Every guide on Clear Water Picks explains where the research came from, what data drove the ranking, and which reader the page is meant to serve. We do not fabricate testing. We analyze real owner data, verified certifications, and community consensus.
Each page starts with the reader's water problem, not the product. We gather source material from long-term owner reports, NSF certification databases, community discussions, and expert reviews. The page goes to the editorial desk that best serves the reader: The Tap Report for everyday households, The Well House for well water and whole-home systems, The Filter Lab for certification-focused comparisons.
We look for convergence across independent sources. A contaminant reduction claim earns trust when certification data, long-term owner reports, and community discussions all point the same direction. Not when a manufacturer says so on a product page. When a guide relies on NSF certification, we check that claim against the public certification database at info.nsf.org.
Reviewer scores reflect how well a filter serves the specific reader the page is written for. The same RO system might score differently on a budget page, a PFAS removal page, or a well water page because the reader's needs are different. Scores use 0.5 increments and are independent editorial judgments, not averages of other ratings.
When relevant to the category, we check NSF/ANSI certification status and which specific contaminants are covered. We give extra weight to owner reports that describe living with a filter well after setup, not just first-week impressions. We also look at replacement filter cost, real-world filter lifespan, flow rate impact, installation difficulty, and the contaminants the filter does NOT remove, because every filter has limits.
Affiliate commissions keep the site running, but they do not control the rankings. A recommendation has to make sense even if the affiliate link disappears. We often use direct affiliate programs when the exact product, pricing, and retailer experience are as good as or better than the Amazon listing. The ranking is always based on how well the filter serves the reader, not which program pays more.
