How we assess products

Start with the exact model

We identify the model, variant and generation, then check its Indian listing, warranty and support documentation. We look for independent ownership and repair reports, including failures. An old model’s reputation does not automatically apply to its replacement.

Every factual claim needs a source. Manufacturer documentation supports specifications and stated policies; it cannot, on its own, establish long-term reliability. Reposts of the same owner’s story count as one report.

Six areas of assessment

Reliability and longevity: 25 points. We look for independent reports that identify the generation, duration of use, conditions, repairs and failures.

Repairability and spare parts: 25 points. We check disassembly information, replaceable components, part identifiers, manuals and practical access to repairs in India.

Materials and construction: 15 points. We assess construction against the intended use and known failure mechanisms. A material name alone does not establish quality.

India service support: 15 points. We check model coverage, official service channels, regional limits and documented repair outcomes.

Warranty: 10 points. Duration matters alongside exclusions, covered components, labour, transport and the claims process.

Ownership cost: 10 points. We need a dated purchase price and supported repair and maintenance costs. Service-life estimates remain clearly labelled assumptions.

When we publish a score

This is methodology version 1.0, dated 21 October 2026. Each area is assessed from 0 to 4, then multiplied by its weight divided by four. The levels run from substantiated severe limitations (0), through restricted capability (1) and mixed but adequate evidence (2), to strong documented capability (3) and unusually strong independent corroboration (4). Missing evidence receives no assessment and no total.

For reliability, level 2 needs two independent accounts of at least two years, with use context and a search for failures. Level 3 needs three accounts, including one covering five years, plus an independent failure or repair source. Level 4 also needs transparent long-term testing or service data with sample size and limitations. These thresholds guide editorial judgment; they are not statistical proof.

For repairability and India support, level 2 needs model-specific documentation and an identifiable support channel. Level 3 adds identified parts and a verified repair outcome; level 4 adds broad evidence of practical access and repair documentation. Warranty assessments consider fulfilment as well as written terms. Construction needs relevant documentation and failure analysis; cost assessments need sourced figures and sensitivity checks.

The current catalog does not meet the requirements for complete scores or approved recommendations. An unscored product is not a zero-rated product. Category weights have not been adjusted. Any change to this method must be dated and explained.

How confidence is described

High: at least three independent evidence groups, including two five-year ownership records, a generation match, verified India warranty and repair documentation, a search for negative evidence and no unresolved critical conflict.

Medium: at least two independent evidence groups, including a two-year ownership record, exact-model documentation and Indian applicability. Important gaps must still be stated.

Low: some evidence beyond marketing, but limited duration, independence or model matching. Insufficient: documentation or anecdotes do not yet support a conclusion about long-term performance. Manufacturer-only research falls in this group.

Review and corrections

An approved recommendation requires a recorded human review of the evidence and any conflicts. Automated updates cannot grant that approval. Personally tested means an identified person tested the exact unit using a disclosed method; no current catalog product carries that label.

If a supporting claim changes, the assessment must be checked again. Prices, parts stock and service coverage can change after the recorded research date, so confirm them before buying.