Skip Navigation

Things to Consider When Planning Your Search Implementation for Sitecore

Resources

Read Time: 10 Minutes

You’ve come to the conclusion that your search experience sucks. People aren’t finding what they need, your conversions are low, and your loyal customers are complaining. 

Think about the money you’re leaving on the table. The people who don’t even care enough to complain are walking away and spending their money with your competitors, who make it easy for them to find what they’re looking for. 

Determining what kind of search experience is going to set you apart and create loyal customers is step one. Step two is a very thorough planning process that ensures your Sitecore search implementation is a success. 

What you need is a detailed plan that outlines the technical functionality, UX and design elements, and Sitecore solution requirements.

We’ve written this article to walk you through that planning process.

Visualizing Your Sitecore Search UX and Design

Many people make the mistake of entering a search project by feature-shopping. Don’t be one of those people. 

Instead of choosing the search software that boasts the most features, you first need to figure out what kind of search your users need. Then you can go shopping for the solution that offers that functionality, without paying for extras you’ll never use.

To define these needs, start by thinking about what role search plays on your website. Is it ubiquitous to the site experience, or will it just be helping some people get to specific documents? In other words, is search a nice to have, or a critical navigation tool?

Next, think about what role can search play? This requires a UX designer. In an ideal world, how would people use search to navigate and more deeply engage with your website. 

At Engagency, we have a 50+ point checklist of design and technical elements that we use to help our clients evaluate their search needs. 

This includes questions like:

  • What should the search box look like? Where should it “live” on your site? Should it be visible on both desktop and mobile? Does clicking the Enter key submit the search?

  • What should the search results look like? Do you want them displayed in a list or pictorial format? Would it be helpful to use visual cues like icons to denote different types of content? Should the search terms the user entered be highlighted in the results? How many results will you include on the first page?

  • What’s eligible for search? Will everything on your site be searchable? Think beyond pages: what about documents in the Sitecore media library or your digital asset management (DAM) system?

  • How is content weighted for relevance? In other words, how do you prioritize different types of content according to the keywords a person entered? Have the relationships between your products and service pages been mapped to other content types (whitepapers, blog articles, etc.) to make that possible? 

  • How does search support other aspects of your site? Search extends beyond the obvious keyword search. The right search platform can power directories, resource libraries, news and blog sections, and more. 

Once you have the answers to these questions, you can begin to look at the technical features that will power them.

Exploring Possibilities and Choosing Your Search Functionality Options

Now that you know what your users need, it’s time to acquaint yourself with how search functions so you can choose the best options. 

Here are the most popular search features and functions used online today.

1. Faceted search

This is the technical name for search results that allow you to narrow down the results by using filters. Faceted search is great when you have a huge repository of information with many different, well-defined attributes, like product catalogs. 

For example, a clothing eCommerce site may create search facets for color, size, length, and product rating.

2. Type-ahead search

Type-ahead search predicts what a user is going to type based on previous searches, or a curated list of applicable keywords. Type-ahead suggestions help people discover the keywords they wouldn’t have thought of on their own, or just save them time by filling the rest of the phrase in for them. 

It’s great for searching documents and resource libraries with a lot of text-based content.

The type-ahead bar can take on a life of its own by going beyond suggested keyword terms to actual promoted content, based on keywords or on some other metric associated with pages that give them additional weight.

3. Natural language search

Natural language search takes type-ahead search to the next level. It goes beyond predicting the words a person is using, and recognizes the intention behind their search—based on the words they use and the action they are most likely to take. 

For example, for a health insurance site, natural language search would be able to infer the appropriate semantic meaning and determine whether a person is trying to contact a benefits advocate vs. look up the available benefits in their plan.

Natural language search is a no-brainer for sites where a user can take multiple actions (contacting you, finding a location, comparing products, etc.). Each of these actions has a clear verb attached to the search, and natural language assumes that verb on behalf of users, delivering the information they need faster and in a more precise fashion. 

We’ve found banks, healthcare organizations, insurance providers, and software vendors most often benefit from natural language search.

