BRENDEN CAVAZOS

Product · Active Promotions Report

How the QA Automation Was Designed

An enterprise promotion-readiness product built through a design- thinking process -- user research, root-cause discovery, consolidated data modeling, and intuitive interface design -- that replaced a manual, error-prone QA process with a fast, self-service workflow.

2,800+

users adopted the product

99.4%

manual effort reduced

$1.6M

annual savings

400+

associates trained

Design Process, at a Glance

Flow diagram: site merchants individually inspected every live product page and variant, then hand-documented findings -- a manual process. User research with 10+ site merchants and cross-functional partners identified the real bottleneck as fragmented signals, not a lack of data. Pricing, inventory, promotion, and content-quality signals were then modeled into one governed, near real-time source of truth. An intuitive, filterable dashboard plus a formula-driven auto-commenting template replaced manual clicking and hand-typed notes. The result: a multi-hour, multi-person weekly process became a single-owner, 30-minute workflow.

A simplified view of the design-thinking-to-automation process behind the Active Promotions Report. Exact systems, formulas, and data sources are intentionally generalized for this public one-pager.

How It Was Designed, Behind the Scenes

The starting point was not a dashboard -- it was user research. Working directly with 10+ site merchants and cross-functional partners surfaced the real bottleneck: the problem wasn't a lack of data, it was that pricing, inventory, promotional status, and content-quality signals lived in disconnected systems and had to be checked one product page at a time. The fix modeled those signals into a single, governed, near real-time source of truth in BigQuery, using a chain of conditional checks to classify each item -- flagging it as promotional, low-inventory, content-incomplete, or another readiness state -- so the right owner could act without manually cross-referencing systems. That source of truth was wrapped in an intuitive, filterable interface built for how site merchants actually work -- plus a formula-driven template that auto-generated QA comments instead of requiring them to be typed by hand.

Technology Behind the Tool

Capabilities Demonstrated

Prepared by Brenden Cavazos as a public, generalized summary. Internal system names, formulas, data sources, and confidential process details are intentionally omitted. See the full case study for context and additional impact detail.