Product type searches
First things first, when optimizing your site search product type searches should be one of the first UX improvements you consider. A product type search is when a user queries for a whole category of products like “sweatshirts” instead of a specific product. This is great for users that are just browsing and don’t have a product in mind or who want to access areas of your site quickly. Or, users might not be able to find what they need from your navigation menu, turning to search. If the latter is the case it’s important to make sure your search returns relevant results.
This can mean lots of detailed categorization and labeling of products in your back end as well as incorporating synonyms and alternative spellings of groups (we’ll get to this later). However, investing time can make the difference in creating an excellent user experience and securing those all-important sales.
Once you’ve got the basics like product type search narrowed down, there’s a whole host of search options available depending on the user experience you want to create…
If you’re a brand with products focused on problem-solving, symptom searches are the feature for you. Using symptom searches your customers are able to enter the problem they have and be presented with products that solve the issue. For example, a user navigating a skincare site might enter “acne scarring”. By incorporating symptom searches the user will see products that will help solve their issues such as serums, creams, and cleansers instead of being presented with “unable to find the product”. Google has this nailed down and so users will come to expect this feature on e-commerce sites, especially as a symptom search is often used when a user doesn’t know where to begin or might be their last point call.
Take Dr.Sam's, as a skincare brand their customers will be looking for and value the ability to symptom search, which is why we used search solution - Boost Commerce (we'll chat about them in a bit!).
However, this often returns multiple and varied product types which can make it hard for the user to find the right product. Consider incorporating search suggestions to remove this issue, it compliments symptom search by guiding the user to a related query path, narrowing the product suggestions, and making a slicker user experience.
Search suggestions as you type
‘Search suggestions as you type’ is a popular feature used by search giants Google and Amazon that involves dynamically updating the suggestions in the search dropdown as the user types. It provides a slick, intuitive experience right from the user’s very first interaction, and allows them to avoid typing out their query in full before searching.
In e-commerce, this feature also acts as an early-stage, real-time filtering system that can carefully point customers to related products, collections, or articles. In fact, you can take this a step further by showing customers products in the search dropdown, including information and imagery, associated with their query. For example, if a user types “black dr” the search dropdown would display a few variations of black dresses sold on your site. This is a great experience for customers who don’t know what they’re searching for - plus it’s a blessing for increasing your traffic to certain products or content.
Algolia does this brilliantly, with auto-complete functionality and widgets that can be pieced together quickly to build a search interface. Once the merchant has this setup, they can populate the results that are delivered to the user as they type. Aloglia’s solution is so powerful for filtering, large, content-heavy documentation websites such as React Native and Tailwind CSS use it to sift through their vast, complex information instantly.
We used Algolia to integrate search suggestions as you type on British Fashion brand, Jigsaw’s Shopify Plus store:
Impressively, Boost Commerce ‘search suggestions as you type’ comes out of the box via their Instant Search functionality. Merchants can control search functionality through the ‘Search’ section of Boost Commerce’s app and adjust the layout by selecting between a one-column, two-column, or full-width dropdown. Merchants also have the ability to adjust the number of products shown in the dropdown and whether those products show in a grid format or list. Enabling them to create a search functionality that delights their users whilst staying true to their brand.
Additionally, Boost Commerce allows you to adjust what’s shown in the dropdown before users type a query - giving you the opportunity to point users to certain products.
Similarly, Klevu’s ‘Smart Search technology’ creates a lightning-fast predictive search function that changes and updates the results with every letter typed by the user. As for the styling, Klevu allows merchants to style their search dropdown using Klevu’s out-of-the-box templates or by creating their own custom style.
One of the most fundamental tasks of any e-commerce team or business is to understand their customers - what do they want right now, and what might they want in the future? What problem are they trying to solve, or what goal are they trying to accomplish? And how can it be made as simple and easy as possible for them to do so? All of this would be a lot easier if you could ask every customer that comes to your online store what they’re looking for. You could rack up sales and improve customer retention, as well as optimize your spending by making better buying decisions.
That’s exactly what search analytics gives you, and Algolia, Boost Commerce, and Klevu provide some of the best in the business.
Algolia contains a powerful suite of analytics right out of the box. They automatically capture the data, with no development setup required and no impact on search speed. Stored on their admin dashboard, merchants are walked through each metric including: measuring the total number of searches, the total number of ‘no results’ responses, top searches, top results, and highest usage by country - just to name a few.
Showing the number of ‘no results’ responses allows merchants to see what products customers are searching for that aren’t currently stocked - informing inventory for new or re-stock. Couple this with an email campaign on the new or re-stocked products and merchants have an opportunity to win back some customers that were lost to a poor experience. Don’t worry we’ll cover more of this in a bit.
Similar to Algolia, Boost Commerce measure and present a range of metrics including total searches, top search terms, and searches with no results.
Next up is Klevu, which offers merchants a real-time view of their analytics enabling them to make instant optimizations during peak times and big sales days. With the ‘live view’ merchants are able to see searches as they happen, as well as the top searches and products at the last minute. And, of course, all of this data is available over longer time periods for more generalized insights.
