Perplexity Buy Now agent hits 2M monthly shoppers, $2B GMV

The 2M-shopper milestone

Ecommerce Times reported on July 2, 2026 that Perplexity's "Buy Now" AI shopping agent crossed 2 million monthly active shoppers, and framed the milestone as a real threat to the DTC funnel. A parallel piece from the same day tied Perplexity's shopping layer to a $2B annualized GMV run rate and cast the combined story as a direct challenge to Google Shopping's product-search moat (Ecommerce Times, July 2, 2026 and Google product-search threat coverage). A third data point published a few days earlier put the same GMV run rate in the same $2B range (The Keyword, June 29, 2026).

The economic model matters more than the shopper count. When a customer completes a purchase through Perplexity's Buy Now flow, the transaction bypasses the brand's owned checkout: no post-purchase email trigger, no loyalty enrollment, no pixel fire, no first-party data. Perplexity reportedly takes an 8 to 12% revenue share on completed transactions on top of Stripe processing. For DTC brands running at 20 to 25% contribution margins, that take rate compresses profit toward zero on Perplexity-originated orders unless the brand treats Perplexity as an acquisition channel with a CAC line, not as incremental revenue.

Amazon's response is telling: in April 2026 it restricted Perplexity's access to real-time Amazon pricing and inventory, per Ecommerce Times sourcing, which cut Perplexity's Amazon-attributed GMV by roughly 30% inside 60 days. The unintended consequence is that Perplexity accelerated direct-merchant catalog build-out, and DTC brands on Shopify now sit at the front of Perplexity's recommendation engine.

Key facts

Timeline

Category signals to watch

Early Perplexity commerce sessions over-index in high-consideration, spec-heavy categories: consumer electronics accessories, home goods, pet nutrition, and specialty apparel. One outdoor gear brand reported that 11% of its Perplexity-attributed orders in May came from queries it had never seen in Google Search Console. The takeaway is not the exact percentage. It is that AI agent commerce is a new funnel entry point, not a reshuffle of the existing Google Shopping funnel.

What Amazon sellers and DTC brands should do

1. 1

   Audit structured data on the top revenue ASINs

   Run product URLs through Google's Rich Results Test and Schema.org validators. Perplexity's indexing crawler evaluates the same signals. This is also the fix for the Amazon Listing Quality Dashboard rollout.

2. 2

   Claim catalog presence in the Perplexity Commerce Channel

   Even brands not ready to transact should install the early-access Shopify app and index the catalog before competitors do.

3. 3

   Model Perplexity as an acquisition channel, not incremental revenue

   Assign a CAC line, plan retargeting from post-purchase data, and negotiate the take rate down where volume or exclusivity supports it (6 to 7% has reportedly been negotiated on annual commitments).

4. 4

   Track AI-agent contribution margin separately in the P&L

   Add Perplexity, ChatGPT Shopping, and Google AI Mode as their own channels in Profit & Loss so the take-rate hit and the lost data value are visible at cohort level.

5. 5

   Get your data model ready for AI agents on Amazon too

   Perplexity is one entry point; Rufus, ChatGPT, and Claude are others. Sellers who can answer "what changed in my P&L last week and why" in plain language have the highest ceiling once agents get real read access. See Amazon AI Agents and the Nova + Claude MCP integration.

Related coverage

• Nova + Claude MCP integration - what agent-native access to Amazon data actually looks like.
• Amazon and Perplexity Comet agentic commerce - the Amazon-Perplexity relationship in the Comet browser context.
• Google Gemini agentic checkout - the other side of the agent-commerce race.

How Nova helps

Nova is the operating system for Amazon brands. As AI agent commerce grows, the brands that hold margin will be the ones whose data model AI agents can actually read, and whose contribution margin is broken out at cohort level by acquisition channel.

• Amazon AI Agents - how Nova prepares an Amazon business for AI agent workflows.
• Profit & Loss - reconciles 40+ Amazon fee types at SKU level so a new acquisition channel plugs in without breaking the P&L.
• Custom Breakdowns - group by channel, campaign, or query intent to see where AI-agent-driven orders show up.