages with links (to them) on the homepage: 1

• Two clicks away: 2, etc.

Typically, the deeper pages are, the less likely they are to be found and explored by search engines.

Internal PageRank

Page depth is complemented by a notion of Internal PageRank, i.e. which pages stand out on the basis of internal links. The more links to a page, the more ways to access it. Yet, Internal PageRank not only takes into account the number of page links, but also how powerful these links are (not all links are equal).

We can consider internal PageRank as a measure how website structure promotes some pages (of the site) more than others. The internal PageRank distribution shows where the website’s internal popularity, or link equity, flows to the most.

HTTP codes and Performance

HTTP codes and Performance tie into accessibility and indexability. Pages returning HTTP error or redirect code are not indexable. Additionally, redirects within your site represent unnecessary hurdles to the search engines trying to access your pages. 

Page speed experience by robots, which is detailed in the performance section, also has a significant impact on search engine crawl - if pages are served fast, then robots can explore more within the same timeframe. 

HTTP codes

HTTP codes section summarises the responses that site crawler received from all pages:

• How many pages returned content successfully?

• How many were redirected?

• Returned client errors or server errors?

“Bad codes” include any type of error or a redirect, whereas Good codes happen when pages return content successfully. 

It’s useful to see HTTP codes by page depth: this highlights the issues that are potentially very visible to users and search engines alike, because they are near to the top of the website.

Also, we have information by segments to see how the issues are distributed among the different types of pages. 

Likewise, errors on top navigation pages require the biggest attention. For example, in the case of redirects, we’ll want to update links to point directly to the page at the end of the redirect. Several redirects in a row create a redirect chain. In the worst case, the redirect chain comes full circle completing the a redirect loop. These are bad as they never lead to a page with content. 

Performance

Performance section is about load time performance as experienced by robots: how long does it take for a full HTML page to be received by the crawler? Looking at this by HTTP code allows us to focus on load time for pages that deliver content successfully, which are the important ones. 

Likewise, some issues can be dismissed. If our slowest pages are only server errors, that’s OK; the server errors are typically slow because it takes time for the server to fail some processes and reach the conclusion that the page cannot be delivered. 

As with other data, looking at performance data by page type is key. 

Inlinks

Inlinks are analysed as signals of internal popularity for the pages receiving them. In this section we analyse the proportion of links with Follow vs No Follow attributes (blocking the ways for robots) and statistics about the follow inlinks, such as total number on the site and the average number per page. 

We also focus on pages which can only be reached by a few inlinks or even a single one, which are basically saying, among the least important pages. They are barely attached to the site’s structure and if a single link was removed for whatever reason the page would become “orphaned”. Robots with very few links are generally hard to find because robots will first have to come across one of the few pages that lead to them. 

What is the ideal example here?

In the image below you can see two charts. Example one (blue colour): too many pages are barely linked, while there’s a “wall” at the far right part of the chart. An ideal distribution might look more like this (violet colour), where there are several inlinks for even the least important pages and a more gradual slope leading up to the top.

Another thing to pay attention to is anchor text diversity. In this example the anchor links are not varied: there is never more than a couple of different anchors on all links to a given page, even when the page receives many links. So we are missing the opportunity to better describe pages via links that point to them.  

Outlinks

Here we mostly focus on unique outlinks, those that matter for our analysis. We analyse the outlinks distribution, i.e. whether they point to pages on your site or outside (internal and external outlinks). We also focus on whether robots can follow them or not. 

Next, it’s important to see if links go to indexable or non-indexable pages. This report also details the types of broken links when a link points to a page that responds with an error or a redirect. 

Internal follow links pointing to pages that robots cannot access are wasting internal linking equity. How much is wasted is summarised here.  

As per links pointing to external websites they are inventoried and consolidated in a table.