Algolia acquires »

Read More

How to Improve Shopify Conversion Rates with Better Search and Discovery

It’s no surprise that the events of 2020 accelerated e-commerce adoption. In 2021, this trend is expected to continue. This increase in adoption and competition also means increased pressure on retailers to retain and increase conversion rates on their websites.

In e-commerce, the conversion rate is the number of purchases divided by the total number of sessions. And the vast majority of visitors will take more than one session to decide on a purchase. At every step of their journey, you have opportunities to make their path to conversion more straightforward and enjoyable.

According to Littledata, the average conversion rate for a Shopify store is 1.6%, and anything above 3.6% would put a store in the top 20% worldwide. Additionally, the average mobile conversion rate should be closer to 2.2%.

Overall, we can think about conversion rate optimization as an exercise in reducing friction for the customer. Things that can add friction to the customer journey include:

  • Slow performance
  • Lack of personalization
  • Poor site structure
  • Poor product data
  • Responsiveness
  • Difficulty finding products

An excellent e-commerce experience starts with an outstanding search experience. This article will go over concepts such as autocomplete and typo tolerance, which are essential for a good search experience and can increase conversions substantially. We’ll also discuss the advanced features and functionality that offers and how they can be leveraged to improve the overall user journey, enhance search capabilities, improve product discovery, and drastically improve your store conversion rate.

Finally, we’ll include examples using the relevance settings for configuring search results and in the yaml configuration for each feature added to the query pipeline. Keep in mind that automatically generates initial pipelines for you, which you can modify or append later on via their built-in pipeline editor.'s query pipelines define how the queries work when searching your collection records and how they’ll rank your results. Steps in a query pipeline can improve query understanding by filtering results, changing relevance, and allowing for advanced queries. offers a no-code relevance setting too for non-developers to make the same adjustments quickly and easily.


One of the first ways you can reduce friction for visitors to your online shop is to implement autocomplete in your store search. This feature allows customers to find relevant results without typing out every single character of their query. uses machine learning to adapt and deliver the best possible search results on every search.

Autocomplete example in a search

In the example above, the search is already starting to suggest recommendations after typing the first two characters of my search. Additionally, visual search (ie, displaying and clicking product search results) directly reduces the customer journey and increases conversion chances.

Autocomplete is included with by default. combines product data from your Shopify store with customer search behavior to provide autocomplete results.

Typo Tolerance

Autocomplete alone is not enough to make site search an effective tool for driving conversion. A smart search solution also has to handle misspellings and user errors; otherwise, you run the risk of users leaving the website thinking the product they want is not offered or not in stock. Typo tolerance helps customers get accurate search results even when they don’t spell words correctly.

An example of typo tolerance

As you can see from the example above,'s search solution is smart enough to realize the user has made a typo and likely meant sneakers. Like autocomplete, spell-check is included out of the box.

This is where is unlike other search solutions. looks at not only each word, but also how the words fit together as a whole and how they match against your shop’s catalog.

For example, try to search for gray had. If everything works correctly, you should get a list of all the grey hats in the store.

Grey hat search results

The ability to handle typos and different spellings drastically reduces friction, improves customer experience, and significantly increases the chances that users will convert in a single session.

So far, you’ve seen how having your search integration handle user input and typos—as well as make recommendations—can be a significant lift to conversion. Next, let’s look at how you can leverage machine learning to provide more advanced features.

Relevance Editing

Relevance is the name of the game for search results. If they’re not relevant, buyers will look elsewhere—simple enough! Except that relevance is tough. Search engines need to parse your product title, descriptions, headings, category, tags, reviews, and more to deliver the best results.

You can help steer more relevant results with the new visual relevance editor; it’s the same as pipelines, but available for non-technical users to adjust relevance settings.

You may want to direct the search engine to put more weight on the title, description field, and publish date. By creating broad rules, you can train the algorithm to deliver better results for your users. You can even create multiple relevance settings and then A/B test them to determine which algorithm provides the most conversions.

Natural Language Processing (NLP)

”Natural language search is a search carried out in everyday language, phrasing questions as you would ask them if you were talking to someone. These queries can be typed into a search engine, spoken aloud with voice search, or posed as a question to a digital assistant such as Siri or Cortana.”—Search Engine Watch

The core idea behind natural language processing (NLP) is that it allows the search to replicate the experience of talking to an associate by phrasing the search query as you would phrase a question. For example, consider a typical, single-word search:

Regular sneaker search

But with NLP, you can specify a color or any other attribute in our query. If you type white sneakers, you should see different results.

White sneakers search NLP capabilities leverage Shopify Product metadata to provide NLP capabilities. Common examples of metadata fields include price, size, color, and brand.

Diagram of an NLP search
Use natural language processing (NLP) to improve the Shopify search experience.

Internally, as part of a query pipeline, a technique called query scoping helps convert an ordinary query into a structured one. For example, in the previous diagram, you can see that mens size 14 nikes under $75 isn’t being processed as a normal query. Instead, it’s extracted as follows:

  • mens is matched to a product property gender
  • size 14 is interpreted as a product property
  • under $75 is matched as a price filter
  • nikes remains as root query

As a result, the customer searches for just nikes against product information, but query scoping automatically filters the search results against size, price, and gender.

