In today’s world, the efficiency and speed of product search on a website play a key role in meeting customer needs and increasing business competitiveness. This is especially relevant for companies participating in tenders, where the specific way product names and article numbers are written can significantly complicate the search for necessary items. Let’s explore how the implementation of advanced technologies helped one such company optimize the search process on its website and achieve impressive results.
Problem: Inefficient product search on the website
The company operates a website on the Bitrix platform, offering more than 8,000 types of specialized equipment. The product database is constantly updated, requiring an efficient tool for fast and accurate searches. However, the standard site search faced a serious issue: in tender applications, article numbers often use Russian letters replaced with similar-looking English ones. For example, the article number “СВ-13457” could be written with the English letter “C” and the Russian letter “В”. The standard search could not handle such variations, forcing users to manually try different symbol combinations, making the search process lengthy and inconvenient.
Solution: Integration of the ELK system
To solve this problem, the company decided to implement the ELK system (Elasticsearch, Logstash, Kibana) — a powerful tool for data search and analytics. For non-specialists, ELK is a set of programs that allows for efficient collection, processing, and visualization of large amounts of data, ensuring fast and accurate information retrieval.
Implementation steps:
- ELK server setup: A separate server was allocated for deploying the ELK system, ensuring the necessary performance and independence from the main website infrastructure.
- Product indexing: All products and their attributes were uploaded into the Elasticsearch index with pre-processed data markup. This allowed the system to understand the data structure and provide more accurate search results.
- Regular data updates: A system for regularly updating and removing products in the index was set up, ensuring information remains relevant and matches the current product database.
- Development of a custom search module: Instead of the standard Bitrix search module, a custom one was created that sends queries directly to the ELK server. The retrieved product IDs are then processed by standard Bitrix commands. This approach significantly accelerated the search process: a single query is processed in about 50 milliseconds.
- Caching frequent queries: To provide an instant response for the most popular queries, caching was implemented, allowing users to receive search results almost instantly.
- Monitoring and search efficiency analysis: A system for tracking user queries and evaluating their effectiveness was implemented. Now it is possible to monitor whether users navigate to a product page after searching, helping to further improve search algorithms.
- Integration with the client’s analytics system: All data on search queries and user behavior was exported to the client’s analytics system, providing valuable insights for making informed business decisions.
Results and conclusions:
- Improved search: The system can now handle complex queries, considering different variations of article numbers and product names, which is especially important for tender applications.
- Reduced database load: The load on the main database decreased by 30%, positively impacting overall website performance and speed.
- Understanding customer interests: The client can now analyze which products are most interesting to visitors and adjust stock availability according to current demand.
- Increased search speed and accuracy: Users can now find the necessary products faster and with greater precision, enhancing their satisfaction and loyalty to the company.
- Expanded analytical capabilities: Integration with the analytics system allowed the client to gain deeper insights into user behavior on the site and make informed decisions for business development.
The implementation of the ELK system and the development of a custom search module not only solved existing product search issues but also significantly improved the user experience, increased website efficiency, and provided valuable data for strategic planning. This approach demonstrates how modern technologies can be successfully integrated into business processes to achieve tangible results and strengthen market positions.
