03 / PMM Strategy

Automating PMM Intelligence @ Achievers

On the PMM team at Achievers, I identified a broken quarterly process — 8+ hours of manual NPS comment analysis producing inconsistent, stale insights — and built an AI text classification pipeline that cut the time to 30 minutes and surfaced feature gaps that went directly into the product backlog.

Company

Achievers

Role

Product Marketing Analyst (Co-op)

Year

Mar – Dec 2025

The Problem

Every quarter, someone manually read through 700–1000+ open-text NPS comments, grouped them into themes in a spreadsheet, and wrote up a summary. The process took 8 hours, produced inconsistent results depending on who did it, and delivered insights that were already stale by the time they reached the product team.

02 / Solution

The Build

Built a Python text classification pipeline using the ChatGPT API. Raw NPS comments went in, structured thematic analysis came out — categorized by topic and sentiment, with recurring pain points identified across the full dataset.

03 / Outcome

The Result

Analysis time dropped from 8 hours to 30 minutes. Three specific feature gaps — reporting limitations, friction in the user recognition flow, and lack of awards customization — surfaced through the tool and went directly into the product backlog.

Protected Content

The full case study is available on request. Reach out to Newgen directly for access.

Get in touch

Next Project

04 / Fraud

Automating PMM Intelligence @ Achievers

The Problem

Transactional Fraud Analytics ft. Mastercard