Adobe's holiday forecast puts a revenue number on AI traffic for the first time at this scale. The company expects AI traffic to U.S. retail sites to more than double this holiday season, up 130% year over year, according to Adobe's annual online shopping forecast. Consumers arriving from AI assistance added items to their carts at a 32% higher rate than those not referred by AI, according to Retail Dive's report on the forecast. Adobe paired the forecast with a complementary survey of 5,000 U.S. consumers.
That growth compounds on a much larger base each year. Adobe measured AI-referred visits to U.S. retail sites up 1,300% in the 2024 holiday season and 693.4% in 2025, so a 130% rise in 2026 still more than doubles last season's volume.
Shoppers who arrive because an AI assistant named your brand are more likely to add to cart. For a marketing leader planning Q4, that changes the math on where to focus product data and page readiness before the spike.
Google's Flipkart pilot shows the next step. Shoppers in India can now buy products from Walmart-owned Flipkart directly through Gemini and AI Mode. Users in the test see a Buy button on select Flipkart product listings that takes them directly to a Flipkart checkout flow without leaving the AI interface. Google plans to roll out the experience more broadly later in October, ahead of India's festive shopping season.
The checkout button moves the transaction inside Gemini. A purchase that starts with a product recommendation can finish without a visit to the retailer's site, and the sale goes to the brand named in the answer.
PYMNTS, in an October 1 analysis, argues that the emerging infrastructure for AI shopping is building a protocol stack. Multiple standards perform different jobs: one helps an AI agent discover inventory, another lets agents communicate, another records what a consumer authorized. The biggest mistake merchants could make in AI shopping, PYMNTS warns, may be waiting for one protocol to win. There probably will not be one.
For a marketing leader, that warning means treating AI distribution as an infrastructure task now. The Adobe numbers give you the reason to audit what an AI platform can discover and transact from your product data. The Flipkart pilot shows the interface where that data gets used. The protocol stack means there is no single standard to wait for.
Start with the audit. Pick a category question your shoppers ask, such as "best running shoes for flat feet." Check whether an AI assistant can name your product, show your price and availability, and complete a checkout. The channel is now both measurable and transactional.
Common questions
How fast is AI traffic to retail sites growing?
Adobe expects AI-referred visits to U.S. retail sites to rise 130% this holiday season, after growth of 693.4% in 2025 and 1,300% in 2024. The percentage is smaller because each year starts from a much larger base.
When will checkout inside AI interfaces be available beyond India?
Google's Flipkart test is live for select users in India, with a broader rollout planned for later in October. The company has not announced dates for other markets.
Do I need to wait for a single AI shopping standard before acting?
PYMNTS argues there will not be one standard. The infrastructure is a protocol stack with multiple standards doing different jobs. The practical move is to make your product data readable to AI platforms now.
Sources
Adobe: U.S. Holiday Shopping Season to Hit Record $275.1 Billion Online, Rising 6.7% YoY
Adobe: Holiday Shopping Season Drove a Record $241.4 Billion Online
Consumer AI use a boon for the holidays as AI traffic doubles | Retail Dive
Google tests buying from Walmart-owned Flipkart through Gemini and AI Mode in India | TechCrunch
PYMNTS | A Vocab Playbook for Merchants Navigating Agentic Standards