loading...
Encore lets shoppers find used goods across all major resale markets with a single natural-language query covering US, UK, and Canada.
ENTRY ANGLES
Natural language search interface for product discovery across resale marketplaces · Conversational search replacing traditional keyword-based search · Personalized curation and feed algorithms tuned to individual user preferences
VERTICALS
CAPABILITIES
Natural language processing and understanding, Personalization and recommendation algorithms, Multi-marketplace data aggregation and search
ENCORE FOUNDER
“I'm a man in my 50s going on a first date. I want something bold and casual that still shows I have a sense of style.”
Encore built a search engine for finding and buying secondhand goods across resale marketplaces and thrift stores.
Despite the secondhand focus, the site includes a mode for searching new products as well – currently covering the US, UK, and Canada.
The core mechanic: search works in natural language. You don't need to know the right keywords or product attributes. You can say something like: "I'm a man in my 50s going on a first date. I want something bold and casual that still shows I have a sense of style." The AI engine proposes initial results, which can be refined through follow-up chat or by tapping contextual suggestions.
The founders also believe that browsing for interesting things to buy isn't a chore – it's entertainment, worth doing regularly. So the homepage shows a curated discovery feed: what people are searching for right now, personalized picks for individual users, and collections assembled by both the AI and the Encore editorial team.
All of these are personalized to what each registered user is likely to find interesting.
The catalog spans more than just clothing and accessories – watches, electronics, cameras, paintings, furniture, sporting goods, and miscellany are all in scope.
Search is free, with daily limits on the number of recommendations and access to a lighter version of the AI engine. The full experience runs $4 per month for enthusiasts.
Encore is also testing a B2B version that marketplaces and stores can embed on their own sites, with the AI trained on each merchant's specific catalog.
Encore is currently in Y Combinator, which contributed its first $500,000. The launch post appeared on the YC blog two days ago.
In its launch announcement, Encore positioned itself as "the Perplexity for online shopping" – implying that traditional product search, like traditional web search, should start dying once AI tools become the better alternative. The analogy is deliberate: forward-leaning users have already migrated from Google to Perplexity for web research, and Encore is betting the same shift happens in commerce.
As a brief aside: it's worth noting how important that initial positioning is. Another startup – one that [recently built a smart search tool for podcasts](/review/5-millionov-chasov-v-nedelju-iz-kotoryh-mozhno-sdelat-startap) – also called itself "the Perplexity for podcasts." Not "the ChatGPT" – specifically Perplexity.
The reason: Perplexity originally distinguished itself from ChatGPT by connecting to live, current information rather than a static knowledge cutoff. ChatGPT has since added real-time search, but first impressions in brand positioning are hard to dislodge. What gets embedded in users' minds early tends to stay.
Back to Encore's thesis: the company believes AI fundamentally changes how online shopping should work. Instead of keyword search, people should have access to a personal AI stylist or advisor who actually understands what they want.
The difference between a good advisor and a bad one: a bad advisor tries to push you toward whatever's in stock. A good advisor takes time to understand your preferences, finds things you'll genuinely love on first sight, senses when your tastes are evolving, and proactively keeps a curated selection ready for you at all times – not just when you ask.
This critique of traditional e-commerce search is genuinely well-founded. After ChatGPT demonstrated what natural conversation with AI could feel like, the limitations of conventional search became much more obvious.
That applies just as much to product search as it does to web search. Right now, shoppers have to reverse-engineer the language that retailers use to describe products in order to find them. The interface forces the human to adapt to the system. But the system should adapt to the human – the shopper should be able to say what they want in plain language, and the platform should figure out which products fit.
Because retailers haven't made that shift yet, visitors frequently leave without finding anything worth buying. Estimates put the annual revenue loss from failed product discovery at $300 billion across online commerce.
Startups have been taking aim at this problem for a few years.
Curated For You ([covered here](/review/nenajdennye-300-milliardov-dollarov)) built a search widget that retailers embed on their sites – enabling natural language queries like "what to wear to a first date in Miami" and pulling results that account for current local weather. The startup raised $5.9 million.
Lily AI ([covered here](/review/nedoponimanie-na-300-milliardov-dollarov)) analyzes product photos to tag items with the language shoppers actually use, bridging the gap between how retailers describe products and how customers think about them. It raised $71.9 million.
There's been a recurring theme in these pages: the conventional online store is due for a rethink. Here's the condensed version of why.
The traditional format – static product images, technical specifications most shoppers don't understand – is losing ground to formats built around real human recommendations. People want to see actual users sharing genuine opinions. That's a structural push toward short-form video from real customers rather than polished catalog photography from brands.
Keyword search by product attribute is also dated. People buy things to use in their actual lives, so they should be able to describe the use case, context, and constraints in plain English – and let the store figure out which products fit. That requires either a fully rebuilt search layer or a conversational interface that replaces search entirely.
The infinite product catalog – where users have to dig through thousands of items hoping to find something relevant – is giving way to a curated infinite scroll tuned to each visitor. For that feed to convert, stores need to know what each individual user finds genuinely compelling.
And perhaps most importantly: the old assumption that people shop for things they need is increasingly wrong. Most purchases are driven by discovery – by seeing something and deciding you want it. In a 2022 survey, 73% of Americans said most of their purchases weren't planned in advance – they were impulse buys. The year before, that number was just 59%. It's growing.
The implication: stores optimized for "discovering interesting things" can potentially capture 73% of the online commerce market. Stores optimized for "finding things you already decided to buy" are competing for the remaining 27%.
That's a compelling reason to think carefully about which mode of commerce you're building for. Or to build the platform that helps the 30+ million existing online stores make that shift – toward discovery instead of search, using modern user behavior patterns and AI as the foundation. A store built on that platform should feel more like a personalized entertainment feed than a product catalog.
With 30 million potential customers, that's not a small opportunity.