Product

How to use site search to avoid “no results” pages
facebooklinkedintwittermail

With sites like Google and Amazon providing instant and personalized search results every single time we search for something, today’s web users expect the same experience from every site they visit. But searches that return “zero results” or “no results” pages are frustrating roadblocks in the user experience.

It’s a dead end for users, since it doesn’t answer the original query and doesn’t suggest meaningful alternatives. It can both annoy the user and damage the business in a number of ways, including:

  • Harms overall UX. A site search that doesn’t return results fails to deliver on its most basic premise: to connect users to what they need to find. When users don’t get what they need, they can’t meaningfully interact with your site. 
  • Undermines users’ perception of site content. When a search fails, users might assume you don’t offer the content or products they seek. Users are more likely to leave without becoming acquainted with everything your site has to offer, which is reflected by high bounce rates after unsuccessful searches. 
  • Negatively affects KPIs. Frequently occurring “no results” pages can have a negative impact on the brand itself. A subpar site search gives users the perception that the site has poor-quality service relative to competitors. This causes increased bounce rates, decreased engagement and conversion rates, and reduced sales, among others. 

Often, companies will try to counteract this effect with a helpful or even cute or funny “no results” page, but that doesn’t exactly help someone who is frustrated that the site hasn’t delivered on its promises. The internal site search engine still has not met the basic expectation of understanding the query and providing the most relevant results, so what reason do they have to stick around? There are a dozen more pages on Google full of websites oversaturated with content, eager to steal your customers. What can we do about this problem?

 

Fine-tune your site search

The first step is realizing that the content that the user is looking for often does exist; the problem is just that the site search is not optimized well enough to find matches. By adjusting relevance, semantic settings, and other components of your search engine, you can eliminate “no results” pages altogether. 

Well-designed site search systems dynamically break down the semantics and structure of a complex search to provide the best matches. This allows the engine to return relevant results, even when the query is misspelled or does not use the exact same terms as the relevant product listing or webpage.

Example of a complex query returning relevant search results

6 ways to avoid “no results” pages

There are a number of ways to optimize the search engine to avoid “no results” pages:

1. Work on building your synonym library

A flexible search engine must be able to handle synonyms that capture the variety of ways visitors may refer to the same item. For instance, a company that sells sodas throughout the US must be able to handle customers from the Midwest that use the word “pop”.

Great search stems from a deep understanding of your customers. You should be able to tell some of the higher-level trends from your site search analytics.

2. Use Autocomplete and Query Suggestions

Autocomplete search and query suggestions offer recommendations and alternate queries – ones proven to have results – as the user types in the search bar. This helps the user to refine their search and potentially discover a new query, leading them to their destination faster.

Twitch search
Example of a rich autocomplete that goes beyond search suggestions

 

3. Make sure your site search is adaptable to human errors

Humans frequently misspell words and misuse punctuation. An un-optimized site search will trip over the smallest user mistakes, but great site search systems are built to handle typos and filter punctuation.

A word of caution: search engines usually correct typos by calculating the distance in characters from the most likely word match, so you’ll need to experiment with and monitor your system to make sure it is not over- or under-correcting words.

query with typo yet relevant result
Example of a query with a typo returning relevant search results

 

4. Personalize results

A user’s history provides valuable insight into what they’re likely searching for now. By using data on past searches, purchases, and self-reported interests, a search engine should be able to better understand a user’s search intent and provide contextually relevant results.

Since the results are more likely to interest users, it makes sense that many companies using Algolia’s Personalization tools are seeing increased click-through and conversion rates.

5. Customize your search based on business needs 

While many search features can be generalized across domains, there are some factors that are unique to your businesses or industries. Refine the custom relevance for your business based on business considerations, like primary KPIs for the website, as well as customer priorities, like top-selling products.

As you refine the relevance of your search to meet business needs and customer needs, searches will return more and more relevant results. 

6. Use analytics to understand customer needs

Optimizing search is an iterative process involving trial and error. Thankfully, each time a user interacts with the site, they generate lots of valuable data about what they need. Analyzing your site search can help you uncover top searches, low-performing content, and popular products, and act on that data.

 

Site Search data presented on Algolia Dashboard

 

What if there are really “no results”?

A great site search experience can turn imperfect user queries into relevant results, but the best site search tools take it further. When there is truly no relevant product or content for a query, the search engine should recommend contextually relevant and popular products or content that the user would be interested in instead of just a “no results” page.

This drives users back to relevant content as quickly as possible and learns from those missed opportunities to improve future searches.

Dig into how you can use Algolia Recommend to avoid “no results” here.

About the authorElliott Gluck

Elliott Gluck

Senior Product Marketing Manager

Recommended Articles

Powered by Algolia AI Recommendations

3 examples to help you transform the “no results” search results page
UX

3 examples to help you transform the “no results” search results page

Elliott Gluck

Elliott Gluck

Senior Product Marketing Manager
How to optimize your ecommerce site search
E-commerce

How to optimize your ecommerce site search

Louise Vollaire

Louise Vollaire

Product Marketing Manager
The (almost) ultimate guide to site search
Product

The (almost) ultimate guide to site search

Ivana Ivanovic

Ivana Ivanovic

Senior Content Strategist