AquaVerdict
Water being poured into a lab beaker next to a test strip for water quality analysis

Our Method

How We Evaluate

Five checkpoints. Every system. Every time. If a manufacturer will not publish a number, we note the gap — we do not fill it with a guess.

By Tim ParkerReads the spec sheets and prices the full install

Checkpoint 1 — True capacity in gallons

We record media life in gallons, not years. Years are meaningless without knowing household size and daily usage. A 1,000,000-gallon media life in a two-person home lasts roughly 20 years; in a large family it may be 8. We pull the manufacturer's published gallonage figure and note whether it comes with a certified test basis or is a marketing estimate. If the number is not published, we say so rather than accepting a vague “up to X years” claim.

Checkpoint 2 — Flow honesty: GPM at service flow

Peak flow claims — the number usually featured in marketing — represent what a system can push at zero backpressure and maximum inlet pressure. Service flow is what you actually get during simultaneous demand: two showers, a dishwasher, a washing machine. We look for the service flow figure, which is typically 30–40% lower than peak. A system with an 11 GPM peak rating and a 7 GPM service flow behaves like a 7 GPM system in real use. We report the service flow number and flag systems where it is not disclosed.

Checkpoint 3 — Certification scope

NSF/ANSI certification is only meaningful when you know which contaminants it covers and at what test conditions. A system may carry NSF 42 (aesthetic — taste and odor) and NSF 61 (materials safety) but no NSF 53 (health effects). We pull the actual certification listing from NSF's public database, not the manufacturer's summary. If a brand claims PFAS reduction, we check whether that claim is backed by NSF P473 at the specific flow rate the system is sold for. The gap between “tested in our lab” and “independently certified” is where most filter marketing lives.

Checkpoint 4 — All-in 10-year cost

The purchase price is the smallest number in the true cost equation. We calculate: system purchase price at typical street price (not manufacturer MSRP) + installation labor at regional plumber rates + media or cartridge replacement over 10 years at published replacement intervals + any salt, chemicals, or consumables. We use the same household assumptions across all systems in a category — same water usage, same plumber rate band — so the numbers are directly comparable. When installation cost depends on existing plumbing configuration, we note the range.

Checkpoint 5 — Failure modes

Every system fails eventually. We document the most likely failure point (head cracking, media exhaustion, control valve failure, UV lamp burnout), typical repair cost at current parts pricing, and whether replacement parts are available outside the manufacturer's own supply chain. A system with a proprietary head that the manufacturer discontinues in eight years is a different product than one with standard MNPT fittings and commodity media. Longevity is not just about build quality — it is about parts availability a decade from now.

CheckpointWhat we pull
True capacityMedia life in gallons, not years
Flow honestyGPM at service flow vs. peak claim
Cert scopeWhich contaminants, which standard, which test conditions
All-in costSystem + install + 10-yr media and consumables
Failure modeWhat breaks first, what it costs, parts availability

What BUY, IT DEPENDS, and SKIP mean

Every system gets a score out of 100 built from the five checkpoints above, then a verdict stamp. The stamp is a summary of the score, not a separate judgment call:

VerdictWhat it means
BUYThe published specs hold up on their own terms — service flow, media life, and certification scope match what the price and category require — and the all-in 10-year cost is competitive for what the system does. No major disclosed gap changes the recommendation for the household type the page names.
IT DEPENDS / SITUATIONALThe system is legitimate but its specs create a real constraint — a lower service flow ceiling, a narrower certification scope, or a design suited to one water source and not another. The verdict is conditional on the reader's specific situation: household size, water source, or the exact problem being solved. We say plainly which households it fits and which it doesn't.
SKIPThe published specs don't support the price, a required certification is missing for the problem the product claims to solve, or a comparable system scores meaningfully higher at the same or lower cost. A SKIP on a page we could earn a commission from is not withheld — see “no score changes for commissions” below.

The numeric score is a relative ranking within a category, not an absolute physical measurement — it exists so two systems solving the same problem can be compared on the same five checkpoints side by side.

When a manufacturer won't publish a number

Some brands do not publish service flow at rated pressure, exact media volume in cubic feet, or a specific NSF certification scope — only marketing language like “high-flow” or “premium media.” When that happens we do three things, in order: check the NSF certified products database directly for any independently-listed data on that model; check the product's own installation manual or spec sheet PDF, which sometimes carries numbers the marketing page omits; and if neither source has the figure, we state on the page that the number is not published, rather than estimating one. An unpublished number is treated as a scoring gap against that checkpoint, not filled with a plausible-sounding guess.

What we do not do

No pay-for-placement. Brands do not buy their way into higher scores or featured positions. A system earns its ranking by scoring well on the five checkpoints, full stop.

No score changes for commissions. We earn affiliate commissions from several brands. Those relationships are disclosed on our affiliate disclosure page. Commission rates have no influence on scores. Our SKIP verdicts on commissionable products are the evidence for that claim — we do not hide them.

No health claims. We evaluate equipment performance — flow rates, media volume, certification scope, cost. We do not make claims about health outcomes. For health information on specific contaminants, we link to EPA, CDC, or NSF directly.

No invented specs. When a manufacturer has not published a number we need for scoring, we note the absence explicitly. A missing spec is information — it tells you something about how transparent the brand is willing to be.