Here’s an example of how the price filtering is set up as part of the pipeline:

The above example looks for a price under a certain dollar amount, which then removes this text from the query and constructs a filter to ensure the price is under the limit. But equally, this can be anything else: colors, sizes, dimensions, and so on.

NLP dramatically increases precision and allows customers to have a quicker path to products relevant to their searches.


Personalized search delivers web search results tailored specifically to a user’s interests by incorporating information about the individual beyond the specific query provided. There are two general approaches to personalizing search results: modifying the user’s query, and re-ranking search results. – Wikipedia

Appropriately leveraged, personalization can be an extremely powerful and effective feature to increase conversion. For example, if you know a customer has purchased Nike branded products in the past, you can recommend additional Nike products on their next visit.

Personalization creates profiles for each visitor and allows you to deliver customer-relevant search results. A few common parameters that could be used for personalization include the following:

  • Location
  • Gender
  • Past purchase history
  • Preferred product categories
  • Customer behavior as they navigate the site

In, you can adjust the relevance settings to personalize results. Filter boosts, combined with conditions, provide a powerful tool to adjust the ranking of your search results. For example, if you know a customer’s brand affinity based on past purchases, you can create a variable called brandAffinity that you can pass into the search query to improve results.

Thankfully, there are many other options to automatically tag customers and orders in the Shopify ecosystem for personalizing results. One well-reviewed tool is TagRobot. Using this solution, you could:

  • Tag customers Price Conscious if they paid with a coupon or a gift card
  • Tag customers with their respective country based on the destination
  • Tag customers as VIP if they place X amount of orders or above a certain monetary threshold

You can then leverage customer tag data as part of your query pipeline to improve result relevance and increase conversion.

Dynamic Boosts

“Boost rules allow you to increase or decrease the importance of specific content in your search results.” – documentation

So far, we’ve covered several ways to leverage search and to improve search relevance and accuracy and to increase conversion. However, in some cases, you might want to have a bit more control over specific search terms and boost certain products or collections.

There are several types of boost steps you can apply to your query; one of the more powerful ones is dynamic-boost. This queries historical performance and dynamically boosts the highest performing records for each specific query.

Use dynamic boosts to adjust search rankings.

A dynamic boost can be as simple as boosting products that have been clicked at least twenty-five times as part of the search results, or as advanced as boosting products based on order data. There are several other types of boost that are available to you, such as:

  • add-conditional-boost, which boosts records that have the specified text present
  • add-multi-range-boost, which boosts records that match any of the specific ranges by a specific amount
  • filter-boost, which boosts records if the giving filter matches a specific condition
  • geo-boost, which boosts a result based on its proximity to geolocation

In essence, you can promote items that sell more frequently, have inventory in stock, have higher user ratings, or practically anything else!

Finally, for most boosting and exclusion rules, you don’t need to touch the pipeline code; gives the option to create ranking rules directly from the dashboard.

Sajari ranking rules

Out-of-Stock Notices

Now that you’ve tweaked and optimized your search for relevance and increased customer experience, there’s a final challenge—and it’s the bane of every online shopper.

When thinking about the online shopper experience and the flow through the website, a crucial thing to consider is inventory status and product availability. All our efforts to improve the search experience are for naught if their search results are out of stock or unavailable.

This not only causes frustration for the customer on that individual transaction, but might cause loss of loyalty. The disappointed customer may never return.

A vital piece of a great search experience is filtering unavailable products from the search results. It also offers an opportunity to present available products more prominently, increasing the chance to convert that customer to a successful sale.

Since directly integrates with your Shopify store, you can leverage inventory data to boost conversion in two ways:

  • Boosting low-inventory products to the top of the search results and adding accurate stock counts in line with the product can trigger the customer’s [fear of missing out FOMO and create an impetus for them to convert. A prime example of this is Amazon’s website, which showcases low stock for products.
  • Boosting high inventory products, in contrast, can leverage product inventory data to promote slower-moving products in specific categories. With a search solution like, you can use a combination of stock levels and discount percentages to influence how your product search results are ranked.


As you can see, site search is not something you want to overlook. In fact, it’s one of the most valuable tools you can have to boost conversion rates and increase customer satisfaction during interactions with your store.

With, not only can you provide your visitors with high-quality search results, but you can also harness a robust series of features and capabilities to build best-in-class customer navigation with detailed analytics, built-in marketing, and personalization capabilities. In short, advanced and accurate search means a smoother customer journey and drastically higher chances of conversion.

Special Offer: If you want to start employing some or all of the features discussed in this article on your site search. Sign up on Shopify to try by May 30 to get 6-months free on our Starter plan for Shopify.

Guest contributor Allan MacGregor is a software engineer and entrepreneur based in Toronto, with experience in building projects and developing innovative solutions.

Similar articles

What's New Can Power Your Shopify Search

Best Practice

How to Optimize Product Browsing and Discovery

Best Practice

How Grizzly Industrial Boosted Revenue with Better Search & Discovery