They’ve also included a handy dynamic word graph so you can search a word to see what terms are used alongside it in customer queries - information like this can be used to create more engaging and useful product descriptions.
This search feature is perfect for brands going international or wanting to elevate their existing global stores. By allowing you to link words such as “pants” and “trousers”, customers receive the same results regardless of the differences in languages. With third-party solutions you can create multiple synonyms:
With Algolia, you can create:
- Regular synonyms - These are two words or a group of words that can be used interchangeably to get the same search results like “pants” and “trousers”. This is the most common category of synonyms that a merchant is likely to use.
- One-way synonyms - As the name suggests, this synonym is similar to the above but only applies one way. For example, a search for ‘laptop’ should display all MacBooks, but a search for ‘MacBook’ shouldn’t display all laptops - this would be frustrating to a user as they searched for a very specific kind of laptop.
- Alternative corrections - Similar to one-way synonyms but items that are an exact match to the query typed are displayed higher in the search results.
- Placeholders - Placeholders allow merchants to display the same results for a number of searches that only differ by one small parameter. For example, a brand selling phone cases might want the search ‘iPhone 7’ to return the same results as ‘iPhone 8’ as a case for both devices will fit the other.
Klevu and Boost Commerce
Klevu and Boost Commerce also allow merchants to create regular synonyms and one-way synonyms which they call ‘Bi-direction synonyms’ and ‘Uni-direction synonyms’ (just to keep you on your toes!).
Klevu makes use of Natural Language Processing - a division of AI that enables the interpretation of human language data throughout their search tool. Klevu combs through product catalog data from the store and adds any relevant linguistic and semantic information. For example, if a product title includes the words “black trousers”, Klevu automatically understands that “black” is the color, and “trousers” is the product type. It then builds out the data using various synonyms and international spellings, enriching the search coverage and adding synonyms for the merchant.
Check out Pangaia's use of search synonyms, appealing to their global customer base:
Next up we’re diving into Typo tolerance. Misspelling, although a common practice for most isn’t always considered by major e-com brands. Baymard states that 27% of major e-com sites don’t provide useful results if a user misspells even a single character - which means missed conversion opportunities.
Typo tolerance is how merchants can ensure their customers who make spelling mistakes when searching can still find the results they need. And supports site accessibility, as some users will find typing and spelling more challenging than others.
Algolia’s ‘Typo Tolerance’ works based on ‘distance’ - the number of letters that differ between the search query and a particular match in the index. For example, a search for “foat” would have a distance of 1 from the word “coat”, so coats would be displayed to the user.
Depending on the length of the search, Algolia allows you a tolerance of 1 or 2 letters that the query can differ from a matched result - longer searches are given a higher tolerance. Using this, Algolia displays exact matches to the search shown first, followed by words with a difference of 1 letter, then words with a difference of 2 letters.
With Boost Commerce, the typo tolerance for a single-word search is built-in. For example, a user navigating a clothing store site that searches for “shift” will be shown shirts on the results page. For two and three-word searches, the tolerance is customizable via the app to offer greater control.
Klevu also has a built-in tolerance. This is a powerful tool as it’s combined with their automated product catalog keyword enhancement, meaning the misspelled words are compared against a much broader selection of possible matches.
We used Klevu to integrate typo tolerance to Skinnydip's search feature for a slick and inclusive customer experience:
Other Search Features to consider
Sitting just outside of product search is ‘non-product search’, it might not seem a priority but introducing this to your Shopify Plus store will take your user experience to the next level. Non-product search includes information like “return policy”, “unsubscribe” and “cancel my order” among others.
Baymard observes that “34 % of users are now expecting the search field on an e-commerce store to search the entire website”, similar to the experience they have on Google and social media. It’s also likely that users search for non-product information as this isn’t readily available or easy to find on a desktop or mobile site. In this case, it’s even more important that users can and do receive accurate and helpful results to alleviate their frustrations and solve their problems.
With all of the many search features available, we’re hoping ‘no results’ would be a rare occurrence. However, there are still times when a user will search and get no results for their query. So what then? It’s key to think of an alternative path for the user by building out a clear ‘no results’ page that gives them tips on how to search again – with a persisted search field so they can see what they searched for and are able to refine it without re-typing. You could also include a list of most popular search terms or a list of best-selling products or collections. All of these enable a user to easily navigate their way to another part of the site without much effort on their part.
It may not be the sexiest e-commerce feature, but it’s clear that site search is becoming the ‘go-to’ first step for users visiting a website and one that can easily turn them away. With that said, Baymard suggests as it stands that “e-commerce search isn’t as easy as it should be” - even the major players in e-com need to make progress.
The good news is this means there’s an opportunity to rise above the rest. Using the query types we’ve taken you through, search can have a transformative impact on customers and merchants: cutting through friction in the checkout journey, improving conversion rates, connecting customers to products they want or information they need, and creating a truly delightful experience.
If you want to take your site search up a notch or need help deciding which third party is the one for you, reach out to our team.