Scaling Lithuania's Leading Online Retailer

The Challenge

With thousands of product listings and growing traffic, Varle.lt's search was becoming a bottleneck. Customers struggled to find what they needed quickly, page load times were climbing, and there was no way to automatically re-engage shoppers when prices dropped or items came back in stock.

The Solution

Rebuilt the entire search experience — starting with Algolia for instant results, then migrating to Elasticsearch when the platform needed more flexibility at lower cost. Overhauled frontend assets to dramatically improve page speed and SEO visibility. Introduced automated customer notifications through Omnisend, so shoppers get alerted the moment a tracked product drops in price or returns to stock. Built modern, interactive UI components with Svelte while keeping server-rendered pages fast with Jinja2 templating.

Results

• Search became near-instant — customers find products faster with smarter filtering and sorting
• Page load times dropped significantly, leading to better SEO rankings and lower bounce rates
• Automated notifications turned passive browsers into repeat buyers
• Reduced search infrastructure costs by migrating from Algolia to Elasticsearch