Website structure SEO audit that helped identify traffic drops

Google Discover Traffic Drops

Over the course of few months, the client had been consistently receiving a substantial amount of web traffic from Google Discover. However, starting in March, the website no longer appeared in the service.

This unexpected change is easily noticeable on the overall Discover traffic chart in Google Search Console:

The Client’s Theory

One theory the client had about the sudden disappearance from Google Discover was related to the transfer of Facebook comments to the website. These comments frequently included offensive and disrespectful language. This could undoubtedly be a reason, particularly because such content violates the policies of Google Discover, which are even more strict than Google’s policies.

Our Take on Sudden Visibility Drop in Google Discover

The comment migration took place in December. Although it might have had an initial impact on traffic from that channel, surprisingly, the website began resurfacing in the service in January, February, and March.

This meant that we had to look further to identify the cause of the sudden visibility drop as the comments clearly weren’t the main culprits.

But before that, we carefully examined the website to make sure it met the important criteria for showing up on Google Discover. This review allowed us to create some content suggestions for our Client.

In short, we offered the Client advice concerning:

• Making the content more clear and understandable
• Including publication dates
• Adding more information about individual authors and the editorial team
• Putting a stronger focus on social signals

This was supposed to help the Client enhance his website’s visibility and user engagement in Google Discover.

Lack of Clear Content Strategy for Google Discovery

What’s more, our SEO audit revealed that our client had no well-defined content strategy. In most cases, such practice has a rather negative effect on the overall quality of a website – as per Google evaluation, of course.

Additionally, we found out that among all types of content published on the Client’s website, one, in particular, was responsible for driving traffic via Google Discover – Cooking tips.

Cooking tips / Recipes / News / Fun facts / Shopping / Show business / Health

As we were reviewing the content strategy, we noticed there was a substantial increase in Cooking tips articles published during November and December. Interestingly, the amount of this type of publication started to decline, starting in early February.

Cooking tips / Recipes / News / Fun facts / Health / Shopping / Information / Other / Show business

To us, there was a clear connection between the number of Cooking tips articles and the website’s performance in the Discover channel.

This highlighted the need to create a well-thought-out content strategy that would deliver two main goals:

Even though the SEO Audit service’s scope we did for the Client didn’t include a content plan, we provided him with some basic recommendations that included:

• Regularly doing keyword research to come up with new topic ideas that actually align with what the target audience is currently searching for. This can be done through Google Search Console.
• Creating topic clusters to demonstrate to Google that the website is an authority in the culinary field. We also suggested connecting articles through internal linking, which has a positive impact on SEO.
• Adapting the new pieces of content to the users’ search intent.
• Constructing unified and consistent information architecture, which depended mainly on the way the information was about to be presented on a website (eg. infographics, tables, bullet points).
• Content refreshing, especially the pieces that gain popularity during specific periods like the Christmas holidays and summer.

Too Many Page Tags

After thoroughly crawling the website, we noticed an excessive number of page tags in comparison to the actual number of pages comprising the entire website.

There were over 3,000 page tags and an additional 1,000 tags, whereas the number of recipes those tags were assigned to was just 700. This is quite an issue because page tags aren’t the same as social media hashtags, which they are often mistakenly taken for.

It’s worth noting that when adding a tag to a website or a blog, CMS, which in this case is WordPress, creates a tag page containing a list of all articles tagged with that specific tag. WordPress’s default settings allow search engine robots to index these pages, so all tag pages end up in Google’s index. This creates two main problems: keyword cannibalization and a waste of Google’s resources.

In plain English, each tag creates a separate page tag. When the number of tags exceeds the number of pages, this leads to a high number of pages that are not only of low quality but also compete against the homepage.

To handle this problem, we suggested the Client go through the following steps:

1. Grouping the tags, minding their topic relevance (eg. recipe-tag potato pancakes + recipe-tag pancakes made of potato)
2. Since the number of tags was pretty high, we advised against categorizing them manually. Instead, we suggested using AI, more precisely doing this through the API Open AI integrated with Google Sheets.
3. From each tag group, we suggested picking just one that has the highest number of impressions.
4. The remaining tags should be deleted, setting the 301 redirects.
5. Tags that generate zero traffic should be deleted with no redirects.
6. Giving the list of tags to the content team so that they can use them while writing and later publishing new pieces of content.
7. In the case of new topics, new tags should be defined.

It’s worth realizing that tags can be valuable for a website as long as they’re set up correctly. This means they shouldn’t duplicate content found in other sections of the website, and there should be several articles associated with each tag. This approach enhances internal linking and contributes to improving other essential SEO factors.

Too Many Pages with Thin Content

The crawl we ran using the Screaming Frog tool also revealed two other issues: low internal link score and insufficient amount of content published on pages.

We compared this result with statistics provided by Google Search Console.

An overwhelming 96% of web traffic came from pages within the Recipes and Tips catalogs. At the same time, the pages that make up most of the website, which are Tags, Recipe tags, and Ingredients, despite appearing frequently in search results, generated almost no traffic.

Similar results were shown by the Google Discover traffic report:

This meant that 87% of both traffic and views were generated by Tips, which made up only 7% of the entire website.

That being said it became clear that the domain was dominated by low-quality URL addresses. This, in turn, negatively affected its ranking with Google and made it vulnerable to changes and updates made to the search engine’s algorithms.

The truth is, that Google algorithm updates had a noticeable impact on the performance of pages in Google Discover over the past year. This resulted in sudden fluctuations in web traffic from this channel, which is quite unpredictable by its nature.