The Atelier Perspective
Merchandising in a store is the elder sibling of this work, still the same instinct expressed differently. A merchandiser walking a shop floor moves the coat to the front because it is warm outside and nobody is touching it. Online, that instinct has to become logic: a sort rule, a boost, a badge, a rule that reads inventory depth, a recommendation trained on what actually gets bought together rather than on what looks good beside it.
Most brands never make that translation. Their collection pages are sorted by "featured", which is upload order, and their bestsellers page is a manual list somebody curated in February. The catalog grew and the logic did not, which is why a larger assortment so often converts worse than a smaller one.
01
Structure Before Volume
A large assortment does not help if the logic around it is unclear. Collection architecture, sort defaults, filter sets, and taxonomy decide whether the catalog feels navigable or overwhelming, and they are decided in the product data long before they are decided in the layout. If two thirds of the catalog has no value in the "fit" attribute, no amount of filter design will save the collection page.
02
Clarity Before Decoration
Presentation should sharpen interpretation rather than add noise. That means a product title a person can scan, the three attributes that decide the purchase visible on the tile rather than buried in the description, and a badge that means something because it is used sparingly. When every product carries a flag, the flags stop being read.
03
Relevance Before Routine
The storefront should respond to behavior, inventory, and margin, not to habit. Sort rules that read sell through and stock depth. Recommendations trained on real purchase pairs. Search results that boost what is available and demote what ships in four weeks. Static sorting is a decision made once and then quietly repeated ten thousand times a day.
04
System Before Effort
Merchandising becomes sustainable when product content, search logic, personalization, and the operational workflow are connected. The test is simple: how many hours a week does someone spend maintaining collection pages by hand, and what would those hours be worth spent on the assortment instead.
our Ecommerce Merchandising Services
Merchandising Strategy
We define how the assortment should be structured, prioritized, and presented, starting from the numbers most brands have not read: revenue and margin concentration by category, sell through by collection, the products that never appear above the fold in any collection, and the share of catalog that has never been seen at all. From that comes a merchandising plan with a hierarchy, sort logic per collection type, and a rule for what earns placement. We also fold in the strategic merchandising questions that sit outside the site: which categories deserve investment, and which are quietly absorbing attention they do not repay.
Collection Page & Product Discovery
Collection pages are the highest traffic templates most brands never test. We rebuild the discovery layer: collection taxonomy that matches how customers describe products rather than how the buying team files them, filter sets built from the attributes people actually use, sort defaults set per collection type rather than site wide, and pagination or infinite scroll chosen on the evidence rather than the trend. Ecommerce product discovery is the phrase for it, and the work is mostly deciding what appears in the first two rows.
Site Search & Zero Results
The search box is the most honest research tool a brand owns, because it records what customers wanted in their own words. We read the query log, fix the synonyms, index the attributes the index cannot currently see, redirect the queries that are really navigation, and turn the zero result report into a standing list of merchandising and content work. A shopper who searches converts at a multiple of one who browses, and a shopper who searches and gets nothing usually does not search twice.
Product Content & Catalog Charity
We strengthen the product layer itself: titles, descriptions, attributes, imagery order, and the supporting detail that lets a customer compare two similar products without opening both. Attribute completeness is the unglamorous heart of this. Filters, search, recommendations, feeds, and personalization all read the same fields, so one afternoon of attribute work improves five systems at once.
Visual Merchandising & Promotions
We manage how the storefront shifts through the year: launches, campaigns, seasonal priorities, and the promotional presentation that goes with them, planned against a calendar so the site changes deliberately rather than accumulating. Banners retired on time. Sale logic that does not bury full price product. Visual merchandising for ecommerce is mostly restraint, which is why so few sites have it.
Search, Filtering, & Personalization Platforms
We build in Nosto, Boost, Athos, and Algolia, configured against real behavior rather than left on default relevance. Nosto is where most of our personalization work lives: category merchandising rules, recommendation placements that are chosen per template rather than dropped everywhere, and segments that reflect real customer states rather than a generic new versus returning split. Personalization is worth doing when there is enough traffic and enough catalog for it to have something to learn from, and we will tell you plainly when there is not.
