INSIGHTS · TECHNOLOGY AND AI DILIGENCE
Agentic commerce: are your portfolio companies ready?
AI assistants are starting to find, compare and buy products on customers’ behalf. The volumes are small but growing fast – and every route to AI commerce depends on one thing most businesses under-invest in: accurate product data.
Palladium Digital Group
·
5 min read
In brief
AI-referred orders are growing fast, from a small base.
OpenAI and Google are taking different routes to AI checkout.
Every platform depends on the same thing: accurate product data.
Picture a shopper asking an AI assistant for a waterproof walking jacket under £150. The assistant compares options, checks stock, reads the returns policy and recommends two. The shopper goes straight to a product page, ready to buy.
This is starting to happen at scale. AI assistants are starting to find, compare and buy products for customers. Shopify reported AI-referred orders up nearly 13x year-on-year in Q1 2026, and those shoppers converted at nearly 50% higher rates than visitors from organic search. That growth is from a small base, and organic search still sends far more traffic, but the direction is clear.
~13x
Year-on-year growth in AI-referred orders, Q1 2026 (Shopify)
~50%
Higher conversion than visitors from organic search
The platforms disagree
The platforms disagree
In September 2025, OpenAI launched in-chat checkout (Instant Checkout) in ChatGPT, but in March 2026 it stepped back, saying the first version lacked the flexibility it wanted. Most merchants now use their own checkout, and ChatGPT focuses on helping people discover and compare products.
Google has taken the opposite route. In January 2026 it launched the Universal Commerce Protocol, an open standard developed with Shopify, Etsy, Wayfair and Target. It then began building checkout into AI Mode and Gemini for eligible US retailers.
Both approaches may change again. A business that builds its plan around one platform is making a bet.
What every platform needs
What every platform needs
An AI assistant chooses products using the data it can get. It needs:
complete, consistent product data
live price and stock
clear delivery and returns information
a site it can access
If data is missing, conflicts across channels, or AI crawlers are blocked by robots.txt or bot-management settings, the product is less likely to be recommended.
Why this matters to investors
Why this matters to investors
In Checkout.com’s 2026 survey, 9 in 10 merchants said they are preparing for agentic commerce and 42% said they are testing it. Investors need to know which portfolio companies have done the groundwork.
In diligence, three questions are worth asking:
1
Can AI crawlers reach the site today?
2
Who owns product data quality, and how is it measured?
3
Is management building data and systems that work on any platform, or an integration with one?
In value creation, the same fixes also help search, marketplaces and conversion on the company’s own site. Start with site access, then product data for the best-selling lines, then ownership and measurement.
Follow along
We’re tracking how AI is changing the way customers find and buy.
Want to know what this means for your business or portfolio? Book a meeting to talk it through with our team.
CONTACT US
Make smarter growth decisions.
Understand the economics behind customer acquisition and build a strategy for sustainable, long-term growth. Whether you’re pricing an asset or improving one, let’s talk.
Your personal data will be processed as per Palladium Digital’s privacy policy. By submitting this form, you consent to the use of your personal data in accordance with the terms thereof.
London
1 Clerkenwell Green, London, EC1R 0DE
+44 203 882 8567
info@palladiumdigital.co.uk
New York
W 17th St, New York, NY 10011
+1 857 222 8188
info@palladiumdigital.co.uk
Oxford
102–104 St Aldate’s, Oxford, OX1 1BT
+44 186 579 3800
info@palladiumdigital.co.uk