What Site Search Technology Do the Biggest Ecommerce Companies Use?

The enterprise search reality

The biggest ecommerce companies typically run one of three patterns:

1. Fully custom built on Elasticsearch or OpenSearch

Amazon, Walmart, and most retailers above $1B+ revenue. Custom ranking models, hundreds of engineers, multi-year build. Not replicable without enterprise headcount.

2. Commercial platform with deep customisation

ASOS, Net-a-Porter, and large fashion retailers often use commercial search platforms (Bloomreach, Constructor.io, Algolia) with extensive custom layers on top. Engineering team still owns the experience.

3. Hybrid: commercial search + custom personalisation

Mid-market and emerging enterprise retailers run a commercial AI search platform with their own personalisation models layered on top via API.

What mid-market teams can borrow from the enterprise pattern

• Treat search as infrastructure. Not a feature you tick off. The platform that runs search is foundational.
• Demand per-query attribution. Enterprise teams measure search ROI at query level. Mid-market should too.
• Pressure-test at scale. Enterprise teams stress-test platforms with millions of queries. Mid-market should test with 50,000+ queries from their own data.
• Plan for the integration layer. Enterprise teams own catalogue feed, behavioural signals, and audience model. Mid-market should own those too, even if smaller.

What mid-market teams shouldn't try to copy

• Fully custom builds on Elasticsearch. The TCO is wildly higher than commercial platforms at sub-enterprise scale.
• Hundred-engineer optimisation teams. The ROI doesn't pencil at mid-market scale.
• Multi-year roadmaps with no measurable lift until year two.

Commercial platforms strong at the enterprise tier

Bloomreach Discovery

Enterprise discovery with CDP. Used by larger retailers. Trade-offs on the Bloomreach alternative page.

Constructor.io

Discovery platform popular with large retailers, especially in apparel and home.

Algolia

Developer-first search powering many large retailers' custom experiences. Trade-offs on the Algolia alternative page.

Commercial platforms that bring enterprise-grade thinking to mid-market

Clerk.io

Mid-market AI search with enterprise-style attribution, per-query reporting and dynamic ranking, without the enterprise pricing or implementation timeline. Optional built-in AI agent handles the configuration work an enterprise team would assign to engineers.

Klevu

Search-led personalisation with the architecture to scale into mid-market and emerging enterprise.

TL;DR

• The biggest ecommerce companies run custom search on Elasticsearch or commercial platforms with deep customisation.
• Mid-market teams should borrow the operating pattern (search as infrastructure, per-query attribution) without copying the cost structure.
• Clerk.io, Klevu, Algolia, Bloomreach Discovery and Constructor.io are commonly evaluated for enterprise-grade search at different price points.