NewNew: The enterprise guide to Agentic AI — 24 min read.

Read →
Case study · Retail & Ecommerce · Agentic AI

63% self-service on post-purchase support at a specialty retailer — agentic returns and orders

A specialty retailer's post-purchase support was 68% of contact center volume and mostly about the same six intents. pronix.ai deployed an agentic post-purchase workflow across chat and voice — 63% self-service, 22-second policy-clear refund resolution and 14-point NPS lift.

Client
US specialty retailer, 1,100 stores + DTC
Industry
Retail & Ecommerce
Platform
Salesforce Commerce Cloud · Amazon Connect · AWS Bedrock
By pronix.ai CX Engineering7 min readQ1 2026
63%
Post-purchase self-service
22s
Average refund policy resolution
+14
NPS on post-purchase contacts
-38%
Cost to serve per post-purchase contact

*Representative outcome; results vary by client, scope and platform configuration.

The challenge

Order status, returns, refund status, exchange, loyalty and warranty made up 68% of contact volume. The IVR shed customers, and agents copy-pasted policy text on every case.

Our approach

Step 01

Intent-bounded agents

One agent per top-6 intent — bounded scope, clean escalation to a human on anything else.

Step 02

Policy as code

Return, refund and exchange rules codified as decision policies the agent consulted — no policy improvisation.

Step 03

Commerce Cloud writeback

Returns and exchanges created directly in Commerce Cloud with the right RMA, right label and right refund path — no callback.

Step 04

Voice and chat parity

Same policies, same tone, same outcomes across voice and chat — the customer picked the surface.

Step 05

Agent handoff with context

Escalations landed with the full intent trail and proposed resolution — agents confirmed instead of restarted.

For Chief Customer OfficerFor VP DigitalFor Head of Contact Center

Illustrative case study. Scenarios, metrics, quotes and client details are representative composites based on Pronix engagements and industry benchmarks unless a named client is shown with written consent. Outcomes vary by client, scope, data quality and platform configuration. Nothing on this page is a guarantee, warranty or professional advice. See our Terms of Use for the full disclaimer.

Trademarks.

AWS, Salesforce, Microsoft, NICE, Genesys, Kore.ai, Google, ServiceNow, Amazon Connect and logos referenced on this site are trademarks of their respective owners. References are for descriptive purposes only and do not imply endorsement, sponsorship or partnership beyond stated partner relationships. See our Disclosures.

Bring this to your program

Talk to the team that shipped it.

Book a working session with a pronix.ai lead on this practice — we'll map the approach to your platform, industry and constraints.