An AI Offer Agent for the Units That Will Not Sell at List

This one is for operators and investors looking for an AI commerce use case that is already shipping. If you want the deal, the offer agent helps you find a price. If you run the brand, read the merchant version.

AI shopping assistants are getting good at the first half of a purchase. They can read a catalog, compare specs, and land on a product. Then they hit a list price.

A list price is a poster. It is not what a person will pay, and it is not what a brand can accept on a unit that will not clear at the sticker. Overstock is the obvious case. A public sale moves the unit and cheapens the brand. An off-price channel can do both.

Buyr is the use case in the middle. A shopper names a price. An offer agent helps them find a number the brand's rules can accept. The list price does not have to move.

The AI use case is the offer agent

The offer agent sits on the product with the shopper. It helps them find a price. If a number would unlock now, it can point at that guaranteed price. The shopper still chooses the number. The agent does not invent a secret sticker from who they are.

What unlocks the deal is a rule the brand set: a floor, an auto-accept line, and in some cases a check against other offers already in the window. If it clears now, checkout collects that price. If it is still an ask, the shopper is not charged until the offer is accepted. Under the floor, it does not go through.

That is an agent a business can explain. It helps a person find a price. It does not quietly change the price because of a zip code. A shopping assistant outside Buyr can use the same public tools: search the catalog, read the offer band, see whether a price would unlock now, and submit an offer. Cost and margin stay private.

The inventory use case is overstock

Brands make more than they sell at list. A slow size. A color that sat. Units bought so the brand would not stock out. Buyr is a way to move those goods without a public markdown and without sending them to a channel that cheapens the brand.

A shopper names a price on the product the brand still sells as itself. The floor decides if it clears. The unit stays on buyr.com or on the brand's own Shopify product page.

Buyr is not a wholesale desk for one-off lots. We do not buy excess and resell it to jobbers. The use case is the overstocked good that still belongs on a storefront, sold to a shopper at a price the brand would accept.

What the business is

Buyr is a marketplace at buyr.com and a Shopify app. Shoppers set a price on authentic products from authorized merchants. Brands configure pricing models and offer rules, see the demand those offers reveal, and pay 4.9% only when an order is captured through Buyr. No subscription. We make money when they make money.

Most stores only see a purchase at list, or a bounce. Buyr records the price the shopper named. That is willingness to pay on a real product, including units that would have sat.

Where to start

If you are wiring an agent to shop, start with For AI Agents, then Use Buyr with AI shopping agents. If you want the commercial version, sell on Buyr or install the app from the Shopify App Store.

Questions we get

Is the AI real, or a label on a rules engine?

Both pieces are real, and they do different jobs. The offer agent helps the shopper find a price. The unlock is the brand's floor and auto-accept rules. Saying the agent sets a secret price would be false. Saying it helps the shopper find a price that can clear is the product.

Is overstock a use case?

Yes. Goods that will not sell at list, including overstock, are why a brand turns this on. The shopper names a price. The unit moves when it clears the floor. It is not a promise to empty a warehouse, and it is not a B2B liquidation exchange.

Can an outside agent use this without seeing merchant cost?

Yes. Public shopping tools expose discovery, offer bands, and offer creation. Cost and margin stay private.