Best ways to recommend articles based on labels? | Community
Skip to main content

Best ways to recommend articles based on labels?

  • December 13, 2024
  • 2 replies
  • 0 views
kyle.kowalsky

I'm hoping to get some guidance. We want to try to give people access to the articles they probably need, ideally before submitting a ticket, whether using the chatbot, suggested articles, email autoreplies, etc. The thing I'm getting hung up on though is recommending articles based on a specific tag or label without me manually entering each article I want to recommend.

Ideally, any of the following would work, but I haven't found anything to let me do roughly what I'm after:

  • On the New Request page, specify suggested articles based on the specific request form (can't seem to find any way to tweak these suggestion settings)
  • Using the conversation bots to recommend articles based on a tag (have to enter articles manually, and only 3 per step)
  • Using conversation bots to link to a Guide search for the content tag (looks like this can be done, but I need to find the content_tag IDs manually to make the links, which is fine for 1 or 2 buttons, but we've got many more to do)
  • Email autoreplies based on the form to include articles with a given tag/label

Does anyone have any idea how we could go about getting the results we're after?

      2 replies

      Elaine14
      • Elaine14
      • July 6, 2025
      Hi Kyle,
       
      Thanks for sharing your detailed use case and the challenges you’re facing with recommending articles dynamically.
       
      Zendesk Guide and messaging tools do have some capabilities for surfacing suggested articles, but they do come with limitations when it comes to fully automating recommendations based on tags or labels without manual intervention.
       
      Here are a few suggestions that might help you get closer to your goal:
       

      • Request Form Suggestions: Unfortunately, the default New Request form suggestions aren’t highly customizable in terms of filtering by specific request forms or tags. You might explore creating separate help centers or sections to guide users more contextually.

         

      • Conversation Bots: While conversation bots can recommend articles or provide search results by tag, as you noted, entry of article IDs or tag IDs is manual and limited in number. If you need to scale this, you might consider building custom bot flows using Zendesk’s API or integrating third-party chatbot platforms that support dynamic content retrieval.

         

      • Guide Search Links: Using conversation bots to link to Guide search by content tag is a good approach for broad article discovery. Maintaining the tag IDs can be automated via scripts or syncing with your CMS if you have developer resources.

         

      • Email Autoreplies: Including articles with specific tags or labels in email responses is possible via placeholders and triggers, but automating this fully based on form inputs requires careful trigger configuration and may need custom integrations.

         

      Overall, achieving fully dynamic recommendations with minimal manual setup usually involves some customization or development work. If your team has developer resources, I’d recommend exploring Zendesk’s APIs alongside your chatbot or automation tools to build a more seamless experience.
      R
      • Rh80
      • August 1, 2025

      Hey there,
       

      Totally get what you're trying to do—serving up relevant help articles before a ticket is submitted is a great way to deflect tickets and improve the user experience.
       

      Here are a few ideas and workarounds that might help:
       

      1. Form-Based Article Suggestions (New Request Page)
      Unfortunately, Zendesk doesn’t natively support customizing suggested articles per request form out of the box. But you could consider:
       

      Using custom triggers or JavaScript on the Help Center to detect the selected form and dynamically suggest relevant content (requires dev resources).
       

      Alternatively, route users through a custom landing page where they choose their issue type, and you display curated articles before the form.
       

      2. Conversation Bot Article Suggestions
      Yep, you’re right, it currently allows only 3 manually selected articles per step. To scale this:


      You could create topic-based flows, each tied to a general tag, and cycle article suggestions across different steps. Or use custom Guide search links filtered by tag.
       

      3. Guide Search Links with Tags

      This is actually a solid workaround. While getting content_tag IDs manually is tedious, once you have a list, you can:
       

      Use URL structures like /hc/en-us/search?utf8=✓&query=&label_names=your_tag_name to link to tagged content.
       

      Build a reference doc with tag names + links to copy/paste when setting up flows.
       

      4. Email Autoreplies Based on Form

      You can’t directly inject article tags into emails, but you can:


      Create separate triggers or automations per ticket form.


      Insert a static list of relevant articles (or tag-based search links) into each auto-reply.


      TL;DR

      There’s no perfect native way to auto-recommend articles by tag dynamically without some manual setup, but combining:
       

      Guide search links by tag

      Bot flows per topic

      Triggers based on form types

      ...can get you 80% of the way there.


      If you have dev support, custom scripts on your Help Center could unlock more automation.

      Terms & ConditionsCookie settingsAccessibility statement