The Purifier Transparency Score

A purifier review usually answers one question: does it clean the air or the water well? That question matters, and we answer it separately with an editorial score. But it is not the question that determines whether someone regrets their purchase eighteen months later.

What determines that is mundane: whether replacement filters are obtainable in their city, what those filters cost every year, how loud the thing is at the fan speed that actually works, and whether the brand published enough about the product for any of this to be predictable before buying. Almost nobody scores that. So we built a framework that does.

The design constraint we set ourselves was that every axis must be computable from a figure we can point at. There is no panel, no vibes, and — deliberately — no per-product override field anywhere in the code. The moment a score can be nudged towards a product we earn commission on, it stops being worth citing, and being worth citing is the entire point of publishing a methodology.

Min-max normalisation has one cruel property: whichever product sits at the bottom of the cohort is forced to exactly zero, however respectable its actual figure. When we first ran the framework, a purifier that publishes eleven of its fifteen specifications scored 0 for Documentation purely because it published the fewest. Read on its own, that says the brand publishes nothing — which was false.

Documentation and Service reach both have a genuine absolute scale, so they now use it. Documentation is simply the share of tracked fields the brand publishes, multiplied by ten. Service reach maps the brand's network rating onto a fixed 0-10 ladder. The remaining axes stay relative because no absolute scale exists for them — there is no objective answer to whether ₹3,400 a year is a good running cost, only an answer relative to the alternatives.

Purification performance. A high Transparency Score does not mean a purifier works well, and a low one does not mean it works badly. Use our editorial review score for that, and read the two together.

Independent verification. Every published figure here comes from manufacturer literature. We have not measured CADR, noise or water recovery ourselves, and the score does not attempt to detect an overstated specification — only to reward publishing one at all.

Real-world service outcomes. Service reach is a proxy derived from brand network ratings, not from measured callout times in your pin code. It is labelled as a proxy everywhere it appears.

Anything about a product we do not track. The database covers the models we cover. A purifier absent from it is not scored badly; it is not scored.