How to make the search in your OpenCart as efficient and user-friendly as possible

Search on the site is one of the key tools that helps users quickly and easily find the desired product, and you, as a store owner, to sell this product faster. Effective search improves user experience and increases conversion.

To make the search on the site as effective as possible, it should have certain functionalities:

Speed. The user should be able to enter a keyword or phrase and get a list of results in seconds.
Relevant results. The search on the site should take into account not only the entered keyword but also its variations (morphology), synonyms, and related terms, as well as previous similar queries.
Autocomplete and typo correction. The site should offer suggestions and corrections when entering a query to speed up the search and improve its accuracy.

In general, the search on the site should be intuitive, fast, and relevant for the user. This will help improve the user experience and increase customer satisfaction.

We have implemented all these recommendations in one module - Live search with morphology, relevance, and history.

Let's consider each point in more detail:

1. Speed. Speed is achieved by indexing product names, options, attributes, and their values for each product. The resulting set of words is stored in one database table in the form of words in their basic form, which allows for amazing responsiveness and search accuracy.

"There is elasticsearch and sphinx," you might say. Yes, but their implementation on shared hosting is impossible, and implementation on a dedicated server is labor-intensive and not as flexible as it seems. All the nuances ultimately affect the cost of such an implementation.

If you need numbers, here they are - 640ms to deliver results when searching among 79 thousand products. This includes morphology, and attributes with options for all products. But more on that below.

2. Relevance. Relevance is the degree to which the found product matches the user's query. Here it is also worth noting the experience of previous users' queries, i.e., to offer the user the product that is most in demand for the current query.

This case was presented to us while working with a store with ~500 thousand products. In our module, it is solved simply - the product that was selected more often in previous searches appears higher in subsequent search results.

Here is a simple example. When choosing a "bomber jacket" in the first query, we will get this jacket first in the search results among the products in the second query.

Synonyms. An important nuance of relevance is synonyms. For example, we have a product named "bomber jacket" and "zip-up windbreaker with hood". A windbreaker is also a jacket, and it should also be shown when searching for jackets. In the module settings, we specify synonyms - "jacket, windbreaker", and we get the following result in the search output:

The bomber jacket is higher because its relevance is higher, as we transitioned to it in the previous point.

Morphology. And it is taken into account. The entire search phrase is broken down into individual words that participate in the search in their basic form, singular, nominative case, here is a simple example:

When requesting "denim jackets" we will get a product named "denim jacket". Thus, it does not matter in which case and number the words are specified in the query and in the product name. The buyer will get what they were looking for.

3. Typos. Typos, incorrect keyboard layout - this is a very common phenomenon. Firstly, the module has a built-in Yandex spell checker that corrects typos. Secondly, the module also has a method for correcting incorrect keyboard layouts, in the query below the word "куртки" is written with errors and with an incorrect keyboard layout:

What else can the module do?

It also searches for categories, shows categories where the product is found, groups products by name, takes into account attributes and product options. It also works on the standard OpenCart search page!