Hosted Catalog: designing the change, not just the publication

• Need
  Planned-spend catalog inside the buyer’s procurement system
• Failure
  Stale catalogs: failed orders and mispriced items; manual refreshes
• Design
  Customer-initiated and Amazon-initiated change engines
• Guardrails
  Refresh frequency, inventory and price thresholds, approval
• Context Internal procurement as a proving ground; a Fortune 100 customer catalog of about 1,200 ASINs.

Product workflow, shown through the decisions and handoffs.

What I owned

The change-management requirements for customer-initiated and Amazon-initiated changes across North America, Europe and Japan; eight requirement versions of the catalog-management tool; the planned-spend requirements; the hosted catalog vision; a workshop presentation; partner insights from three procurement and spend-management partners; a change-management operating procedure; and a one-pager for the internal procurement pilot.

The decision

Two change engines. Customer-initiated: changes to source criteria such as curated business lists, guided-buying policies, planned-spend agreements, category codes or product lists trigger a refresh. Amazon-initiated: a catalog listener watches availability, inventory thresholds and price-change percentages for catalog products. Customers configure the guardrails.

Guardrail	Range	Why the customer sets it
Refresh frequency	Immediate to yearly	Procurement calendars differ
Inventory threshold	For example, below 50 units	Avoid publishing items that will fail
Price-change threshold	A percentage the customer chooses	Approval workflows key off price
Approval before publish	On or off	Supplier-managed versus internally hosted models

How it moved

Publishing had to work as download, secure file transfer or API depending on what the buyer’s platform supported. A single source of truth for planned-spend agreements separated custom offers, driven by the source, from marketplace offers, driven by product-level thresholds. An internal customer was the at-scale proving ground before external rollout: de-risk at scale with a partner who will tolerate iteration.

What I learned

A catalog is a continuing agreement about what a buyer can rely on. More requirement versions refine the wording; at some point a bounded pilot teaches more than another round.

From my original account

There are two industry catalog types: internally hosted, loaded by the buyer, and supplier-managed, pushed by the supplier with buyer approval. When availability, inventory or price changed on Amazon’s side, or a customer changed source criteria, the buyer’s catalog did not update. This caused failed orders and mispriced items; refreshes were manual. I wrote the change-management requirements across North America, Europe and Japan.

Across regions, one review

I authored and presented quarterly business reviews for the whole ESI portfolio across the US, Europe and Japan. The Q2 2023 review covered eight portfolio goals, including Hosted Catalog and Punch-in. I also worked on weekly reviews for search, Hosted Catalog and quoting, and a sourcing-partner monthly business review in February 2024.

For Hosted Catalog, I wrote eight versions of the catalog-management tool requirements, the planned-spend requirements, a vision, a workshop presentation, partner insights, a change-management operating procedure, and a one-pager for the internal procurement pilot. The internal procurement program was the proving ground for Hosted Catalog and related purchasing capabilities before external rollout.

Related purchasing capabilities had dependencies of their own. I worked on the relationship between who places an order and who receives it, and used an internal customer as a proving ground before wider introduction. The strategic choice was to learn with a customer who could tolerate iteration while still providing a meaningful test of the experience.

I ran weekly Punchout insights and authored quarterly reviews for the ESI portfolio. Punch-in provided a separate adoption lesson: partner readiness was green, but customer adoption was well below goal. I shifted to demand-side levers: European customer-advisor incentives, staged ERP-partner pilots, four procurement partners’ lead lists, and market-development funds tied to adoption targets. Partner readiness does not equal adoption.

Tools in this chapter versioned requirements · catalog listener design · SFTP / API publishing · quarterly and weekly business reviews · regional requirements

The pipeline included 192 customers and a Fortune 100 consumer-goods customer’s catalog of about 1,200 ASINs. The 2023 quarterly business review set goals to launch by July 31 and onboard three e-procurement systems by December 31. Planned delivery on direct orders was a company-level goal. Together, these described the opportunity and intended rollout.