Unnat Patel

The book · Chapter 7 of 16 · 2022–2024 · Evidence · 6 minutes

Next · 08 Across regions, one review →

The integration advantage

By Unnat Patel

Chapter 07 · Explore the idea

The buyer’s workflow is the product boundary.

A procurement experience crosses the marketplace, approval, ordering and post-order reconciliation.

Quoting eligibility, unit interpretation and search visibility were separate causes; each required its own evidence.

Inside the decision

Procurement system

Enterprise buyers worked through procurement platforms with approvals, purchase orders and invoice reconciliation.

Read supporting chapter text

Enterprise buyers do not shop. They procure — through platforms with names like Coupa, SAP Ariba, Jaggaer and Oracle, with approval workflows, purchase orders and post-order invoice reconciliation. At an enterprise-commerce business I owned the products that connected a marketplace to a range of those platforms, across the US, Europe and Japan.

Inside the decision

Marketplace shopping

Punchout let a buyer shop in the marketplace and return a cart for approval.

Read supporting chapter text

The core integration lets a buyer punch out from a procurement system into a marketplace, shop and return the cart for approval. Integrated accounts were strategically important to the business. The product problem extended beyond adding capabilities: customers faced a lengthy path to useful adoption.

Inside the decision

Return for approval

The integrated cart returned to the buyer’s procurement workflow.

Read supporting chapter text

The core integration lets a buyer punch out from a procurement system into a marketplace, shop and return the cart for approval. Integrated accounts were strategically important to the business. The product problem extended beyond adding capabilities: customers faced a lengthy path to useful adoption.

Inside the decision

Place the order

Ordering through an API or XML order was distinct from post-order reconciliation.

Read supporting chapter text

I wrote the go-to-market strategy across five levers — sales targets by vertical, a shared launch-readiness onboarding framework for the professional-services team, self-service product tooling, marketing, and partner-generated leads — and ran the weekly business review that tracked it. The integration work also distinguished ordering through an API or XML order from post-order reconciliation through the Reconciliation/Invoice API.

Inside the decision

Reconcile after ordering

The Reconciliation/Invoice API addressed the post-order portion of the journey.

Read supporting chapter text

I wrote the go-to-market strategy across five levers — sales targets by vertical, a shared launch-readiness onboarding framework for the professional-services team, self-service product tooling, marketing, and partner-generated leads — and ran the weekly business review that tracked it. The integration work also distinguished ordering through an API or XML order from post-order reconciliation through the Reconciliation/Invoice API.

Inside the decision

Diagnose the actual constraint

I distinguished quote eligibility from unit-of-measure interpretation and search relevance from visibility.

Read supporting chapter text

Then there was quoting. The product that let a buyer request a quote through a sourcing platform was not converting, and the team was working on the quote engine. I pulled funnel data and separated product matches from requests eligible for the existing response path. The analysis also exposed a unit-of-measure interpretation problem: a price expressed per pack could look wrong to a buyer comparing individual units. These were different causes requiring different changes.

The Integrated Quoting strategy reconsidered the response path for suitable demand. A won quote includes price lock and inventory reservation while the customer orders; volume demand goes to a seller quote. I chose a faster implementation over more extensive alternatives to meet the date, and recorded which design should follow later. Correcting the unit-of-measure interpretation was a focused change with a clear customer benefit. We evaluated behavior in shadow before allowing it to affect live responses.

Integrated search taught the same lesson from another angle. Low conversion could not be understood from relevance alone: visibility within the buyer’s experience mattered. I used the available evidence to distinguish whether customers rejected the result or rarely saw it. You cannot fix what you have not measured, and you cannot measure what you have not instrumented.

Enterprise buyers do not shop. They procure — through platforms with names like Coupa, SAP Ariba, Jaggaer and Oracle, with approval workflows, purchase orders and post-order invoice reconciliation. At an enterprise-commerce business I owned the products that connected a marketplace to a range of those platforms, across the US, Europe and Japan.

The core integration lets a buyer punch out from a procurement system into a marketplace, shop and return the cart for approval. Integrated accounts were strategically important to the business. The product problem extended beyond adding capabilities: customers faced a lengthy path to useful adoption.

I wrote the go-to-market strategy across five levers — sales targets by vertical, a shared launch-readiness onboarding framework for the professional-services team, self-service product tooling, marketing, and partner-generated leads — and ran the weekly business review that tracked it. The integration work also distinguished ordering through an API or XML order from post-order reconciliation through the Reconciliation/Invoice API.

Then there was quoting. The product that let a buyer request a quote through a sourcing platform was not converting, and the team was working on the quote engine. I pulled funnel data and separated product matches from requests eligible for the existing response path. The analysis also exposed a unit-of-measure interpretation problem: a price expressed per pack could look wrong to a buyer comparing individual units. These were different causes requiring different changes.

The Integrated Quoting strategy reconsidered the response path for suitable demand. A won quote includes price lock and inventory reservation while the customer orders; volume demand goes to a seller quote. I chose a faster implementation over more extensive alternatives to meet the date, and recorded which design should follow later. Correcting the unit-of-measure interpretation was a focused change with a clear customer benefit. We evaluated behavior in shadow before allowing it to affect live responses.

Integrated search taught the same lesson from another angle. Low conversion could not be understood from relevance alone: visibility within the buyer’s experience mattered. I used the available evidence to distinguish whether customers rejected the result or rarely saw it. You cannot fix what you have not measured, and you cannot measure what you have not instrumented.

Tools in this chapter Punchout / cXML · CIF catalogs · UNSPSC · Coupa · SAP Ariba · Jaggaer · Oracle · price-lock API · SSO · funnel analytics · WBR / QBR mechanisms