From Keywords to Intent: Retail Search Gets a Reset

A central theme of the "Case Studies: Redefining Site Search and Discovery" session was clear: product discovery has outgrown the traditional search bar.

On stage, Rafe Colburn (Etsy), Colleen Waters (Steve Madden), and Tess Kornfield (ThredUp), alongside moderator Sarah Engel (January Digital), shared how search, personalization and discovery are being rebuilt across modern retail.

Rafe Colburn, Etsy: Intent Over Keywords

Rafe Colburn traced Etsy's journey from basic keyword search to today's AI-driven systems. With over 100 million items and largely unstructured, seller-generated data, Etsy has long faced a complex challenge: matching shopper intent to messy inventory.

That is now shifting with generative AI.

Large language models are helping Etsy better interpret listings and shopper behavior, turning what was once a limitation into an advantage. Instead of relying purely on keywords, the platform is moving toward understanding meaning.

Colburn summed it up simply: intent is the new interface.

He also pointed to conversational search as a growing opportunity, offering shoppers a more natural way to express what they want. Still, he emphasized that adoption will take time, and retailers must balance innovation with familiar user behavior.

Colleen Waters, Steve Madden: Discovery Moves Off-Site

For Colleen Waters, the biggest change is where discovery happens.

Customers are increasingly finding products on platforms like TikTok before ever visiting a brand's website. That shifts the role of ecommerce from discovery to conversion.

Steve Madden has leaned into this with TikTok Shop and partnerships like DoorDash, aiming to meet customers in unexpected moments. The strategy is less about driving traffic to a homepage and more about staying present across channels.

Waters also highlighted the importance of cultural relevance. Moments like the viral "Steve Madden is for the girls" trend show how quickly brands need to move. Lean teams and a test-and-learn mindset make it possible to respond in real time.

On-site, the focus is on bridging inspiration and purchase through features like shoppable lookbooks and outfitting, helping customers discover full looks rather than single items.

Tess Kornfield, ThredUp: Reducing Friction at Scale

Tess Kornfield addressed one of the hardest problems in retail search: scale.

With millions of unique secondhand items, ThredUp cannot rely on traditional merchandising. Search and personalization are essential to making the experience usable.

The company has moved toward more visual and descriptive search. Customers can now search with detailed language, upload images or even shop from Pinterest inspiration. The goal is to reduce friction and make discovery more intuitive.

Visual search has been a key step forward, especially for shoppers who cannot easily describe what they want. Kornfield noted that this shift was driven by observing behavior, not direct demand.

She also highlighted the challenge of context. A broad query can mean very different things depending on the shopper's intent. Helping users refine that intent quickly is critical to avoiding overwhelm.

A Reset for Retail Search

Across all three speakers, the message was consistent.

Search is becoming more contextual, more visual and more connected to the full customer journey. Discovery is no longer confined to a website, and product content must now work across both human and AI-driven environments.

For retailers, the shift is clear.

Success will depend less on matching keywords and more on understanding intent.