Site search | Doghouse SaaS Knowledge Base     [Skip to content](#main)  [ ![Doghouse SaaS Knowledge Base](/storage/branding/01M3N8PQ7PSDNQBBKKJM6XGW03.svg) ](https://nginx.deploy-lagoon-production.districtcms-docs.dh1.amazee.io)  [Product documentation](/articles) [Roadmap](/roadmap)   Toggle navigation      

  [Product documentation](/articles) [Roadmap](/roadmap)  

  1. [Home](/) ›
2. [Product documentation](/articles) ›
3. [Navigation &amp; site search](/articles?category=12) ›
4. Site search

 Site search
===========

Updated 5 days ago · 4 min read

 Search     Search  

Overview
--------

The header has a search box on every page, and the home page has a second one in its banner. Both send people to the same results page at */search*. Behind it sits an index, a copy of the site's content built for fast searching, which the site keeps up to date as content changes.

> **⚠️ Important:** The search index is administrator work. Content editors cannot reach these screens. What editors control is the content itself, which is most of what makes search good.

---

What search actually reads
--------------------------

The index does not look at particular fields. It reads each page as a visitor who is not signed in would see it, through the **Default** display set under **Manage display**, plus the title separately so that a title match counts for more.

Three things follow from that.

- Anything visible on the page is searchable.
- Anything hidden from the page is not. A field switched off in the **Default** display is not searched. See [**Controlling how a field appears**](/articles/controlling-how-a-field-appears).
- Unpublished pages are never in the results, whatever they contain.

### Documents are not searched inside

Only pages are indexed. A PDF, Word file or spreadsheet attached to a page is never opened and read, so a word that appears only inside that file will not match a search.

What is searchable is the page around it, including the words you use for the link. Give a document a descriptive link rather than *Download* or *PDF*, and people will be able to find it.

### Making your pages easier to find

Not every word on a page counts the same. Words in headings carry more weight than ordinary body text, so a page broken up with headings is easier to find than the same words written as a block of prose.

Where the words are

How much it counts

The page title

The strongest signal there is. It is indexed on its own as well as within the page.

**Heading 2**

Three times ordinary text

**Heading 3**

Twice ordinary text

**Heading 4**, **5** and **6**, and bold text

The same as ordinary text

The alternative text on an image

The same as ordinary text, but it is read, so it is worth writing

> **💡 Note:** The editor offers **Heading 2** downwards and no **Heading 1**. That is deliberate: Heading 1 is the page's main heading, which the site builds from the title, so there is never a reason to add one yourself.

---

What the results page shows
---------------------------

Results are ordered by relevance, the closest match first, under a count such as *About 15 results for "orbital"*.

- Ten results to a page, with paging underneath.
- Each result shows its content type, its news categories and date where it has them, the title, and the page's summary.
- Down the left, **Content type** and **News category** filters narrow the results, each with a count.
- When nothing matches, the page says *No results found. Try different or fewer keywords.*

![The search results page for orbital, with the Content type and News category filters and the list of results each outlined in red.](/storage/tQsGtzvRTSKFdxMS32LyaQ6hZzFzksmpTAHCFKfu.png)

---

Keeping the index current
-------------------------

An administrator will find it under **Configuration**, then **Search and metadata**, then **Search API**. The index is called **Content**.

Open it and the first thing on the screen is **Index status**, shown as a percentage and a count, such as 38/38 indexed.

![The Content search index screen with its Index status bar outlined in red, showing 38 of 38 items indexed.](/storage/pC8PKrFgQs72JmfDrnWCb8iekcuryTfJf6vtRmvf.png)

An item is added to the index as soon as it is saved, so that figure normally sits at 100%.

Index status

What it means

100%

Search is current. Nothing to do.

A few short of the total

Something was saved while indexing could not run. The site clears the backlog on its own, 50 items at a time, or open **Start indexing now** and press **Index now** to do it immediately.

0 indexed

Search will find nothing at all. Press **Queue all items for reindexing**, then **Index now**.

> **💡 Note:** Because saving an item indexes it, a page that cannot be found is rarely a matter of waiting. Check instead that it is published. If a page can be found by its title but not by words in its body, raise it with an administrator, because something is stopping that page being read in full.

  Related articles 
------------------

- [ Controlling how a field appears → ](/articles/controlling-how-a-field-appears)
- [ How people navigate your site → ](/articles/how-people-navigate-your-site)
- [ Page addresses → ](/articles/page-addresses)
- [ Meta tags and search appearance → ](/articles/meta-tags-and-search-appearance)
- [ XML sitemaps → ](/articles/xml-sitemaps)

 Was this article helpful?
-------------------------

Your feedback helps us prioritise what to rewrite.

      Yes     No  

 [    Back to Navigation &amp; site search ](/articles?category=12) 

 On this page 

**Need additional help?** 
Submit a support request through the [**support portal** ](https://doghouse.atlassian.net/servicedesk/customer/portals)or email 

 © 2026 [Doghouse SaaS](https://doghouse.agency) [Glossary](/articles/glossary)
