# Fast Simon tests AI agents on 50,000 shoppers, lifts product discovery conversion to 22%

*Dual-engine system pairs AI agent recommendations with traditional search, outperforming search-only setups by documented margin.*

By **Jenny Huang Goodman MPA MSc MHSA, Principal** — The Stash Edge, Hako Shikin LLC.
Published 2026-06-06.

Canonical: https://www.pops4.com/stash/articles/fast-simon-2026-06-06t03-2
Subject: Fast Simon
Tags: ai agents, product discovery, conversion optimization, ecommerce search, recommendation engines, dual-engine commerce

---

Fast Simon ran AI agent recommendations through nearly **50,000 e-commerce shopper sessions** and recorded product discovery conversion at **22%**, according to Business Insider Markets. The system pairs an AI agent that proactively suggests products with traditional keyword search. The dual-engine approach beat search-only conversion rates by a margin the company documented but did not break out separately in the public release.

The AI agent operates as a second discovery path. A shopper lands, browses or searches, and the agent surfaces recommendations based on behavior signals and catalog fit. Traditional search remains available. The shopper chooses which path to follow. Fast Simon tracked sessions where the agent fired recommendations and measured conversion when shoppers engaged with agent-surfaced products versus search-only flows. The **22%** figure represents the blended conversion rate across agent-assisted sessions.

The mechanism is recommendation timing and redundancy. Search depends on the shopper knowing what to type. An AI agent removes that dependency by serving candidates without query input. When both engines run in parallel, the shopper has two routes to the right product. More routes mean fewer zero-result dead ends and fewer abandoned sessions. The agent also adapts faster than static merchandising rules. It reads session behavior in real time and adjusts suggestions mid-browse, which shortens the path from landing to cart.

A small physical-product brand cannot deploy Fast Simon's full stack, but the dual-engine principle scales down. Start with your existing site search and layer a simple recommendation widget on the homepage and product pages. Use Shopify's native recommendations or a lightweight app like Wiser or LimeSpot. Configure the widget to show "Frequently Bought Together" or "Customers Also Viewed" based on order history and session data. The widget becomes your agent proxy. It surfaces products the shopper did not search for, adding a second discovery path.

Run both paths simultaneously. Do not hide search. Do not force the recommendation flow. Let the shopper pick. Track conversions by traffic source in your analytics. Tag sessions that clicked a recommendation widget versus sessions that used search only. Measure cart adds and completed orders for each path. If the recommendation path converts at or near your search rate, you have successfully replicated the dual-engine model at low cost. Budget **$29 to $79 per month** for a recommendation app, zero developer time if you use a Shopify-native tool.

The Fast Simon result proves the value of parallel discovery paths in e-commerce. A shopper who cannot find the right product through search alone will often find it through a well-timed recommendation. Brands that run both engines simultaneously capture more of that intent and turn more browsers into buyers.

## The takeaway

Run search and AI recommendations in parallel; shoppers convert when given two paths to the right product instead of one.

---

## Publisher

**Hako Shikin LLC** — Virginia Beach, Virginia. Founded 1997. ASI 217876 · DUNS 18-204-6339.
Principal and author: **Jenny Huang Goodman MPA MSc MHSA**.

- Author: https://www.huanggoodman.com/about
- LLM context: https://www.pops4.com/stash/llms.txt
- MCP endpoint, for AI agents: https://mcp.pops4.com/mcp
- Client dashboard: https://dashboard.pops4.com/
- Catalogue: 70,000+ products, 200+ brands