Search is possibly the most used and expected service to be found on websites. Users will often turn to the search box on your site before exploring the information architecture you have carefully crafted for them. Moreover, users will often prefer to use Google Search than interact with a site which is not delivering on findability!
Search will therefore be a vital feature on sites which have enough scale to make browsing the content impractical. Getting it right is a important for the success of the site.
The right foundations
The success of a search solution is not just down to the technology being used to deliver the search results. The chances of success are increased when the following foundations are in place:
- The right content: A content strategy needs to be in place and delivered on. Does your site have the information your users are looking for. Are the right keywords on the page and given the correct prominence?
- The right structure: Faceted solutions will rely on on data being categorised and classified in consistent, cross cutting ways. Good factors must be "mutually exclusive" and "collectively exhaustive". Designing the content model with this in mind will ensure that the content can be filtered properly.
- Indexing all the things: All of the relevant content on the page should be indexed, including the content in components and associated metadata. What the search engine sees may not be everything the user sees.
- Excluding unnecessary things: Some content is not suitable for the search engine. Removing whole content types or individual pages can help with the results.
- Extending with keywords: Sometimes a page may be about a certain thing but the keyword doesn't appear on the page. It is helpful to have the capacity for editor to augment the content with keywords.
Full functionality
Users will also be expecting features such as:
- Autocomplete for search terms
- Synonyms
- Typos
The right solution
At Morpht we generally recommend one of the following two solutions
Search API: Drupal comes with a very capable search solution called Search API. It is a framework for integrating in search backend such as the database backend or Solr. Solr is a highly scalable solution based on Lucene which provides strong full text and facetting capabilities. If you are looking for a Drupal based solution this is the most reliable way to go.
Sajari: Sajari is a AI powered SaaS solution which offers a number of advanced features for those looking for extra control over boosting and curated search results. Sajari is able to track the results users click on so that the engine is able to learn which results users favour and then promote those for a better search experience.
