Search is never a user’s end goal but rather a means to an end; search engines are enablers for solving tasks.
In this post, we’ll discuss how Architech crafts superior user experience for search by using the powerful, open-source search engine, Apache Solr. Solr is built atop of Apache Lucene, another open-source project used to power services as large as iTunes search.
Whether the task is to find information about a topic or to find products for purchase, a search engine is usually the starting point. We expect that search engines will know what we want and present the information that is relevant to our searches in a meaningful way, search therefore needs to be efficient, comprehensive, and relevant.
A well-designed and effective search experience should enable and encourage the user to continue moving towards their end goal.
Search should guide users to find what they are seeking by providing relevant information and options to narrow down the results, and should never present dead ends. When the search terms are generic, the search engines should offer recommendations and allow the user to continue forward. As such, the true value of the search winds up being in the process of accumulated learning and information acquisition through discovery.
Not all search engines are built the same way, though. Some search engines are free services (e.g. Google, Bing) while others can be paid products (e.g. Oracle Knowledge, Oracle Endeca, Microsoft FAST). However, if you strip away the packaging, at the end of the day they all have to fulfill the user’s needs and satisfy expectations. Solr is a free, open-source software package that can be rolled into your project, service, product, or site.
Let’s step through the user’s search strategy and explore how Solr’s capabilities supports the delivery of an enhanced search experience on each of the following search stages:
1. Search formulation
2. Results comprehension and analysis
3. Query refinement and reformulation
The most critical part of information discovery is arguably query formulation as a user tries to determine where to start and what words to use to begin their search. Users are not always sure of what information they’re trying to find, most often because they cannot recollect details. Therefore, choosing the correct search terms can be a challenge. In this situation, a user will usually start by typing a keyword and add more as they type.
Solr’s auto-suggestion is a feature that transforms a recall problem into one of recognition. As the user composes the query, the system predicts and provides immediate feedback. The system can also offer corrections for spelling errors (a must-have feature for mobile users).
Auto-suggestion in action – a mainstay of effective search implementations.
Auto-suggestion and auto-correction cannot be undervalued as exploratory tools since they encourage query reformulation and disambiguation. Users are provided with clues on further refining their search terms and can be guided towards particular results.
Solr’s auto suggestion can be set up based on top questions, information that’s in index or based on custom question lists. The flexibility of Solr allows this functionality to be designed based on the needs of your website and your customers.
Results comprehension & analysis
The search results page is the most visible focus of the search experience, and can make or break your site’s conversion rates. Simply displaying search results sorted by keyword occurrence provides very little value and is taxing on the user’s comprehension of the results.
A meaningful visualization of search results makes complex information tangible, and allows users to rapidly assess search result quality, and if required, quickly reformulate their query for a follow-up search. This effectiveness of results comprehension is therefore hugely dependent on the manner in which the results are displayed.
Poorly optimized search results with meaningless descriptions and titles.
An effective search results page that provides snippets of info and other areas for explorations and related search.
Solr has the following features that can be used to enhance context on the results display and improve the user’s comprehension of the results:
- Sorting of results based on relevance – Recency, relevance, price, and distance from your location are all very useful candidates for result sorting.
- Location based results page – Where results are displayed on a map allows information to be put into context for the user. Map-based views are ideal for location-driven searching, i.e. searching on a mobile device.
- Keyword-in-context or query-oriented summary – Highlighting of the search terms within each result snippet allows users to quickly scan the verbiage around the highlighted keyword areas and validate the context of each result, and therefore the quality of the result.
Etsy.com search results page contains a left facets that includes location and user preferences.
Query refinement and reformulation
A common user search strategy is to start from general terms, review the top results, and if unsatisfactory, reformulate the query in an attempt to improve the results. This process is generally repeated until the information is found or the user gives up, making it critical for your design to guide the user in the process of search refinement and reformulation.
Once again, Solr has features that can help facilitate this process.
Faceted search permits users to refine or navigate a collection of content based on its structure or additional metadata. Users can drill down into the result set for a more focused perspective, or hop across multiple categories to more quickly cover the breadth of what’s available.
Hotwire.com faceted search filters can aid user to quickly sort and explore flight and hotel options.
The search and filter action of faceted search encourages users to explore and as a result will trigger new goals enabling the user to search in new directions.
Search for all
Without a doubt, search has transformed the way we think about the web, the way we interact with the vast amount of information and how we navigate to find what we need. If your site doesn’t feature a fast, powerful, and flexible search experience, then your users will simply leave and look elsewhere.
Apache Solr allows you to expose your site’s content in specific and meaningful ways, and allows your users to browse your site more effectively by way of discovering new information or focusing in on something specific.
Auto-suggestion, faceted search, keyword highlighting and location-based searching are some of the features of Solr that can enhance the user experience of search and increase customers engagement.