Skip to content
Work

Amazon

AI-assisted vendor onboarding at catalog scale

Used catalog and vendor data across 5M+ listings to find what was hurting customers, then launched an AI-assisted onboarding program and smarter categorization.

Senior Product Manager · Business, Industrial & Scientific Supplies (BISS)2021 – 2022

Product

Outcomes

vendor onboarding time
−35%
AI-assisted onboarding program
strategic vendors activated
28
Within 90 days
products added
85K
From newly onboarded vendors
attributed revenue
$140M+
From the onboarding program
classification accuracy
78% → 91%
AI-powered recategorization

How it fit together

  1. 01 · Diagnose

    Catalog & vendor data

    Find customer-facing issues and data-quality gaps

    5M+ listings

  2. 02 · Automate

    AI-assisted onboarding

    Automated catalog creation and support workflows

    −35% time

  3. 03 · Enable

    Vendor self-service

    Webinars and automated email campaigns

  4. 04 · Correct

    AI recategorization

    Business rules defined with Catalog & ML teams

    78% → 91%

Result → 28 strategic vendors in 90 days · 85K products · $140M+ attributed revenue
The four parts of the program, from data diagnosis to automated onboarding, enablement, and recategorization.

Context

Amazon’s Business, Industrial & Scientific Supplies (BISS) categories span more than 5 million product listings. I led product strategy and operational improvements there, partnering with Engineering, Vendor Management, and Analytics to improve catalog quality and the shopping experience.

Challenge

Analysis of catalog data and vendor performance surfaced customer-facing issues and data-quality gaps, which hurt search relevance, product discoverability, and catalog completeness. Bringing strategic vendors and their products onto the catalog quickly and correctly was a lever for both customer experience and revenue.

Approach & key decisions

  1. Let the data set the priorities

    Step 1

    I analyzed catalog data and vendor performance across 5M+ listings to find customer-facing issues and data-quality gaps, then turned them into prioritized product requirements.

  2. Automate onboarding with AI

    Step 2

    I launched an AI-assisted vendor onboarding program that automated catalog creation and support workflows, so strategic vendors could go live faster.

  3. Scale through enablement, not headcount

    Step 3

    I built vendor enablement programs using webinars and automated email campaigns, standardizing catalog-management guidance to improve self-service and product content quality.

  4. Fix classification at the source

    Step 4

    I worked with the Catalog and Machine Learning teams to define business rules and deploy an AI-powered recategorization solution that corrects categories automatically.

Execution & collaboration

  • Partnered with Engineering, Vendor Management, and Analytics on catalog quality and the shopping experience.
  • Defined business rules with Catalog and ML teams for automated category corrections.
  • Ran vendor webinars and automated email campaigns to standardize catalog guidance.
Visit Amazon

You might also like