NewsIntegrations for PHP, JavaScript, and commerce platforms

Find the right product on the first search

Lexis turns store search into a relevance system: it interprets the query, orders results by real signals, and surfaces where demand is being lost.

Fewer lost searches · more relevant results · decisions from real behavior

Exemplu de căutareMagazin online
Works with
OpenCart
Magento
WooCommerce
Shopify
PrestaShop
Laravel
CodeIgniter
Node.js
PHP
</>
REST API

Classic search breaks on the queries that matter.

A store doesn't lose sales on perfect queries. It loses them when the customer types fast, uses a different word, searches by code, or expects the system to understand context. That's where a default engine or a plain LIKE starts to fail.

Typos
upright vaccum

A missing letter shouldn't break the path to a product that's already in the catalog.

Synonyms
sneakers / trainers

Customers search in their own words. The catalog sometimes uses a different name.

Zero-result searches
0 results

An empty page kills the session right when intent has already been expressed.

SKU searches
BKC-48291

Product codes are maximum-intent queries and must be handled differently from vague text.

Ambiguous searches
battery

The same word can mean a product, a category, or an accessory. Without context, the order becomes noise.

Lexis turns a weak query into a useful result.

Relevance is built before products are shown: normalization, exact matches, intent, and catalog context.

01
Normalization

Lexis cleans the query before the index: typos, grammatical forms, plural, singular, and searches without diacritics.

02
Catalog binding

Customer terms are mapped to the real products, categories, SKUs, and names in the catalog.

03
Recovery

When an exact match is missing, the system proposes the closest useful search instead of an empty page.

Products asked for often rise where it counts.

Lexis uses precise searches as ranking signals. If a product is asked for explicitly or shows up consistently in very specific queries, the order of results starts to reflect demand.

Signals that shift the order
Signal

SKU searched directly

What it learns

The product is asked for explicitly, not discovered by accident.

Effect

The exact product gets priority in relevant contexts.

Signal

Search with 1-2 results

What it learns

The query shows clear interest in a narrow set of products.

Effect

Those products gain weight in broader searches.

Signal

Repeated specific searches

What it learns

Customers return to the same intent even when the terms vary.

Effect

Ranking gradually shifts toward what customers actually look for.

Real searches become indexable pages — only when they deserve to.

Frequently searched terms can become dynamic pages ready for Google, but only when there's clear demand and real coverage in the catalog.

How an SEO page is decided
01
Read the real searches

The system starts from customer language, not from invented keyword lists.

02
Verify the intent

The term must signal a commercial need clear enough to justify a dedicated page.

03
Validate catalog coverage

A page isn't generated if it only has 1-2 products. Without coverage, it stays an internal signal.

04
Build the dynamic page

The page gets a stable route, text, and relevant products, then updates as demand and the catalog change.

See what customers search for and where they get stuck.

Search performance must be measured: volume, unanswered searches, corrections applied, and terms that point to missing products or new pages.

The data below is a sample report, used to show what you track after connecting the catalog.

Analyzed searches
18,420

Volume for the selected period, separated from accidental spikes.

Zero results
312

Terms where a correction, a mapping, or a real catalog gap needs a decision.

Corrections applied
1,864

Queries that got a useful direction instead of staying missed searches.

Searches that need attention
Search
20000 mah power bank
Correction / direction
clear intent on capacity and category
Results
24
Status
resolved
Search
carbon hood filter
Correction / direction
relevant category, products available
Results
18
Status
resolved
Search
women's running shoes
Correction / direction
term eligible for a dynamic page
Results
9
Status
seo
Search
kids ski helmet
Correction / direction
repeated demand, catalog doesn't cover the term
Results
0
Status
opportunity
Analysis windows
Same data, three different rhythms.
1 day
Issues from today

Good for quick interventions after campaigns, product imports, or catalog changes.

7 days
Stable trends

Good for seeing whether an adjustment works and whether a term repeats often enough.

30 days
Catalog and SEO decisions

Good for weighty decisions: indexable content, missing products, and merchandising priorities.

It plugs into the catalog and becomes the store's search.

Lexis belongs between the catalog and the customer, not beside them. The integration syncs products, builds the search index, and exposes results via API, JavaScript, PHP, or connectors for commerce platforms — without moving the source of truth out of the store.

01
Connect the product source

The API, JavaScript/PHP code, or platform connector sends products, categories, prices, stock, SKUs, and the attributes used in search.

02
Build the index

The data is prepared for exact matches, semantic relevance, and behavior signals.

03
Activate search

The store requests results from Lexis, and the response comes back with products ordered by relevance and learned signals.

04
Control the change

Important rules stay visible: what runs automatically, what goes into review, and what must not be promoted.

Start free. Pay when you scale.

Three plans, no per-feature surcharges, no setup fees.

Free
Freeforever

For small stores or evaluation.

  • 10k docs · 100k searches/month
  • API + OpenCart plugin
Start free
Most popular
Pro
€29per month

For storefronts in production.

  • 1M docs · 1M searches/month
  • Reranking + SEO pages + email support
Start with Pro
Scale
Customby volume

Heavy volume, multi-region, negotiated SLA.

  • Custom limits · active failover
  • Dedicated support + onboarding
Contact us

Quick answers to what people ask us first.

How long does integration take?

On OpenCart with the official plugin: 15-30 minutes (install + initial catalog sync). For custom storefronts via REST API or SDKs, an evening for a mid-level developer. The free playground is available before any decision.

What platforms do you support?

Official plugin for OpenCart 3.x and 4.x. REST API + PHP/JavaScript SDKs for any other stack — Magento, WooCommerce, Shopify, custom. Dedicated connectors for Magento and WooCommerce are in roadmap.

Where is my data stored?

The cloud runs in Frankfurt (EU), GDPR-compliant. Your customers' data stays isolated per organization and is never aggregated cross-tenant. For self-hosted (on-premise), everything runs on your infrastructure — nothing leaves.

What happens after the Free tier?

Free stays free forever (10k docs, 100k searches/month). If you exceed the limits we email you first and you can upgrade without downtime. We don't hard-block.

Can I leave if it doesn't work out?

Yes, any time. Export your catalog and analytics from settings in one click. No penalty, no minimum term, no sales call required to export.

What support do I get on each plan?

Free: public docs + community. Pro: email with same business-day reply. Scale: direct line to the team + negotiated SLA. All plans include the playground and public docs.

Measure search performance on your catalog.

Start with a demo on real data or go straight to integration. The goal is to see how many searches turn into useful results, not just how many queries ran.

What you measure first
Unanswered search rate

Where the current engine fails to connect demand with the catalog.

Recovered searches

How many queries get a useful direction after normalization and interpretation.

Growth signals

Terms that deserve better ranking, new content, or a catalog decision.