Our Data & Methods
Every fit claim and price comparison on this site comes from a process you can check. This page documents where our data comes from, how we verify it, and — the part most sites hide — where it is weak.
Where Our Fit Data Comes From
Our compatibility data is compiled from public retailer catalog listings, cross-checked against each other. For every replacement part number, we collect the model-coverage lists published by multiple retailers and compare them line by line. Where the sources agree, the pairing enters our fit tables. Where they disagree, it does not — conflicting pairings are held out of our recommendations entirely rather than guessed at.
We are explicit about what that is: an inference from consistent public listings — and a verifiable one, because the same catalogs we drew from are public. It is not a bench test, and the limitations section below does not pretend otherwise.
How We Verify Fit & Compare Prices
The exact-fit workflow has four gates. A pairing that fails any gate is not published as a recommendation.
1 · Collect
Pull the model-coverage list for a part number from multiple public retailer catalogs — never a single source.
2 · Cross-check
Compare the lists line by line. Keep the intersection (pairings every source confirms). Flag divergences as unverified and exclude them from recommendations.
3 · Map & spot-check
Load confirmed pairings into the Filter Finder fit table, then spot-check stated dimensions — length, width, height — against the catalog values.
4 · Re-verify
Fit tables carry a last-verified date. Catalogs drift and model lines get renamed, so the cross-check re-runs on a cycle.
How the “vs OEM” Price Comparison Is Calculated
We compare against the OEM list price for the same part number — not bundle prices, not “compare at” theater. When a listing says 40–73% below OEM, that is per-SKU arithmetic: our price divided by the OEM list price for that exact part number.
Prices move with supplier costs. Our comparison percentages are re-checked on the same cycle as the fit data, so a displayed percentage can lag the live checkout price by a few days. The checkout price is always the current truth.
Limitations: What Our Data Cannot Do
Honest limits beat confident vagueness. If you rely on our fit tables, rely on them knowing all of this:
Fit data is inferred, not physically tested
Compatibility is compiled from public retailer catalog listings and cross-checked across sources. We do not install filters on appliances to confirm fit. A pairing in our tables means multiple independent listings agree — a strong inference, not a bench test.
We do not run independent filtration tests
Contaminant-reduction claims, flow rates and filter-life figures repeat what the cartridge’s own certification documents state. We verify that the stated certification exists and matches the SKU; we do not re-test the chemistry.
Prices move without warning
Supplier costs change between our verification cycles. The checkout price is always the live truth; the “vs OEM” percentage on a listing can lag it by a few days.
Coverage has edges
Fit tables reflect catalog data as of the last-verified date. Models released in the current model year may not yet appear, and a small tail of older models is absent from every catalog we can access.
Sources can be wrong
Retailer catalogs contain errors, lag model renames, and occasionally list the wrong part for the right appliance. When a confirmed error reaches us we correct it and note the correction date on the affected listing — we cannot guarantee we have caught them all.
About the “ClearFit Filters Technical Team” Byline
Guides and fit tables on this site publish under one byline: ClearFit Filters Technical Team. It is the in-house team that compiles the catalog data, runs the cross-checks described above, and maintains this page.
When an article states a measurement, a price comparison, or a compatibility claim, it has passed through the workflow on this page. Corrections, once confirmed, are noted on the affected page. If something looks wrong, send us the listing URL and we will re-run the check.