When you publish an article, or you save changes to an already-published article, we reindex that content for search. (We also do this for Custom content categories and Topic display categories!)
We will automatically index words entered in these fields:
- Title
- Permalink
- Body
- PDFs
- Meta description
- Search phrases
So as long as you are placing keywords in your article's title, permalink, body, or meta description, you don't need to do anything extra for that content to show up in search; it just does. 😊
When a reader interacts with your knowledge base, they'll see two main forms of search:
- Autosuggest search: As a reader types the phrase, a dropdown will appear below the search bar, showing up to 10 relevant articles.
- These suggestions are ranked based on the title, permalink, meta description, and search phrases only.
- We ignore body and PDFs here to keep the autosuggest search very responsive.
- Full search: Once a reader hits Enter or clicks the Search icon after typing in a phrase, they'll be taken to the full search results page.
- These results support pagination and factor all of the indexed fields into a ranking algorithm.
Since autosuggest uses a shorter list of indexed fields, the results you see in autosuggest may differ from what you see in full search.
Here is a breakdown of the main differences between autosuggest and full search.
Autosuggest | Full search |
---|---|
Suggests below the search bar as you type | Displays on search results page after hitting enter |
Searches title, search phrases, permalink, and meta description (no body or PDFs for speed) | Searches title, body, pdfs, search phrases, permalink, and meta description |
Doesn't use synonyms | Uses synonyms |
Limited to top 10 suggestions | Displays full results with pagination |
Highlights differences in title | Highlights matches in blurb |
There's also a third search option:
- Tags search: Tags aren't one of the fields that we automatically index as part of main search.
- The tags search is a separate search altogether, and will just pull a list of articles with the searched tag.
- A reader can perform a tags search by beginning their search with a colon--the autosuggest list will suggest tags they can search for.