Notifications
Clear all
Topic starter
10/07/2026 9:11 pm
TL;DR: AI agent referrals in Commerce Network were 13.5 times higher in October 2025 than a year earlier, with sales from those referrals up 894%, as ecommerce brands face a new visibility and checkout environment, according to Signifyd. The governance challenge is not only discovery quality but also whether fraud and identity controls can interpret agent-shaped buying behaviour without blocking legitimate sessions.
NHIMG editorial — based on content published by Signifyd: 9 Steps to Optimize for AI Agents in Ecommerce and Protect Revenue
By the numbers:
- Orders from AI agent referrals in Signifyd’s Commerce Network were 13.5 times higher in October 2025 than they were a year earlier.
- Sales from those referrals climbed 894%.
- At $75 average order value, 10,000 shoppers with a 5% conversion rate represent a $37,500 purchase opportunity.
Questions worth separating out
Q: How should merchants make sure AI assistants describe their brand accurately?
A: Merchants should test the same prompts across multiple assistants and compare the answers to their source-of-truth pages.
Q: Why do AI agents complicate ecommerce fraud controls?
A: Because AI assistants can compress the discovery phase outside the merchant site, legitimate sessions may look unusually fast, direct, or shallow once they reach checkout.
Q: What do security teams get wrong about AI-assisted checkout?
A: They often assume that low page depth or rapid checkout is automatically suspicious.
Practitioner guidance
- Audit AI-facing brand signals Test how multiple assistants describe your brand, products, and policies, then compare the answers against your source-of-truth pages.
- Strengthen machine-readable product facts Update product detail pages, FAQs, schema markup, and policy pages so product attributes, use cases, return terms, and availability are available in clear, consistent language that agents can parse.
- Rebalance checkout fraud rules Review rules that over-penalise direct-to-product visits, low page depth, or rapid checkout.
What's in the full article
Signifyd's full post covers the operational detail this analysis intentionally leaves for the source:
- Step-by-step guidance for auditing product pages, FAQs, and policy text so AI assistants can summarise your brand more accurately.
- Examples of the exact prompts merchants can use to test how different assistants describe their brand and products.
- Practical advice on balancing rapid checkout signals against identity, device, payment, and delivery context in fraud review.
- Workflow suggestions for making AI visibility testing a recurring part of ecommerce operations.
👉 Read Signifyd's analysis of how AI agents are changing ecommerce discovery and checkout risk →
AI agents in ecommerce: are discovery signals and fraud rules keeping up?
Explore further
10/07/2026 9:34 pm
AI-assisted commerce creates a recommendation governance problem, not just a search problem. Merchants are no longer optimising only for human browsing and keyword ranking. They are optimising for how external assistants interpret product intent, policy language, and differentiators. That makes machine readability a governance control, not a marketing preference. Practitioners should treat content consistency, schema quality, and policy clarity as part of the trust boundary.
A question worth separating out:
Q: Who is accountable when AI systems misrepresent product or policy information?
A: Accountability sits with the merchant because the assistant is summarising the merchant’s public signals. That makes content governance, policy accuracy, and structured data quality part of operational control. Merchants should own the review process the same way they own pricing, returns, and checkout risk decisions.
👉 Read our full editorial: AI agents in ecommerce are changing discovery and fraud controls
