Skip to content
Documentation
All docs

Analytics Reference

Product Risk Map

Map products against their average risk scores to prioritise interventions.

The Product Risk map is a scatter chart that lets you see every product in your catalog as a single dot, plotted by return volume (x-axis) and average risk score (y-axis). The upper-right quadrant — high-volume, high-risk products — is where to spend your time.

The four quadrants

  • Lower-left (low volume, low risk) — Healthy long tail. Ignore.
  • Upper-left (low volume, high risk) — Niche problem. Could be a flawed SKU or a single bad customer ordering it repeatedly. Worth a quick look but not urgent.
  • Lower-right (high volume, low risk) — Your hero products. Lots of returns but they look like normal customer behavior. Don’t change anything.
  • Upper-right (high volume, high risk) — These are the products costing you money. Investigate first, intervene second.

What to look at when you find an upper-right product

  1. Click the dot to see the product’s return list, recent scores, and the top reason clusters.
  2. Reason cluster mix — If 60% of returns are SIZING_FIT, the problem is the size guide or the variant lineup. If they’re QUALITY, the problem is the supplier or QC. If they’re CHANGED_MIND on a single customer cluster, you have wardrobing.
  3. Top customers — Is the high return rate spread across many people, or concentrated on a few? Concentrated → customer-side problem (work the customer profiles). Spread → product-side problem (work the product page).
  4. Time pattern — Did the returns spike in a specific week? That’s usually a single bad batch or a confused influencer post that drove the wrong audience to buy.

Filtering

The chart respects the global date filter and adds two of its own:

  • Minimum sales count — Hide products with fewer than N total orders so you don’t over-react to noisy small samples.
  • Product type / category — Narrow the view to a single department. Useful when comparing apples to apples.

Common interventions

  • Pause the SKU — Most extreme. Useful for a clearly defective item.
  • Update the size guide and product photos — Highest leverage for SIZING_FIT and EXPECTATIONS-driven returns.
  • Add a workflow filter — Use workflows to flag every return for the affected product so your team reviews each one.
  • Replace the supplier — Last resort, but the right answer for chronic QUALITY issues.

Next steps