Enter Password
Client
Fractal
Year
2026
Category
Enterprise SaaS,
AI,
E-Commerce
Services
UI/UX Design,
Data Visualization,
Agentic AI Workflows
Enterprise brands managing $2M+ in Amazon sales frequently lose revenue due to slow decision cycles and siloed data across media, supply chain, and account management teams. Cogentiq eCommerce was designed as a plug-and-play SaaS platform that detects revenue leakage early, triangulates root causes across 70+ signals, and delivers prioritized, executable recommendations—reducing decision time from weeks to minutes.
THE CHALLENGE
Managing multiple brands on Amazon involves navigating complex data streams, leading to delayed action, wasted ad spend, and lost sales. Key challenges included:
1.
Revenue Loss from Delayed Action:
Out-of-stock events and listing suppressions caused brands to lose 5% to 15% of weekly sales, with recovery cycles stretching between 2 to 8 weeks.
2.
Wasted Media
Budget:
Disconnected workflows led to ad dollars continuing to run on out-of-stock or suppressed ASINs.
3.
Data Fragmentation & Silos:
Data was scattered across Seller/Vendor Central, ad reports, and supply chain tools, forcing manual weekly reviews to diagnose issues.
4.
Lack of Decision Transparency:
Users lacked clarity into why issues occurred or how financial impact estimates were calculated, hindering rapid executive approvals.
Cogentiq eCommerce Platform Overview
DESIGN PROCESS
An Action-Oriented AI Work Environment
The design process focused on converting complex, high-density marketplace data into an intuitive, recommendation-driven experience that empowers users to act confidently.
From Passive Dashboards to the "Decision Inbox" & "Sparks"
Traditional dashboards require users to find problems manually. We introduced the concept of "Sparks"—actionable decision units representing detected issues like Buy Box losses, conversion drops, or listing suppressions.
The Decision Inbox: Sparks appear directly in a centralized inbox prioritized by calculated financial impact.
Proactive Recommendations: Users start their workday with clear root-cause insights and pre-configured resolution steps ready for review.
The Decision Inbox featuring prioritized "Sparks"
Root Cause Triangulation & Impact Transparency
To build user trust in automated recommendations, the UI triangulates data across 6 critical domains: Suppression, Buy Box status, Promotions, Out-of-Stock Risk, Content Quality, and Retail Media.
Transparent Logic: Designed explicit UI patterns explaining the rationale behind estimated financial loss numbers.
Persona-Tailored Views: Customized workflows for Retail Media Leads, Supply Chain Leads, and Amazon Account Leads to focus on their respective domain signals.
Multi-domain root cause analysis and impact calculation transparency
Closed-Loop Execution & Human-in-the-Loop Controls
To give enterprise clients complete confidence, we established flexible execution modes ranging from automated guardrails to human-in-the-loop approvals.
One-Click Execution: Users can review, adjust parameters (e.g., ticket creation, price adjustments, media pausing), and approve recommendations in one click.
Closed-Loop Status Tracking: Clear progress monitoring with continuous ML logging to verify resolution and impact.
One-click approval and continuous monitoring workflow
Key Highlights & Impact
1.
Accelerated Speed to Action:
Compressed issue detection and root-cause resolution cycles from weeks down to minutes.
2.
Eliminated Wasted Ad Spend
Automatically paused media campaigns on suppressed or low-stock ASINs.
3.
Frictionless Onboarding
Designed for a 1-to-2 day plug-and-play SaaS deployment utilizing native Amazon APIs with zero custom integration needed.
IMPACT & OUTCOMES
By transforming fragmented, multi-domain Amazon data into an action-oriented AI workflow, Cogentiq eCommerce redefined how enterprise brands manage marketplace performance.
Resolution Speed Compressed by ~95%: Reduced the time required to detect issues, perform root-cause analysis, and execute corrective actions from weeks down to minutes.
PPC & Media Spend Optimization: Prevented thousands in wasted ad spend by instantly suppressing or adjusting ad campaigns on out-of-stock or suppressed ASINs.
Rapid Plug-and-Play Onboarding: Delivered actionable "Sparks" within 1 to 2 days of onboarding via standard Seller and Vendor Central APIs—eliminating the need for custom data pipelines.
Enhanced Decision Confidence: Achieved high user adoption by replacing black-box AI estimates with transparent impact calculations and explicit action rationales.
LEARNINGS & REFLECTION
Designing an enterprise AI product operating at the intersection of supply chain, media spend, and live marketplace metrics provided several key takeaways:
Trust Requires Transparency in AI Systems: When AI estimates financial loss or suggests automated repricing, users need to see the "why." Designing transparent calculation logic and clear rationales was critical for user acceptance and executive buy-in.
Designing for High-Density Data Without Cognitive Overload: Enterprise users deal with overwhelming amounts of raw data. The core challenge wasn't showing more numbers, but structuring high-density interfaces so users could scan, evaluate, and act on critical signals within seconds.
Cross-Functional Workflows Require Persona-Specific Views: Supply chain teams and retail media leads view the same product through fundamentally different lenses. Tailoring the dashboard experience around specific user roles while keeping a shared source of truth was essential for seamless cross-team alignment.
