Skip to content
Documentation
All docs

Analytics Reference

Financial Impact

Track the monetary cost of returns over time and measure the ROI of your policies.

The Financial Impact tab is the chart you show your CFO. Everything else in RefundSentry is about diagnosing and operationalising fraud detection — this one is about translating that work into dollars saved and dollars lost.

The four headline numbers

  • Total refunded — Sum of all refund amounts in the selected period. The baseline cost of returns to your business.
  • Refund rate — Total refunded ÷ total order value, as a percentage. The single most useful long-term KPI.
  • Estimated saved — Refunds avoided thanks to RefundSentry actions: workflow holds, manual rejections following a HIGH-risk score, chargebacks blocked by an early customer tag, etc. See the methodology section below for how this is computed.
  • Net impact — Estimated saved minus the cost of false-positive holds (refunds that would have been processed normally but got delayed by a workflow). This is the conservative number to quote.

The trend chart

Below the headline numbers is a stacked area chart showing dollar value over time, split into:

  • Refunded (LOW risk) — Normal cost of doing business. Hard to change.
  • Refunded (MEDIUM/HIGH risk) — Where the leverage is. Reducing this stack via tighter workflows and better thresholds is the main job of RefundSentry.
  • Saved by action — Returns blocked, chargebacks prevented, wardrobing avoided. Plotted as a positive band above the baseline.

How “Estimated saved” is calculated

Conservatively. The number includes:

  • Returns where a HIGH-risk score led to a workflow hold and the merchant subsequently rejected the return. Counted as 100% of the refund amount.
  • Returns where a HIGH-risk score led to a manual rejection. Counted as 100% of the refund amount.
  • Chargebacks the chargeback predictor flagged at order creation that were then held or pre-emptively refunded. Counted as the order value.

It does not include:

  • Soft impact like “customer was deterred from abusing because they got tagged.” Real, but unmeasurable.
  • Returns the merchant rejected manually with no preceding HIGH score. Those weren’t RefundSentry’s contribution.
  • Long-term loyalty cost of false positives. Negative impact, also hard to measure.

The methodology errs toward under-counting, which is the right error mode when your goal is a number you can defend in a board meeting.

What the chart can’t tell you

Whether the saved dollars are worth the operational time spent on review. Use the dashboard’s “average review time per return” metric (visible in your team’s settings) and your team’s loaded hourly cost to compute that yourself. For most stores, even a 60% accurate model is cost-effective.

Next steps

  • Period comparison — see how impact has changed since you turned RefundSentry on.
  • Refund methods — change the cost mix from refunds to exchanges where possible.