---
url: https://pmfolio.xyz/demo/fraud-review-queue
title: Cut fraud rate 64% in 90 days with a redesigned review queue — pmFolio demo
description: Finova's fraud operations team was manually reviewing flagged transactions using a brittle internal tool built in 2019. As transaction volume grew 3× in 18 mont
type: pm-case-study-demo
---

# Cut fraud rate 64% in 90 days with a redesigned review queue

**Company**: Finova
**Role**: Lead PM
**Duration**: 4 months
**Team**: 6-person squad
**Skills**: 0→1, Risk, Data, Ops design, Stakeholder management

## Key Outcomes

- Fraud rate reduced from 3.2% to 1.1% of transaction volume
- Analyst review time cut by 52% (from 8 min to 3.8 min per case)
- False-positive rate dropped from 22% to 9%, reducing customer friction

## 01 — The Problem

Finova's fraud operations team was manually reviewing flagged transactions using a brittle internal tool built in 2019. As transaction volume grew 3× in 18 months, the backlog of unreviewed fraud alerts hit 48 hours — meaning real fraud was being caught too late, and legitimate transactions were being declined while customers waited. The tool surfaced alerts in chronological order with no risk scoring, no case history, and no bulk actions.

> Every percentage point of fraud directly hit the P&L, and regulator SLAs required us to action high-risk alerts within 4 hours. We were failing that target 40% of the time.

## 02 — Research & Insights

**Methods**: 12 contextual interviews with fraud analysts, 3 weeks of shadow sessions, Mixpanel funnel analysis, 6 stakeholder interviews with Risk, Compliance, and Engineering leads

Analysts spent the first 2 minutes of every case hunting for context scattered across four tabs: transaction history, device fingerprint, account age, and prior flags. The tool showed alerts in FIFO order regardless of risk score, so high-risk fraud was often reviewed after low-risk noise. 78% of analysts said they had a mental 'pattern' they checked every time — but the tool didn't support it. We also discovered that 60% of false positives came from a single category: first-purchase, high-value orders from new accounts in the EU.

## 03 — Solution

We redesigned the review queue around the analyst's mental model rather than the database's record structure. The new queue surfaces cases ranked by a composite risk score (ML model + rule-based signals), consolidates the four context tabs into a single card with progressive disclosure, and adds one-click bulk actions for common patterns. We shipped in three phases: risk-ranked queue first, then the consolidated case card, then bulk actions.

### Key decisions & trade-offs

The biggest debate was whether to build a fully automated decision engine or improve the human review flow. I pushed for the human-in-the-loop approach: our ML model had only 78% precision — not good enough to automate high-stakes declines, and analysts' domain knowledge was genuinely catching edge cases the model missed. We agreed to improve tooling now and revisit automation in 6 months once precision improved.

## 04 — Results

| Metric | Before | After | Delta | Timeframe |
|--------|--------|-------|-------|-----------|
| Fraud rate (% of volume) | 3.2% | 1.1% | −64% | 90 days |
| Avg review time per case | 8 min | 3.8 min | −52% | 90 days |
| False-positive rate | 22% | 9% | −59% | 90 days |
| SLA compliance (4-hr rule) | 60% | 97% | +37pp | 90 days |

## 05 — Challenges & Learnings

### Challenges

The biggest challenge was the 6-week data migration from the legacy tool. Our fraud data was spread across three systems with inconsistent IDs. We hit a 3-week delay when we discovered that ~8% of historical cases had duplicate entries, which would have corrupted the ML training data.

### What I'd do differently

I underestimated how much analysts relied on muscle memory with the old tool. Even though the new interface was objectively better, we saw a 2-week productivity dip after launch as people adjusted. I'd budget a longer hands-on training period next time and consider a gradual rollout rather than a hard cutover.

---

*This is a demo case study built with pmFolio. Build your own: https://pmfolio.xyz*

*Back to demo portfolio: https://pmfolio.xyz/demo*