Kidam Mun

I spent some time working through the limitations of labels and, although not the most elegant solution, I found a method that works well. I’m sharing it here for others who may be facing similar issues.

The approach involves manually boosting the search ranking by adding hidden, repetitive sentences to the document. For example, if you want a page to rank higher when searching for the term “scan depth,” you can add the following hidden sentence multiple times at the end of the source code.

After adding this hidden sentence about 30 times, the page consistently appeared at the top of search results for “scan depth.”   

Hope Zendesk provides a more elegant solution in the future. 

Hi Katrina

I think custom pages are what you need.

Custom pages don't appear in a section or a topic. 

Rather, they have their own URLs that you can use to direct users to the pages.

https://support.zendesk.com/hc/en-us/articles/4409012911770

I do agree with Bradley. 

Lazy-loading is essential for articles with multiple videos.  

But the source code doesn't allow loading="lazy" in iframe. 

Hi,

Does it support the structured data required for Google’s rich results feature?

Here is the workflow I tried today.

Recently, I have become interested in applying rich results to my FAQ and Q&A articles.

So, I tried both their example codes (JSON-LD and Microdata) in the source code of my help center article.

https://developers.google.com/search/docs/appearance/structured-data/faqpage

However, it doesn’t work.

I think it’s because the supported HTML for help center articles is limited.

https://support.zendesk.com/hc/en-us/articles/6644509092378

Using HTML blocks, I can copy and paste unsupported HTML elements into the source code editor.

So, I used HTML blocks to add sample code in content blocks and inserted that content block into the test article.

https://support.zendesk.com/hc/en-us/articles/6739380623770

However, I cannot see the result when I check the test article on the test site.

https://search.google.com/test/rich-results

I’m not sure if Zendesk supports “structured data,” but if it does, a guide or tutorial would be really helpful for users.

If Zendesk cannot support “structured data,” I hope it will be supported soon.

Thanks.

Last year, I integrated Zendesk with Google Analytics 4 (GA4) through the integration section of Zendesk Guide settings. 

Now I want to use GA4 through Google Tag Manager (GTM). 

Do I need to cancel my previous integration after following the instructions in this article? 

Will this cause any data discontinuity between the existing GA4 data and the new data collected through GTM?

It would be helpful if admin can control the search results for specific article and keyword easily.

Currently, according to  “About help center end user search”, the weight of Labels is 2.8 and title is 3.0.

So I think the article include the labels should be the top rank article if there is no article which title is exactly matching the search keyword. 

But it's not. Relevance scores are impacted by a text analysis process, and this factor is really dominant, especially semantic search is introduced. 

When I use the keyword as a label, I can find the article in search rank #20. 

( When I use the keyword as a title, I can find the article in search rank #1. )

As you know, title cannot include all keyword when the article is long. 

So label is needed, but it doesn't work effectively now and hard to control the search rank of keyword directly.

Please consider the change of label weight, or other solution. ( I made ticket about this issue, but cannot find clear solution. )

Hi Ifra, 

In this case, I can open/close the accordions by clicking the TOC but cannot do so by clicking the headers. 

Is there any way that both methods are possible to open/close the accordions? 

( When I use Step 1 code in this article, and Step 2 code and Step 3 code in Tip: Collapsible headers in articles or templates (accordions) (https://support.zendesk.com/hc/en-us/community/posts/4408860552730)

It works. 

In this case, the source code is changing to following.

Section One