Sample Root Cause Analysis Report.
A full sample output for a fictional running shoe with a rising "sole separation" theme. Fishbone-style cause tree with evidence weights, leaf-level hypotheses, and recommended investigations per branch.
What this sample shows
A fictional GlidePath Running Shoe v3 with a theme called "sole separation after ~200 miles," built from 64 related reviews and returns notes over the last 30 days. The analysis produces a fishbone with four branches — Materials (weighted 0.42), Manufacturing (0.31), Design (0.18), Usage (0.09) — and specific leaf hypotheses per branch. The Materials branch is the likely starting point because two leaves there — adhesive batch variance and a new supplier cohort — have the strongest evidence support.
GlidePath Running Shoe v3
Theme: sole separation after ~200 miles · 64 related reviews + returns · last 30 days
Fishbone cause tree
Leaf hypotheses with source language
Adhesive batch variance
The review corpus contains 14 explicit references to "glue line" giving way at the forefoot. Language clusters around a 180–220 mile window, consistent with a curing issue rather than a runtime wear issue. Eight of fourteen reviews mention shoes shipped in October 2025 or later — consistent with a batch date range.
New supplier cohort (Q4 2025)
Cross-referencing the review timestamps against the supplier ledger would confirm. Eight of the fourteen glue-line-specific complaints align to units produced after the October 2025 supplier transition noted in internal QA notes. Strong candidate for supplier audit.
Bond-press temperature drift
Glue-line failures concentrated on the forefoot are consistent with a bond-press temperature excursion during curing. If factory records for October–December 2025 show a temperature drift on line 3 or line 5, this becomes a primary hypothesis. No direct review language supports this leaf — it's inferred from the failure geometry.
Forefoot flex geometry
v3 introduced a more aggressive forefoot rocker compared to v2. If v2 at the same mileage window showed the same failure rate, this branch weakens significantly. Pulling v2 warranty data for the 150–250 mile window is the fastest way to isolate.
Runner weight > 200 lb outlier
Four of 64 reviews mention weight above 200 lb. That's not enough to explain a rising theme. Usage factors are a minor contributing variable, not the root cause. Unless a usage branch is confirmed, don't route this to a warranty-exclusion discussion.
Recommended investigations
- Supplier audit — adhesive lot trace. Pull the supplier ledger for adhesive shipments between September 2025 and January 2026. Cross-reference against the eight review timestamps clustered in the Q4-production batch. Escalate if lot-level tracing matches.
- Factory record review — bond press. Pull thermal logs for bond-press lines 3 and 5 across the same window. A temperature excursion of even 4°C over the cure window is typically enough to produce a failure cluster this shape.
- v2 baseline comparison. Pull v2 warranty and return data for the 150–250 mile window over the equivalent time period. If failure rates match, Design branch weakens. If v3 failure rates are materially higher, Design becomes a secondary investigation.
- Customer outreach — specific cohort. For the 14 reviewers who described glue-line failures specifically, offer warranty replacement with a request for the returned pair for teardown. This is the fastest path to physical evidence.
How Indellia built this sample.
The sample uses a fabricated scenario — the GlidePath Running Shoe v3 is fictional — but follows the pattern of real RCA investigations consumer brands run when a defect theme starts rising in reviews. The four-branch fishbone (Materials, Manufacturing, Design, Usage) is an Ishikawa structure adapted for physical-product feedback. Each branch has 2–4 leaf hypotheses with explicit evidence weighting.
On the live Indellia platform, the Defect Agent (Beta) runs this analysis continuously. When the Anomaly Agent flags a rising defect theme, the Defect Agent automatically produces a cause tree, pulls the supporting review passages, and — where data integrations exist — cross-references returns volume via Loop Returns, Narvar, or AfterShip. The QA team sees the cause hypothesis within hours of the theme appearing.
For RCA explained in depth, see the root cause analysis glossary entry. For the full defect-detection workflow, see detect product defects from reviews.
Keep going.
Have a specific question?
Indellia's AI agents answer with citations from real customer feedback across Amazon, Walmart, Best Buy, and 20+ retail channels.
Cause trees continuously, every rising theme.
The Defect Agent (Beta) runs fishbone RCA automatically when the Anomaly Agent flags a rising theme — with reviews, returns, and factory records cross-referenced.