Nobody was watching where the returns went.

A business analyst case study on cutting returns losses through real-time SKU visibility — built from a 950-record retail returns dataset, start to finish: data, requirements, process redesign, and a live backlog.

Filed by Saneha Gill · Business Analyst / Solutions Consultant · London

View Jira backlog Requirements doc Download full write-up

EVIDENCE LOG — filter by return channel to recalculate the numbers below

—

total returns logged

—

refund value at stake

—

average days to process

—

of volume from top 3 SKUs

01 — INTAKE LOG

Where the volume is actually coming from

Live charts, recalculated from the same filter above. The concentration in 3 SKUs — and the online processing gap — hold at every slice.

Returns by SKU (top 8)

Avg. processing time by channel

02 — OPEN CASE FILES

Three SKUs, three different root causes

Not one generic "reduce returns" problem — each of the top 3 SKUs fails for a distinct, fixable reason.

SKU · CM-4001

Foundation Shade 3

174 RETURNS

Wrong shade matchlead cause

Changed mind2nd

Ordered wrong size/shade3rd

Diagnosis: shade guide / product imagery doesn't set accurate expectations online. Fix sits with listing content, not product quality.

SKU · SK-1001

Retinol Serum

156 RETURNS

Skin reaction / allergylead cause

Changed mind2nd

Seal broken on arrival3rd

Diagnosis: needs clearer patch-test guidance at point of sale, not a packaging or logistics fix.

SKU · EL-6001

Hair Dryer 2200W

144 RETURNS

Doesn't match descriptionlead cause

Arrived damaged2nd

Faulty item3rd

Diagnosis: transit damage + listing accuracy — worth a supplier packaging review.

03 — ROUTE MAP

From week-long blind spot to same-day signal

Toggle between the current process and the proposed redesign.

Bottleneck: no SKU-level visibility until a manual weekly batch review — by which point ~50% of return volume is already concentrated in 3 SKUs with no one flagging it.

04 — CASE STATUS