Stars lead where available.
4–5 positive; 3 mixed/neutral; 1–2 negative. Unrated text uses a conservative polarity rule. We do not infer negative sentiment from a company reply.
STATIC RESEARCH CUT · 01 Oct 2026
The organised used-car experience,
in customers’ words.
A transparent, explorable sample of recent public reviews—not a league table, not a lifetime app rating, and not a claim of market-wide sentiment.
01 Jul–30 Sep 2026India storefrontsreview-intelligence-v2.0
PUBLIC RECORDS COLLECTED5,6871,979 excluded by disclosed rules
IN-SCOPE SAMPLE3,708Dated, deduplicated records
OBSERVED NEGATIVE43.8%1–2 stars, or conservative text rule
HIGH / CRITICAL IMPACT18All 18 audited; reported, not verified fault
Why this does not match the App Store headline
Different question. Different denominator.
The storefront headline is typically a cumulative product rating. This tool asks: what did qualifying public reviews posted in one fixed quarter say, after scope and duplicate checks, across several channels?
Interactive explorer
Read the signal. Inspect the denominator.
Static · reproducible · PII-minimised
Selected cohort
All observed brands
3,708 reviews3.22 observed avg.43.8% negative452 / 541 buyer / seller signals
Multi-label classification
% of negative reviewsComplaint themes
Classification rubric
What the labels actually mean
Multi-label, transaction-stage taxonomy.
Negative records can carry several broad categories and narrower subcategories. Counts are mentions, so they do not add to the number of reviews.
Reported impact, not angry language.
Critical requires an explicit safety/privacy event or firsthand legal/title exposure. High requires material loss, vehicle stranding or comparable outcome. Negation and reviewer-confirmed resolution reduce false alarms.
Evidence strength stays separate.
First-person transaction signals raise evidence confidence. A company response is not marked resolved unless the reviewer confirms resolution. All 18 high/critical records were manually audited; allegations are not treated as verified facts.
Investor-use boundary. This is a public-web listening sample, useful for finding recurring product and operating signals. It is not a customer-satisfaction survey, market-share measure, independent verification of allegations, or exhaustive census. Brand comparisons are especially sensitive to source mix and sample size.