Strategic initial callChoose a time
Who I am Competencies References Services Process Insights DownloadsNEW Contact
Book a call
E-Commerce · Pricing 7 min read

Why complete looks can outperform individual items online.

Pairing a top and bottom instead of leaving customers to put the outfit together themselves may sound like a simple merchandising choice. But it brings together product presentation, recommendations, pricing, and cross-selling. Major fashion platforms have long built dedicated “Shop the Look” and “Complete the Look” systems around this idea.

Anyone running an online fashion store knows the challenge: A customer likes a top, but the purchase decision doesn't end there. What pants go with it? Which shoes? Does the color work? What does the full outfit look like? That's where co-ords, matching sets, and “Complete the Look” come in. Instead of selling an isolated product, the store also shows customers what a finished look could be.

1M+
outfits used as training data for “Complete the Look” (Pinterest, KDD 2020)
4M+
fashion items in the underlying Pinterest dataset (Pinterest, KDD 2020)
$2.1B
Stitch Fix revenue in fiscal 2022 (Stitch Fix, Annual Report 2022)

The real problem: Putting an outfit together takes work

In The Paradox of Choice (2004), Barry Schwartz describes why more choice doesn't automatically lead to better decisions or greater satisfaction. The underlying research is more nuanced, however. A meta-analysis by Scheibehenne, Greifeneder & Todd (2010) found no universal choice-overload effect across the studies examined. Whether more choice helps or overwhelms depends on factors such as the situation, the task, and the complexity of the decision.

Fashion e-commerce adds another challenge: Customers aren't just evaluating a product. They're often evaluating how to style it. What pants go with this top? Which color? Which shoes? Will this work for the occasion? The fact that this is a commercially relevant problem is reflected in the technology built by major platforms. Pinterest, Amazon, and Zalando have developed systems that recommend products to complement individual items or complete a look.

A complete outfit doesn't eliminate the decision. But it gives the customer a concrete starting point.

The science behind it: four mechanisms

Choice architecture

How options are grouped and presented affects how easily people can compare and choose between them. A pre-styled look changes that choice architecture. Instead of having to combine several isolated products from scratch, the customer starts with a possible solution and can then decide which pieces are relevant.

Cognitive ease

In Thinking, Fast and Slow (2011), Daniel Kahneman describes how ease of processing can influence perception and judgment. For a fashion store, that doesn't automatically mean “less thinking = higher conversion.” But a clear visual presentation can make the task easier: Customers don't have to imagine entirely on their own how several products might work together.

Complementarity

The underlying commercial mechanism is often simpler than the psychology: Products can complement one another. A top can create interest in the pants shown with it, the pants can lead to the shoes, and the complete outfit can expose customers to several products from the same collection. The condition is simple: The combination has to make sense.

Visual context

Fashion is a highly visual product category. An individual product image answers the question, “What does this item look like?” A complete look also answers, “How could I wear it?” That's why major fashion platforms have invested in visual recommendation systems that surface compatible products and complete looks.

“A set doesn't make the decision for the customer. It gives them a starting point: This is what the finished look could be.”

How individual items and complete looks differ in the shopping experience

Individual items (top or bottom separately)Complete sets / “Shop the Look”
Product is evaluated in isolationProduct is evaluated in the context of an outfit
Customer has to create the combinationCombination is already suggested
Cross-selling happens separatelyComplementary products are part of the presentation
Focus is on a single itemMultiple products become visible at once
Customer starts with the productCustomer can start with the finished look

What the market has done with it

Pinterest

“Complete the Look” as a recommendation system

Pinterest developed “Complete the Look” to recommend products from other categories that complement a particular fashion item. The company describes the task itself as complex because creating a good outfit requires creativity and a sense of style. To develop the system, Pinterest built a dataset containing more than one million outfits and four million fashion items.

Outfits become the recommendation unit
Amazon

“Shop the Look” as visual product discovery

Amazon developed a large-scale “Shop the Look” system for fashion and home. Instead of relying exclusively on product names or categories, customers can discover products in the context of visual scenes and find relevant items.

The look becomes the entry point to product discovery
Zalando

From inspiration to matching products

Zalando Research has also worked on “Shop the Look.” The research focused in part on identifying clothing in fashion and street-style images and matching those items with relevant products available in the store.

Inspiration connects directly to the assortment
Stitch Fix

Curation as a business model

Stitch Fix has made reducing choice a central part of its business model. Customers receive a curated selection of items based on data and styling. The company generated approximately $2.1 billion in revenue in fiscal 2022. That figure doesn't prove that sets inherently outperform individual products, but it does show that curated fashion selection can operate as a business model at significant scale.

$2.1B in fiscal 2022 revenue

The pricing logic behind it

Why sets can be positioned differently on price

Sets create pricing options that go beyond the price of an individual item. The customer isn't necessarily evaluating only the top and pants separately. They may also value the convenience of having a combination already put together.

Set price vs. individual prices

Top CHF 49 + pants CHF 69 = CHF 118 purchased separately. Priced as a set at CHF 109, the customer saves about 8% while the store sells two items instead of one. Whether that actually makes the transaction more profitable depends on margin, conversion, returns, and whether the second item would have been purchased without the set.

Anchoring through sets

A set creates an additional reference price. Customers can compare the price of the complete offer with the individual item prices. Whether that comparison makes the set or the individual product more attractive depends on the actual pricing structure.

Track returns

There's no basis for assuming that sets automatically produce lower return rates. That's exactly why returns should be part of the test. A higher AOV means little if customers also return more items or if the set discount cuts too deeply into margin.

Cross-sell trigger

“Complete the Set” or “Shop the Look” can serve as a cross-sell mechanism. A customer viewing a top can immediately see matching pants, skirts, shoes, or accessories. The metric that matters isn't a generic industry benchmark. It's the incremental contribution margin those recommendations generate in your own store.

What this means for your store

The core principle

Fashion customers aren't always shopping for an isolated product. They're often also looking for an answer to a simple question: How do I wear this and what do I wear it with? That's where sets and “Shop the Look” earn their place. They turn individual products into a concrete styling suggestion while creating natural opportunities for cross-selling.

A set isn't just a product format. Done well, it's also a styling and decision-making tool.

Sources: Li et al. (2020), Bootstrapping Complete The Look at Pinterest, KDD 2020 · Schwartz (2004), The Paradox of Choice · Scheibehenne, Greifeneder & Todd (2010), Journal of Consumer Research, 37(3), 409–425 · Kahneman (2011), Thinking, Fast and Slow · Du et al. (2022), Amazon Shop the Look, KDD 2022 · Zalando Research (2018), Shop the Look with Deep Learning · Stitch Fix (2022), Annual Report 2022.

Want to build your assortment and pricing strategically?

I help you bring set logic, AOV optimization, and conversion strategy together, from assortment to checkout.

Request a conversation →