Every verdict published on WaterHeaterLab comes out of the same three-step process. It is not complicated, but it is consistent: the same documents, the same math and the same questions, applied to every water heater category we cover. This page explains how the process works, what it can tell you and where it stops.
Step 1: Collect the specs
We start with documents, not impressions. For every model class in a category — 50-gallon gas tanks, condensing tankless units, hybrid heat-pump water heaters — we pull the manufacturer’s data sheets, the DOE/ENERGY STAR Uniform Energy Factor (UEF) listings and the full warranty documents, then log them side by side. For each model class we record:
- Capacity and recovery — tank size, first-hour rating or flow rate.
- Efficiency — the certified UEF figure, not the marketing headline.
- Warranty — length in years, and what the fine print actually covers: tank, parts, labor.
- Installation drivers — venting, electrical service, condensate drainage, clearances: the requirements that quietly raise the installed price.
Logging everything in one table keeps the comparison honest. No model gets credit for a spec its competitors were never scored on.
Step 2: Model ten years of ownership
The sticker price is the smallest part of the story. For each model class we build a ten-year cost of ownership: the equipment price and the energy the appliance will burn. We leave installation labour out of it on purpose, because labour is local and any national figure would be a guess dressed up as a number. The models run on one fixed set of assumptions so that no two pages on this site can disagree with each other: 64 gallons of hot water a day, a 70 degree Fahrenheit temperature rise, natural gas at $1.50 a therm and electricity at $0.17 a kilowatt-hour, over ten years. That works out to roughly 136 therms, or 4,000 kilowatt-hours, of delivered heat a year, which each appliance then has to supply at its own efficiency.
Incentives get named, never assumed. The federal 25C tax credit for heat pump water heaters expired on 31 December 2025, so nothing on this site models it as still available. Where a utility or state program applies we say that it applies and leave the money out of the total, because those schemes vary by zip code and close without notice, and a total that quietly depends on one is not a total you can act on. The point of the ten-year model is simple: a cheap sticker can’t hide an expensive decade, and an expensive sticker gets the chance to earn its keep.
Step 3: Publish the verdict
The output is a plain-English recommendation by household type: the family of five running two showers and a dishwasher before 7 a.m., the couple in a two-bedroom condo, the cabin used ten weekends a year. Each verdict names a first pick and a runner-up, and explains in a few sentences why one beat the other. When the numbers are close, we say so instead of manufacturing a dramatic winner.
What we don’t do
- No sponsored rankings. A manufacturer cannot buy a position in any comparison.
- No referral fees from installers or manufacturers.
- Advertising stays separate. Display ads on the site are served independently of our editorial verdicts; advertisers have no influence on them and no advance notice of them.
- No installation tutorials or DIY repair walkthroughs. We help you choose the right unit; a licensed professional should install and service it.
About the images on this site
The pictures illustrating our comparisons are computer-generated images, not photographs of real products. They exist to show a type of appliance — a storage tank beside a tankless unit, a heat pump model beside a gas one — at a consistent scale and under consistent light, which is something stock photography of water heaters simply does not offer.
They deliberately carry no brand marks, model names or rating labels, because the equipment they depict is generic rather than a specific unit you can buy.
Each guide also carries a second, closer image: a cutaway, an overhead, a part on a bench. Those are generated too, and they need the same warning twice over. When you see a scaled-up heat exchanger or a corroded anode rod on this site, you are looking at a drawing of a failure mode, not a photograph of one we found. It is there to make a mechanism legible, not to prove that it happened.
Treat all of them as diagrams, not as evidence: every figure that matters — capacity, efficiency, warranty, ten-year cost — comes from the documents described above, never from a picture.
Updates and corrections
Prices move, tax credits get revised, and manufacturers retire models mid-year. Every comparison and guide carries a “Last updated” date so you can see when its numbers were last checked, and we revise pages when the underlying data shifts.
If you spot an error — a stale price, a discontinued model, a spec that no longer matches the current data sheet — email us at 
or use the contact form. We correct verified mistakes and note material changes on the page itself.
The limits of the method
Our ten-year models use national averages. Your local electricity and gas rates, your water hardness, your climate and the complexity of your installation can all shift the numbers — sometimes enough to change the verdict between a first pick and its runner-up. Treat our figures as a structured starting point, not a quote.
Everything on WaterHeaterLab is informational only. Before you buy, get written quotes from licensed local installers, and leave installation and repair to a licensed professional — local codes vary, and permits and inspections are part of doing the job right. If you want to know who is behind this method, see our About page.
Last updated: August 19, 2026