The Future of Product Discovery: From Search to AI Shopping Agents

AI-driven product discovery: what it means for recommerce

Product discovery is shifting from traditional search engines toward AI agents. Instead of visiting a website or typing a query into Google, consumers will increasingly ask an AI assistant to find the exact item they want.

This change has major implications for recommerce businesses — especially because every item in resale is unique.

From search to agents

Historically, customers discovered products by:

Visiting the brand’s website
Googling a product
Browsing marketplaces

Search results surfaced product pages and Google product cards.

But with AI-driven discovery, the consumer may never visit the store directly. They will instead ask an AI agent (like ChatGPT) to:

Find a specific item
Compare multiple options
Evaluate condition, pricing, and availability
Pull results from several retailers automatically

This makes the AI agent an active shopper — not just a search tool.

Why this matters more in recommerce

Linear, new-product e-commerce is simple:

A new iPhone is the same everywhere.

The AI only needs to compare price and shipping.

Recommerce is completely different:

Each product is unique.

Condition varies.

Battery health differs.

Scratches, wear, color, accessories — everything changes the value.

This is where AI becomes incredibly valuable.

AI agents will:

Analyze product images with vision models
Read condition notes
Compare unique defects
Surface the most relevant matches
Present options to the user like a concierge

Example:

“Find me a 1970s Jimi Hendrix vinyl in near-mint condition under $200.”

The AI fetches results across stores, compares photos, checks condition, and presents the top three.

The new requirement for resale stores

To appear in AI-driven search, merchants must make sure their:

Catalog
Inventory
Item-level condition data
Images
Availability

…are all machine-readable and real-time.

This means small shops can no longer rely on their website alone.

They need software that exposes structured, up-to-date recommerce data to AI agents.

What this unlocks

For resale and rental businesses, AI-driven discovery enables:

Better visibility for niche stores
Faster matching between buyer preferences and supply
Higher conversion because the AI handles filtering
Reduced friction for customers hunting for unique finds

AI will not only replicate human evaluation — it will improve it.

Tuomo: Now I think the product discovery route might be that if you need a product you might already go directly to the brand's website or then you just simply Google it. And if you think about Googling stuff you usually have the websites as results, the online stores, but you also see those product cards there on the top.

Tuomo: Now AI-driven discovery is kind of instead of going to Google, you might go to a chat client like ChatGPT. You’ll have a discussion with an AI agent and say, “I want to buy an iPhone, where can I find one?” or something more specific like, “I'm looking for an old Jimi Hendrix vinyl from this era and in this condition.”

Tuomo: Instead of you searching for the right website, you are hiring an AI agent to do the search for you. That’s the big change in product discovery: the agent goes out and finds the product from different endpoints such as websites and marketplaces.

Tuomo: For smaller stores this means it becomes necessary to make sure their catalog and inventory are constantly available — not only on their website, but indexed at the individual product level so AI agents can access it. Just like Google shows product cards now, AI agents will read individual items.

Karri: How do you see this differing between linear commerce, where new products are identical, and recommerce where everything is unique? Especially because in recommerce images matter a lot.

Tuomo: Yeah, I think there are multiple aspects. First, the agent doesn’t always need to crawl your website — it could access your catalog through an API layer, which is even better. But then images become important. You need images that AI can analyze with vision models.

Tuomo: At minimum, your catalog needs AI-friendly descriptions. Not just “iPhone X,” but things like: “iPhone X with a slight scratch on the bottom right corner.”

Tuomo: In linear commerce, comparison is simple. In recommerce, the agent becomes incredibly valuable. It can show you three different used phones and explain the pros and cons of each: scratches, battery life, color, accessories.

Tuomo: That becomes a concierge experience. Instead of doing hours of manual browsing, the agent surfaces curated results. This removes friction from the buying experience and is especially powerful in recommerce, where small details matter.

‍