Skip to content

Anatomy of an e-commerce search query

Anatomy of an e-commerce search query

Users combine 12 query types mapped into 3 groups: spectrum, qualifiers and structure.

  • Query spectrum: base of the search query. This is the foundation of the user’s query.
  • Query qualifiers: they are used to refine the boundaries of the query spectrum.
  • Query structure: the structure helps explain how the query should be interpreted.

Spectrum, qualifiers and structure help design search logic that aligns with user behavior and expectations.

Query Spectrum: setting the ranger

The query spectrum is used to indicate the range of what should be searched. Here are the types of searches that belong in the query spectrum:

  • Exact search
  • Product type search
  • Symptom search
  • Non product-related search

Exact Search Queries

Exact search queries contain the product title or the product number. You should handle phonetic mistakes and products that have alternative titles. The logic should search the entire data set to broaden the query’s scope. Why? Well because some people know products so well that they search for them by SKU number for example. If onsite search tools do not pan all the way to the SKU field, then, nothing would be returned. Exact search is the simplest query type in e-commerce.

Product Type Search Queries

Product type queries are used when the user knows the type of product they want but not which particular product they seek to buy. To handle this type of on-site search, you should have detailed categorization and product labels in place. You should also handle synonyms and alternate spellings within those product groupings.

Symptom Searches

Users aren’t always looking for a product. Oftentimes, they are looking to solve specific problems and want a product that helps them take care of their problems. A good example would be someone looking to remove a stain from their carpet. That person isn’t looking for a cleaning product, they are looking for something to make the stain go away.

Non-Product Searches on E-commerce Site Search

Site search engines should also handle auxiliary content search like that as often users will have a hard time finding these in the navigational links. What this means is that your search should go beyond product pages. Oftentimes, people want specific information that assists their decision purchase. This type of information isn’t often present in the first level navigation menu or the footer. Hence, many visitors will use your internal search engine to find pages.

Query Qualifiers to Delineate the Boundaries of the User’s Search

Query qualifiers are conditions that the users want to see included or excluded in their queries. These qualifiers fall under 5 big categories:

  • Feature search
  • Thematic search
  • Relational search
  • Compatibility search
  • Subjectivity search


Thematic, compatibility and subjective searches are a little more challenging from a technical perspective, but they are often used by users searching for things. They should be thought about within an internal search project.

Feature Search Queries

Feature searches are the most common qualifier used by customers. Here’s how to define a feature search qualifier: it is any type of product aspect or attribute. This includes:

  • Color
  • Material
  • Performance specs
  • Format
  • Price
  • Brand

Thematic Search Queries

This type of search is a common browsing pattern and product arrangement. Store have displays and thematic aisles so customers look for these online as well. Here are some common themes:

  • Seasons…let’s not forget pumpkin spice season (most people call it fall), summer BBQs, cozy winters and joyful spring displays!
  • Intended usage (outdoors, office, etc.)
  • Occasions (birthdays, weddings, divorce parties…)
  • Events (Olympics, Superbowl, Thanksgiving)

Good product categorization and labelling will get you halfway there when it comes to handling these types of searches.

Relational Search Queries

Relational searches are searches where users enter the name of entities involved with or related to the product. Ever wonder all the movies Tom Hanks has been in? That’s a relational search! Amazon is awesome at handling these.

Compatibility Search Queries

Users often don’t know the name of the accessory or spare part they need. Instead, they focus on the details of the product that they already own. A common structure for a compatibility search would be: name of brand + type of accessory or spare required. One of the most common compatibility search queries we all know is ink cartridges for our printers.

Subjective Search Queries

Subjective qualifiers such as “high-quality” or “cheap” are often vital to the customer’s purchase decision. Except that no search engine is equipped to decide on its own what should be considered high quality or cheap in a product catalog.


Here are some tips to get started:

  • Approximate user intent by using one or more attributes as proxy. Maybe you already have a label in your internal taxonomy to help you decide what cheap is? Use that label as a proxy. Another way you can handle the “cheap” query is by offering filters for items under price ($5, $10, $100).
  • Identify an attribute that could serve as a useful proxy. Deconstruct what the query is about and offer a proper way to handle these subjective searches.

Query Structure : Constructing the Search Query

Query structure deals with how the query is constructed by the user. The query structure takes into account the context, the syntax and the search interpretation. Here are the big 3 things you need to handle query structure properly with your internal search engine:

  • Slang, abbreviations and symbol search
  • Implicit search
  • Natural language search

Implicit Searches

Certain aspects of search queries are left out by visitors because it’s part of their obvious context. Detecting implied components of a search query can alter the search experience quite a lot. Think about the following things to handle implicit searches:

  • Bias: this is what happens when a search is done from a category. If I am in cocktail dresses, don’t offer me ALL of the black dresses, show me the black cocktail dresses.
  • Suggest relevant search scopes or query clarifications. This is often seen in the “did you mean…” messages we get under search bars as users.
  • Auto-refine and auto-correct to include implied components as you go.

Natural Language Searches

It’s all about understanding semantics, context and relationships of the query instead of parsing the query as a set of keywords.

Giving Excellent Search Improving Your Onsite Search

You should handle as a priority at least 5 query types in your internal search engine:

  1. Exact search
  2. Product type search
  3. Feature type search
  4. Thematic searches
  5. Relational searches