Reference · built on requestOperator by FibricCustomer support

Knowledge Gap

Compares resolved conversations with what the help center covers and what people searched for. Proposes a draft article for review.

Request accessPricing ↗

About

The questions agents answer by hand every week are the articles the help center is missing. Knowledge Gap reads solved tickets and the tags on them, the articles the help center already holds, and the terms people typed into search. Where a question recurs and no article answers it, or an article exists but searches for it fail, it writes a draft.

The draft goes to the help center as unpublished, or to a document, with the tickets it was written from listed. An editor reviews it, changes it, and publishes it. The operator never publishes. The record shows the tickets counted, the search terms, and the article proposed.

This is a reference listing. It documents what Fibric would read from Knowledge Gap and what it could propose, based on the vendor's published interfaces. Fibric builds it under a managed deployment when you request it; selecting it here installs nothing.

Inputs

  • Solved tickets in Zendesk with their tags, subject, and solved_at, and the agent's public replies
  • Articles in the Zendesk help center, read through the incremental articles endpoint by update time
  • Articles and collections in Intercom, and articles by collection in Help Scout Docs
  • Search terms from Google Analytics 4: the view_search_results event and its search_term parameter
  • Whether a term finds anything: the Help Scout Docs search endpoint, and the Zendesk help center search
  • Your style rules for articles and who may publish in each section

Proposed actions

  • Target capability: propose a new article as a draft in Zendesk, with draft set at creation and the source tickets listed
  • Target capability: propose an article in Intercom in the draft state, under the collection you name
  • Target capability: propose an unpublished article in Help Scout Docs, in the collection you name
  • Target capability: propose a revision to an existing article as a document for the editor, when the article exists but is not found
  • Target capability: propose retiring an article that no ticket has needed and no search has reached

Proposed actions are target capabilities. Every action runs propose-first and needs a validated deployment and the appropriate permissions.

What you can build

  • Draft the article the tickets keep asking for

    The same question appears in solved Zendesk tickets each week and no article matches. The operator proposes a draft in the help center with the tickets listed. The editor reviews and publishes.

    With Zendesk, Google Analytics 4

  • Fix the article nobody finds

    Google Analytics 4 shows a search term repeated; the Help Scout Docs search returns nothing for it. The operator proposes adding the term to the article's keywords and a revised title.

    With Google Analytics 4, Help Scout

  • Write the draft in Intercom's own collection

    For an Intercom help center the operator proposes the article in the draft state under the right collection, with the source conversations listed in the body for the editor.

    With Intercom

  • Keep a shared draft for a longer review

    When several editors must agree, the operator proposes the draft as a document in Google Drive with the ticket references, and files the help center draft only after they approve.

    With Google Drive, Zendesk

Requirements

  • A support connector for solved tickets and their tags: Zendesk, Intercom, or Help Scout
  • A help center the operator can write drafts to: Zendesk help center, Intercom Articles, or Help Scout Docs
  • Google Analytics 4 with enhanced measurement on the help center, so search terms are recorded
  • An editor who owns the help center and approves every draft
Authentication
Runs on the credentials of the help desk, help center, and analytics connectors you attach. It holds no key of its own and publishes nothing.

Limits

  • Google Analytics 4 records the term searched, not whether results came back. Misses are tested by running the search again
  • In Zendesk the draft flag can be set only when an article is created, so a revision is proposed to the editor, not written over
  • An Intercom article needs an author_id that is a teammate, so the draft is attributed to the approving editor
  • It reads what tickets say. A question answered by phone and closed without notes is not counted

Access and pricing

Reference listing. Fibric builds the operator under a managed deployment when you request it. Your quote covers the build, capabilities, usage, and support.

Request Knowledge Gap ↗

Questions and answers

What does the editor approve?
The draft article: title, body, section or collection, and the list of tickets and search terms it answers. The editor can rewrite it or decline. Approval creates the draft, unpublished. Publishing is a separate action the editor takes in the help center.
What record is left?
A receipt with the tickets counted, the search terms and their result counts, the article created and its id, who approved, and when. The draft itself lists its source tickets so a later editor can see why it exists.
Will it publish or edit a published article on its own?
No. New articles are created unpublished. Changes to an existing article are proposed as a document for the editor rather than written to the article. It never changes what customers see.
Ask about Knowledge Gap

Ask about the capabilities and requirements in this listing.

For project-specific requirements, contact Fibric.

This operator is developed, published, and supported by Fibric. Third-party names and logos identify the systems an integration connects to; they are the property of their respective owners, who are not affiliated with Fibric and do not sponsor or endorse this listing. Trademark policy