+19%

Increase in Search Conversion Rate

+17%

Increase in Search Revenue per User

Industry:Fashion

Business Unit:E-Commerce

Use Case:Product Search & Visual Discovery

Channel:Website, Mobile App

Key Objectives:Increase Search Conversion, Enable Visual Search, Reduce Merchandising Overhead

Products:

ApparelAccessoriesFootwear

How Fashion Nova Achieved a $130M Revenue Uplift with Marqo

Company Overview

Fashion Nova is one of the most-searched fashion brands in the world, offering fast fashion apparel, accessories, and footwear. Founded in 2006 and headquartered in Los Angeles, California, the brand operates primarily through a direct-to-consumer model. Running on Shopify as one of the platform's largest customers, Fashion Nova relies on fast, accurate product discovery to support its high-volume e-commerce business.

As Fashion Nova's catalog expanded and customer expectations increased, the team saw an opportunity to elevate its on-site search and discovery experience. Shoppers were using more descriptive, nuanced language to find products, and Fashion Nova needed relevance that could interpret intent more accurately and evolve with an assortment that changes week to week.

Marqo stood out for its ecommerce-built foundation and its ability to deliver catalog-trained relevance designed for fast-moving retail environments.

Why Fashion Nova Evaluated Marqo vs. Algolia

Fashion Nova evaluated Marqo as a replacement for Algolia after finding that keyword-based relevance couldn't keep pace with the velocity and complexity of their catalog. New collections drop every week, and the team was spending significant time on manual merchandising to compensate for relevance gaps.

Marqo's catalog-trained approach — learning from Fashion Nova's specific product data and real shopper behavior — offered relevance that improves automatically, plus visual search capabilities that match how Fashion Nova's shoppers actually browse and buy.

Challenges

Fashion Nova's catalog changes constantly. New drops, limited-time styles, and an enormous variety of colors, fits, and occasions demanded a search system that could adapt as fast as the assortment — without requiring a full-time merchandising team to keep up.

Long-Tail and Descriptive Queries — Highly descriptive, multi-attribute, and misspelled searches underperformed, causing shoppers to re-search or abandon without finding what they were looking for.

Catalog Velocity — Consistent discovery performance was difficult to maintain as product content and attributes fluctuated with each collection and launch cycle.

Visual and Semantic Relevance — Limited semantic and visual relevance capabilities meant shoppers could not find items by style, occasion, or color when queries were imprecise.

Multilingual Search — Multilingual search functionality was needed to accommodate a diverse, global customer base without per-language manual configuration.

Fashion Nova's customers search with terms like 'bodycon dress for girls night out' or 'monochrome set for brunch.' When search understands style intent — not just keywords — the conversion lift is immediate.

Solution

Fashion Nova deployed Marqo to power search and discovery across its fast-fashion catalog. Marqo's visual and semantic understanding enabled relevance for style-driven queries that change with every new collection drop — handling misspellings, multilingual searches, and imprecise descriptions without manual configuration.

Behavioral signal learning from Fashion Nova's high-volume traffic continuously improved rankings over time. Merchandising controls gave the team the ability to boost, bury, and pin products by category or attribute without requiring engineering support.

Enhanced Intent Recognition: Improved interpretation of descriptive queries, automatic handling of misspellings, and support for multilingual searches without heavy manual configuration.
Visual Search Capabilities: Visual signals from product imagery improved semantic relevance, enabling shoppers to find items by attributes such as color, style, and occasion even when queries were imprecise.
Merchandising Agility: Control over ranking through boosting, burying, and pinning rules targeted by category or attribute, without requiring engineering intervention.
Behavioral Signal Learning: Ongoing ranking improvements over time by leveraging anonymized onsite behavioral and conversion signals from Fashion Nova's high-volume traffic.

Results

Fashion Nova tested Marqo against its existing search solution in a controlled, production A/B test over a two-month period, with 30% of traffic assigned to Marqo. Fashion Nova attributed these gains to improvements in discovery quality across the shopping experience, supported by semantic understanding, image-based relevance, and behavioral signals.

Search performance

+19%

Increase in Search Conversion Rate

+17%

Increase in Search Revenue per User

Overall business impact

~80%

Decrease in Manual Merchandising Hours

$130M

Additional Annual Revenue Attributed to Marqo

What This Means

Fashion Nova's $130M revenue uplift demonstrates what happens when search truly understands fashion. With Marqo powering discovery, trend-forward drops reach the right shoppers faster, descriptive queries return confident results, and the team spends less time on manual merchandising — and more time on what matters.

Style-driven queries return confident results, even for vague or multilingual searches.
~80% reduction in manual merchandising hours as relevance learns automatically.
$130M in additional annual revenue attributed to Marqo's search and discovery.

Results reflect a controlled A/B test conducted over two months with 30% of traffic assigned to Marqo. Outcomes may vary depending on catalog size, traffic volume, and implementation configuration.

Moving Forward

What Comes Next for Fashion Nova

With Marqo powering search and collections, Fashion Nova is continuing to evolve product discovery across the shopping experience. Fashion Nova is exploring agentic discovery capabilities such as automated collection creation, conversational search experiences, more predictive personalization, and context-aware refinements designed to reduce friction while browsing.