Every figure and every claim this site has got wrong, what it said before, and what it says now.
Last reviewed September 7, 2026 by the developer who builds and maintains this site. First published. Every entry was written from the decision record for the change it describes, and the before figures were read back out of the superseded code rather than recalled.
What this page is
There are 11 corrections here, 5 of which could have changed what somebody bought or decided. Each says what was wrong, what a visitor would have seen, what they see now, and why the new figure is right.
The old value is stated every time. A correction that does not say what it replaced cannot be checked by the person who relied on the old number, and that person is the only one it is for.
Most of what follows was found during a full audit of the calculators rather than reported by a reader, and was corrected before the site carried advertising. That is said here rather than left for you to assume either way: these are real defects that were really in the code, and the log exists because the next one will not be caught by an audit.
If you find a figure that looks wrong, [email protected] reaches the person who can check it, and a reply comes within two working days. What to include is on the contact page.
Who publishes this site
Could have changed a decisionCorrected September 7, 2026 across the site
The about page and the editorial policy described a group of people with backgrounds in residential construction and cost estimating. Nobody by that description has ever worked on this site.
- What it said before
- "Nest Quantum was built by a small team with backgrounds in residential construction, cost estimating, and software development", and "the people who write our calculator explanations and guides have direct experience with the kind of projects these tools cover".
- What it says now
- Both pages say the site is built and maintained by one developer, and the editorial policy states that the earlier description was of nobody and is withdrawn.
A credential a reader cannot check is worth less than a method a reader can. The claim was removed rather than softened, and what replaced it is the thing that can be verified: every constant, its value, its source, and the date that source was read.
Window replacement energy savings
Could have changed a decisionCorrected September 7, 2026 on /calculators/window-replacement-cost-calculator/
Annual savings were calculated per window, at up to $160 each, with no reference to the house or the climate. ENERGY STAR, which the page cited, publishes whole-home figures three to ten times smaller.
- What it said before
- Twelve windows reported about $1,560 a year in savings and a payback of 4.8 years.
- What it says now
- The same job reports $117 to $186 a year against a $7,524 cost, so a 40 to 64 year payback, and the page says plainly that windows are rarely paid back by energy savings alone.
Savings do not scale with the number of windows independently of the house, and they depend far more on what is being replaced and on the climate than on which new glass is chosen. The model is now a whole-home range from the ENERGY STAR city table, scaled by the share of the home being replaced, and it needed two new inputs to be answerable at all. A test now fails if the shortest payback ever drops below 15 years.
Source: Cost and energy savings estimates for ENERGY STAR qualified windows, U.S. cities table, ENERGY STAR, U.S. Environmental Protection Agency and U.S. Department of Energy, published 2005, read 2026-09.
Heat pump running cost and savings
Could have changed a decisionCorrected September 7, 2026 on /calculators/heat-pump-savings-calculator/
The seasonal heat loss coefficient was 0.0005 BTU per hour per square foot per degree, about 330 times too small, and the annual saving was clamped so it could never be negative.
- What it said before
- Heating costs were reported at a small fraction of any real bill, and a heat pump that costs more to run than the system it replaces reported $0 saved rather than a loss, beside a ten-year figure of negative $13,000.
- What it says now
- The coefficient is 0.165, calibrated against measured national consumption. Where a heat pump costs more to run, the headline reads "Annual cost increase", the payback reads "Does not pay back", and a note explains that the electricity-to-fuel price ratio is what decides it.
EIA's Residential Energy Consumption Survey measures 34.4 million BTU of site energy per household for space heating, over a 1,614 square foot average heated area at about 4,300 heating degree days. Clamping a negative saving to zero hides the answer on the page where a reader is deciding whether to spend five figures.
Source: Residential Energy Consumption Survey 2020, Table CE3.1: annual household site end-use consumption, U.S. Energy Information Administration, published 2024-03, read 2026-09.
Cooling cost load factors
Could have changed a decisionCorrected September 7, 2026 on /calculators/cooling-cost-calculator/
The cooling load factors were a thousand times too small, and the calculator's own test file contained a different value from the calculator.
