Data Analyst Mission — Product Quality Risk: ZCPTM-SS-M-BW

Mission/open goal Description

Data Analyst — Investigate a hidden product quality issue in Zava's data using the PromptathonDb SQL Server database.

Harness and model

Used Copilot Agent mode with SQL MCP tools to investigate the database, following this path: SupportTickets → SupportChats → Products → SalesOrderLines. Combined quantitative ticket/sales metrics with qualitative complaint themes from support chat transcripts, as required by the mission.

Turn-by-turn journey

SELECT RelatedSKU, COUNT(*) AS TicketCount
FROM SupportTickets GROUP BY RelatedSKU ORDER BY TicketCount DESC
→ ZCPTM-SS-M-BW had 9 tickets, the highest of any linked SKU.

SELECT RelatedSKU, AVG(SatisfactionScore) AS AvgSatisfaction
FROM SupportTickets WHERE RelatedSKU = 'ZCPTM-SS-M-BW' GROUP BY RelatedSKU
→ Average satisfaction score: 1 (lowest observed).

SELECT SKU, SUM(Quantity) AS TotalQuantity, SUM(LineTotal) AS TotalValue
FROM SalesOrderLines WHERE SKU = 'ZCPTM-SS-M-BW' GROUP BY SKU
→ 228 units sold, $19,825.05 in revenue.

Ticket categories: ProductQuestion (4), TechnicalSupport (3), Returns (2).

Support chat themes: recurring complaints about app connectivity/setup
friction, unexpected billing charges, and return intent.

Full evidence and additional queries:
https://github.com/76Meharzad-Waseem/sql-ai-promptathon/blob/main/missions/data-analyst.md

Completion

• Yes, the agent completed the mission or goal.
• [] No, the agent did not complete the mission or goal.

Bonus work

SKU ZCPTM-SS-M-BW (Premium Short Sleeve Men's Top) is the strongest
product-quality risk candidate — it combines meaningful sales volume
with the highest support ticket count, lowest satisfaction score, and
recurring negative themes. Recommend Zava prioritize this SKU for
immediate product-experience investigation.
What worked well: Combining ticket volume, satisfaction score, and chat
transcript themes created a persuasive, evidence-backed story rather
than relying on a single metric.

What didn't work as well: The Codespace's SQL Server connection wasn't
reachable at first and required troubleshooting. Some support tickets
weren't directly linked to a SKU, requiring careful filtering.

Read missions/data-analyst.md and follow the documented SQL investigation path from SupportTickets through SupportChats, Products, and SalesOrderLines using the PromptathonDb SQL Server database. Compare ticket volume, satisfaction, sales, and chat themes; the mission is done when the evidence supports a documented product-quality risk conclusion.