Types of search queries ecommerce merchants need to know

The 9 types of search queries ecommerce merchants need to know are listed below

1. Exact queries
2. Product type queries
3. Symptom/problem queries
4. Non-product queries
5. Feature queries
6. Thematic queries
7. Compatibility-related queries
8. Queries with slang, abbreviations and symbols
9. Natural language queries

If you’re an ecommerce store owner, providing powerful search capabilities to your visitors is absolutely critical. When visitors land on your store for the first time, you can’t simply count on your store layout or product images to dazzle them. Instead, they’ll probably quickly search for something they’re interested in buying, and if the search results they see are unhelpful or disappointing, they’re quite likely to leave right away and try their luck on a different store (i.e. one of your competitors!). 

You might think that providing decent site search is quite straightforward: after all, for the most part, the search engine just needs to match the words in a search query with the products in your database, right? Unfortunately, it’s not that simple. 

It’s easy to underestimate the potential complexity of user search behavior, and thus fail to cater to the expectations of your store visitors. This, in turn, can lead to your brand not making much of an impression, which is bad news for your revenue. 

In fact, 84% of companies don’t actively optimize their site search. You need to make sure you don’t join their ranks.

One important aspect of customer search behavior that you should have some knowledge of is the various kinds of search queries site visitors typically use to get what they want. Moreover, you need to make sure that your site search engine can handle as many of these query types as possible. So let’s dive into these query types.

9 types of search queries your site search engine needs to handle

1. Exact queries

This kind of query is typically used by searchers who know exactly what they’re looking for: e.g. ‘toblerone milk chocolate large bar’. On the face of it, it might seem that handling such queries involves simple keyword matching. However, for exact search queries to be truly handled well, there are some additional tricks that search engines need to have up their sleeves.

For instance, searchers might mistype a name as they might only have heard it but never seen it spelled. In addition, many products can have different names in different countries, so an online store that caters to a global audience needs to take that into account. Searchers often also copy-paste model numbers into stores and expect to find the corresponding product, so search engines need to be able to handle such scenarios as well.

2. Product type queries

This is also a very common query type, wherein people look not for a specific product but a particular category, such as ‘laptops’ or ‘carpets’.

There are a few important aspects that a site search engine needs to consider here. For one thing, if a category page already exists for the category a user is searching for, then the search engine should simply redirect them to that page. 

Secondly, the category a user is looking for may be available on your store, but under a different name. For instance, your store might have a category called ‘pants’ but not ‘trousers’ or ‘slacks’. Thus, if a store visitor were to search for the latter two terms, they might come up empty-handed. To deal with such a situation, the search engine needs to know that ‘pants’, ‘trousers’, and ‘slacks’ are actually synonyms. This can be accomplished through creating such lists manually, or, if the search engine you’re using is based on artificial intelligence (AI) and machine learning (ML), it might come with such “knowledge” right out of the box.

In addition, the fact that the user is searching for a category means that ideally, the search results page should provide some filters and facets tailored to that category.

3. Symptom/problem queries

Store visitors may sometimes search for a problem that they have, in the hope that the store can provide them with a product that solves their problem. For instance, a user might search for ‘dirty keyboard’ or ‘wine on dress’.

Symptom-based queries are particularly important and common in certain industries, such as medicine, hardware and DIY, cosmetics, housekeeping, etc.

A common way to help users who might have symptom-based queries is to guide their query formulation with the help of scope suggestions. Scope suggestions can help point visitors to potential categories that might contain products that might help them solve their problem.

4. Non-product queries

As the name suggests, these are queries that are typed in because a user is looking not for a product, but for some other kind of information, such as your store’s returns and refunds policy, or a customer support number. 

Thus, in effect, customers nowadays expect site search to fetch results not just from the products database, but from all over the site. Your site search engine needs to take this expectation into account.

5. Feature queries

These are queries that indicate not just a particular product but also some particular attributes or features of that product. These features can be of several different kinds, including brand, color, price, size, etc. An example query of this type would be: ‘size 10 nike sneakers’.

A good way to handle such queries is to transform each feature into a filter on the search results page. For instance, the example query given above could lead to a results page for ‘sneakers’ with the ‘Brand’ filter already set at the value ‘Nike’ and the ‘Size’ filter already set to ‘10’.

6. Thematic queries

Such queries involve concepts that can be somewhat fuzzy or subjective. For instance, when a user looks for ‘summer dress’, what constitutes a summer dress might be somewhat vague, and yet a search engine should ideally return all relevant results, not just results that have ‘summer dress’ in their title or description.

A common way to handle such queries is to use ‘thematic tagging’ for each product: this refers to manually adding tags to products that indicate their thematic categories. A much better alternative is to use a search engine that uses artificial intelligence (AI) and machine learning (ML) to automatically “understand” the thematic categories of various products.

7. Compatibility-related queries

An example of such a query is ‘iphone 14 usb charger’. Here, you’re looking for an accessory that is compatible with a different product that the user probably already has. 

However, while users might sometimes indicate the exact product that they already have, they might sometimes use a generic product category instead (e.g. ‘smartphone usb charger’), or might even leave out such a product altogether (e.g. ‘usb charger’).

One way that your site search engine could handle such queries is by displaying the search results for the “main” product with some kind of filter toggled on for accessories or related products.

8. Queries with slang, abbreviations and symbols

Some examples of such queries are: 

• using colloquial terms like ‘shades’ to mean ‘sunglasses’ 
• using abbreviations of measure units, such as ‘ft’ and ‘in’, and 
• using symbols such as ‘-’ or ‘%’, as in ‘sweater for -10 degrees’ or ‘4% rubbing alcohol solution’.

While slang and abbreviations can be handled relatively easily through synonym mapping, symbols can be trickier to handle, as a single symbol can have different meanings in different contexts. 

For instance, the symbol ‘-’ can be interpreted as a minus sign (as in the example given above) or as a range indicator (as in ‘clothes for babies 3-5 kg’). Thus, your site search engine will have to be able to properly interpret certain symbols based on their context to deliver relevant results. And yet, 49% of sites don’t have site search that supports symbols and abbreviations. 

9. Natural language queries

These are queries that are phrased in a way similar to what you would say in natural speech, such as ‘real leather shoes for men that don’t have a brand name written on them’.

Such queries cannot be handled well by search engines that rely primarily or exclusively on string matching. Instead, you’ll need a search engine that makes use of Natural Language Processing (NLP) to “understand” and then suitably process such queries.

Such NLP-based search engines rely on various bits of semantic and contextual information to deduce the search intent behind a query. Moreover, they can handle not just phrases and statements but also questions (e.g. ‘Which laptop is great for gaming?’), thus greatly reducing the cognitive load on store visitors.

Search query types: handle them better to keep customers hooked

While search might seem uncomplicated from the outside, there is more nuance to it than meets the eye. Online business owners looking to provide their site visitors with the best experience possible, and who are hence shopping around for a site search engine that meets their requirements, need to look into the ability of such search engines to handle various kinds of search queries effectively. Remember: conversion rates for people who use site search are up to 50% greater than the average.
‍

If you wish to provide your customers with a powerful AI-based site search engine that is blazingly fast, provides highly relevant search results and handles a wide variety of search query types, you might want to try Zevi. By harnessing the power of artificial intelligence (AI) and machine learning (ML) under the hood, Zevi can understand the search intent of your customers, even for naturally phrased  queries, and offer the search results most relevant to your business needs. Install our Shopify app now, or try Zevi out for free today!