- What it said before
- A typical home reported an annual cooling cost of around $1.34.
- What it says now
- The factors are calibrated so that a national-average home reproduces EIA's measured 2,056 kWh a year for air conditioning, to within 1.3 per cent.
Neither the shipped value nor the one the test expected survived contact with published data: the test's value would have implied roughly five times the measured national average. The table is now derived from EIA consumption, average cooled floor area and typical installed efficiency, rather than chosen to make a test pass.
Source: How much electricity is used for air conditioning in the United States?, U.S. Energy Information Administration, published 2026, read 2026-09.
Payback periods that never arrive
Could have changed a decisionCorrected September 7, 2026 across the site
Six payback figures across five calculators returned either infinity or zero when the investment never pays back, and both rendered as "0 years".
- What it said before
- A solar array or an efficiency upgrade that never repays its cost showed a payback of 0 years, which reads as an instant return.
- What it says now
- All six report "Does not pay back", and a test asserts that string can never contain a zero.
This was the exact opposite of the truth on the two pages where a reader is deciding whether to spend five figures. An unfavourable answer is stated rather than hidden.
Heating cost load factors
Changed a quantity or a costCorrected September 7, 2026 on /calculators/heating-cost-calculator/
The heating load factors overstated measured household consumption by about 2.1 times.
- What it said before
- An 1,800 square foot home at 5,000 heating degree days with a 0.96 AFUE furnace at $1.20 a therm reported about $1,125 a year.
- What it says now
- The same home reports about $450 a year.
At EIA's own national averages the calculator predicted 72.3 million BTU of site energy where the survey measures 34.4. The table is recalibrated the same way as the heat pump coefficient, and the two calculators now agree with each other and with the survey. Propane and heating oil energy content were corrected at the same time, from 91,500 and 138,700 BTU per gallon to EIA's 91,452 and 138,500.
Source: Residential Energy Consumption Survey 2020, Table CE3.1: annual household site end-use consumption, U.S. Energy Information Administration, published 2024-03, read 2026-09.
Decking board count
Changed a quantity or a costCorrected September 7, 2026 on /calculators/decking-calculator/
Every row of decking was charged a whole board and every offcut was scrapped, so the calculator over-ordered by roughly two thirds.
- What it said before
- A 16 by 12 foot deck reported 696 linear feet of decking for a surface needing 416.
- What it says now
- Offcuts are carried between rows, which is what happens on site. Pieces shorter than two feet are still treated as scrap, because a butt joint has to land on a joist.
Over-ordering by two thirds is not a conservative estimate, it is a wrong one, and it is money spent on timber that goes in a skip.
Plant spacing cost
Changed a quantity or a costCorrected September 7, 2026 on /calculators/plant-spacing-calculator/
The calculator told you how many plants to buy including a 10 per cent overage, then costed the number without it.
- What it said before
- A job that told you to buy 47 plants quoted you for 42.
- What it says now
- The cost covers the quantity you are told to buy.
Every other calculator on this site costs the quantity you buy rather than the quantity that ends up installed. This one disagreed with the rest of the site and with its own page.
Kentucky bluegrass seeding rate
Changed a quantity or a costCorrected September 7, 2026 on /calculators/lawn-seed-calculator/
The seeding rate for Kentucky bluegrass was 25 per cent above its published range.
- What it said before
- 2.5 pounds per 1,000 square feet for a new lawn.
- What it says now
- 2.0 pounds per 1,000 square feet, the top of the published band.
Purdue Extension gives 1.5 to 2.0 pounds per 1,000 square feet for 100 per cent Kentucky bluegrass on a new lawn. The top of the range rather than the middle, because running slightly heavy on a new lawn is the forgiving direction.
Source: AY-3-W: Establishing Turfgrass Areas From Seed, Table 2, recommended seeding rates, Purdue University Extension, published 2005, read 2026-09.
Logged and maintained by the developer who builds and maintains this site. One person is answerable for every figure on this site, which is the point of publishing where each of them came from. Corrections go to [email protected] and get a reply within two working days.