Which search functionality is right for you?

As you can see, there’s a lot you can do with search, but you don’t need the search on your site to do all of it. Introduce too many features to your search, and things can quickly lean into overkill territory. For instance, faceting seems like a great idea, but you can overload your users with filter options (just take a look at Amazon).

When choosing your search functionality, remember this: You don’t need all possible features of search. You just need the essential features your particular users need. 

Choosing the Right Search Software for Sitecore

Once you’ve defined the search experience you want, it’s time to go shopping for the appropriate solution.

The two most popular search solutions for Sitecore are Azure search and Solr search. They’re functionally equivalent and Sitecore interfaces with both, so you’ll get the same features either way. Which one you choose depends on what makes the most sense for your Sitecore setup and server environment. 

Azure search

As with most things Azure, Azure search is easier to manage. You spin it up and it’s there. You don’t have to deal with the underlying VMs. 

However, that ease of use comes with a hefty price tag. The cost of Azure search is based on how many resources you use—so if customers are using search on your site at a high volume, this gets expensive fast. 

Like Azure PaaS, Azure Search is fully managed, which is why many companies are willing to pay the higher price. There are no updates or maintenance to worry about. You simply turn it on and it works.

However, Azure Search has some basic limitations that are surprising given the cost. For example, we've run into limitations with Azure search in terms of providing the ability to make individual content items have search weighting assigned. This means that content editors cannot directly influence a single piece of content to force it to the top of the list. We've also found that more complex filtering and faceting can be done with Solr as you have more complete control over the data.

Solr search

Solr gives you the opportunity to build any search experience you want. But the true value of Solr search is that it’s an enterprise-class search engine that solves for the shortcomings of Azure Search—at an open-source price. 

You can install Solr search on your own physical or cloud environment, in which case you’ll have VMs to manage. This is a more expensive approach, as you’ll have to pay for both the direct and indirect management costs, like replacing physical hardware and paying internal IT staff to maintain it and keep things up to date.

The smarter option is to take a SaaS approach by going with a third-party managed Solr as a Service provider like SearchStax, WebSolr, OpenSolr, and others. This makes things a lot easier and cheaper.

Coveo

A discussion about Sitecore search wouldn’t be complete without mentioning Coveo, another search software for Sitecore that leverages AI to take a lot of the thinking and planning out of the search implementation process. 

If you’ve got a product-driven website like an eCommerce site, Coveo’s AI-powered search experience could potentially be a very good fit. It allows you to flip a switch and have an intelligent search engine helping people find what they’re looking for without giving the planning process much thought. 

But, Coveo is very out of the box. You’re not going to be able to make customizations with the kind of granularity you can with a tool like Solr. Also if budget is a concern, you’ll want to consider other options. 

Which search software makes the most sense for Sitecore sites?

Of the three options, fully managed Solr search provides the best balance of cost and ease of management. While Azure search and Solr search are functionally equivalent, Solr’s got a leg up on Azure search lacks some basic functionality that you would expect from an enterprise search engine. This makes fully-managed Solr the best value, because it is scalable according to your needs.

Implementing Search for Your Sitecore Website

There are a million exciting things happening in search. For your site search to be exciting—for both your users and your CFO—you need it to provide the features everyone expects for the best cost.

Don’t just pick the trendy buzzword features because they’re cool. Instead, really ask yourself—and work with the professionals who can help guide the way—which search features are essential for your site’s unique user experience. 

Define these needs with your UX agency, and then select the software that provides that functionality.

At Engagency, we’ve led search implementations for numerous Sitecore customers. This always starts with a roadmap, where we review the latest search best practices with our clients, and evaluate the importance of each feature according to their users’ needs. Then, based on these discussions, we put together a presentation summarizing our findings and recommendations for the search design, technical functionality, and finally, the infrastructure requirements for the production instance.

Interested in what search can do for your Sitecore website? Contact Engagency today.
download
Things to Consider When Planning Your Search Implementation for Sitecore

Download this Article

Download a PDF copy of this article to share with your team! Download Now

Related Content