PIM and Catalog Governance
We support the product data workflows that make all of this sustainable: taxonomy governance, required attributes by product type, enrichment workflow, and the operational consistency that keeps a catalog clean as it grows. We work in Plytix for PIM. Merchandising built on dirty data is a set of instructions the site cannot follow.
Merchandising Platforms We Support
We currently support Nosto, Boost, Athos, and Algolia for search, merchandising, and personalization, and Plytix for PIM, on Shopify, Shopify Plus, and BigCommerce.
Many brands hold several of these platforms and use a fraction of each. Enterprise capability without discipline creates inefficiency. Structure creates advantage.
The right tools deliver more when they're built on the right strategy.
Our pricing
Choose the plan that best fits your needs. From a solid foundation to a fully optimized solution.
Foundation
Basic
$4,500
$4,500
/month
For brands whose search and navigation are quietly losing sales.
What's Included:
Monthly query review across the top 250 terms, with zero-result remediation
Up to 25 curated collections kept in order
Up to 20 rule, synonym, and redirect changes per month
Recommendation placements maintained and monitored
Filter and facet logic maintained against the existing attribute set
Monthly reporting on search conversion and revenue contribution
Signature
Most popular
$7,500
$7,500
/month
For growing brands treating discovery as a channel, on a catalog that needs real attribute work.
What's Included:
Everything in the Foundation program
Product catalog enrichment: attribute mapping and population for filters and facets
Twice-monthly query review across the top 1,000 terms
Up to 75 curated collections and up to 50 rule changes per month
Personalization and recommendation tuning against performance
Two merchandising tests per month, plus seasonal and promotional programming
Bespoke
Premium
Custom
Custom
Contact us for more
For established brands running discovery across multiple storefronts or markets.
What's Included:
Everything in the Signature program
Multi-storefront and multi-locale discovery, including translated attributes
Larger catalogs with continuous enrichment
Advanced segmentation and behavioral personalization
Discovery unified with paid search intent and SEO signals
Continued partnership with a dedicated strategist and custom attribution reporting
Programs run on Shopify and BigCommerce with a personalization and merchandising platform partner. If the platform is already in place, we begin inside the monthly engagement and the first cycle is audit, cleanup, and optimization. If it needs to be built from scratch, implementation is scoped as a separate project. Attribute enrichment runs as a rolling program by priority category.
Scale deliberately.
Automate intelligently.
Expand confidently.
Where does platform limitation currently restrict your growth trajectory?
FAQ
Everything you need to know about our Strategic Merchandising approach.
What is ecommerce merchandising?
Ecommerce merchandising is the practice of deciding what a shopper is shown and in what order, across collection pages, site search, recommendations, and personalization. It covers collection architecture and sort logic, filters and taxonomy, product content and attributes, promotional presentation, and the product data underneath all of it. Done well it raises revenue from traffic the brand already has.
What does an ecommerce merchandising agency do?
An ecommerce merchandising agency owns the logic that arranges the catalog rather than the design that displays it. In practice that means auditing category and margin concentration, rebuilding collection and filter structure, fixing site search relevance and zero result queries, completing product attributes, and configuring personalization and recommendation platforms so the rules reflect real behavior instead of vendor defaults.
How is ecommerce merchandising different from CRO?
Conversion optimization asks whether a page persuades. Merchandising asks whether the right product was on it. The two are measured differently: CRO tests a change against a control, merchandising is judged on discovery metrics such as search exit rate, collection to product click through, catalog coverage, and revenue per session across the assortment. Most brands need both, and merchandising is usually the cheaper of the two to improve first.
What are the most common ecommerce merchandising mistakes?
Four appear in almost every audit. Collection pages left on the platform's default sort, which is usually upload order. Filters built from the attributes that happen to be populated rather than the ones customers use. A site search index that cannot see the attributes held in metafields. And a bestsellers collection that is a manual list somebody last updated two seasons ago.
How do I improve my ecommerce site search?
Start with the query log rather than the platform. Read the top searches and the zero result searches, add synonyms for the words customers use that your catalog does not, index the attributes the search engine currently cannot see, redirect the queries that are really navigation requests, and boost in stock products. Search users convert at a multiple of browsers, so relevance work here pays faster than almost any other merchandising change.