In this post, learn how to make sure your long tail ecommerce search queries are relevant and context-aware.
Site Search on e-commerce sites, along with navigation and product recommendations, is one of the primary product discovery and purchase channels for online shoppers. Based on on-site search data across our top clients, conversions through site search are 2x more than the site-wide average. This signifies just how critical site search is to improve conversions and findability on an e-commerce site.
Relevancy & Context-Awareness: The Cornerstone of Effective e-commerce Site Search
Relevant site search greatly increases the chances of visitors finding the desired product in their first search. This is critical if your site goals include reducing the number of clicks required to reach the product and lowering the bounce rate, in this case, the search exit rate. Context-awareness takes relevancy a step further. Think of this as intelligent relevancy – where your site’s search can understand the visitor’s intent and show him/her results that will increase the chances of a click-through. A great example of this is a standard search like ‘linen shirt’. Here, understanding that the visitor’s object of intent is a shirt is imperative. For instance, if your store doesn’t stock linen shirts, showing a no results page or showing results for linen dresses would be counterintuitive. The correct way to handle this query is to simply return results for shirts.
The Importance of Relevancy in the Case of Long Tail Queries
With long-tail queries, relevancy becomes an even more difficult problem to solve. Some typical long tail queries on an online apparel/accessory store could be, ‘White Cotton Maxi Dress’, ‘Blue Formal Shirt’, or Charles & Keith Leather Pumps’. In case of such queries, an ineffective e-commerce site search engine would consider any/all of the words as separate search queries and the results would represent a mix of all these. Making sure long tail queries are relevant is highly critical to ensure a better user experience and lower search exit rates. Here are a couple of reasons why: Higher Purchase Intent Long tail search queries on e-commerce sites usually comprise 3+ word queries and like long tail queries on a search engine, they represent a higher purchase intent. This makes it doubly important for e-commerce site owners to get relevancy for long tail queries right. The large number of Long-Tail queries On average, long-tail queries can comprise as much as 40%-50% of all the search queries on a site. That’s a large chunk of revenue on the line for sites that don’t handle such queries effectively.
How Unbxd’s Relevant Search Provides 100% Relevancy for Long Tail Queries
Unbxd Search, our relevant site search solution offers an out-of-the-box system for handling long-tail queries. Here are a few ways that make Unbxd the perfect choice for e-commerce sites looking for a relevant site search solution. Context-Aware The Unbxd Search solution is above all, a contextually-aware site search engine. It understands user intent and is adept at showing relevant results for all kinds of queries – short and long tail. Here are a few examples detailing how Unbxd handles long tail queries on an e-commerce site. The screenshots here are from a demo site we’ve created and contain data from the top online apparel stores in the US.
In the example above, even though the site contains only 3 maxi dresses (printed or otherwise), it still shows relevant results on the top and printed dresses after that. Powerful, Real-Time Merchandizing Controls Relevancy, though essential, can be a tricky affair, especially with ever-changing business needs. For example, your site search may be relevant but your merchandisers may need to promote high-margin products or products that have a large stock. Our merchandising dashboard makes this incredibly easy. It gives merchandisers a chance to set business rules to optimize search results, something that can be done to further improve relevancy for long-tail queries. For instance, we increased the relevancy of the long tail query, ‘Cheap Dress Shoes’ by adding a filtering rule to show only dress shows below $499. Now, this is something that any site search engine wouldn’t be able to pick up (unless of course cheap is an attribute in the product data), making this a query that can be optimized using our merchandising controls.
With merchandising controls, merchandisers can set filtering/sorting rules for search results, pin products to specific places, and boost products based on attributes like brand, color, profit margin, inventory status, and virtually any other attribute imaginable (yes, you can actually specify which attributes can be considered for merchandizing rules!)Actionable Metrics Maintaining search relevancy is an ongoing task. Relevancy can always be tweaked and improved upon to consistently provide shoppers the best search experience possible. Unbxd’s Search Dashboard also includes a comprehensive insights section that shows key insights like top searches, zero result queries, the conversion funnel per query, and more.
Keeping a close watch on this can give merchandisers and product managers a keen insight into the search queries that require optimization. A great way to know which long-tail queries need optimization is to monitor the conversion funnel for long-tail queries and pick out the ones that aren’t performing adequately. Zero result queries should also be monitored to identify long-tail queries that don’t show results and fix them. What tactics do you use to correct irrelevant results for long-tail queries on your eCommerce site? Do comment and let me know!