NOAA and USDA publish all of this for free. Getting it into a table you can join to your own data takes a weekend and a tolerance for fixed-width station files. That work is done here.
Pay, download, done. No account to make and nothing recurring. Build whatever you like on any of it, commercially, with no royalty and no seat count. Redistributing the raw files is the one thing that differs between packs: the weather data has no restriction at all, the frost pack asks you not to resell the file as-is, and the HVAC pack forbids reselling its files as a competing dataset. Each pack carries its own licence and that file governs. The terms spell out the difference.
Monthly normals for 40,306 ZIP codes. 483,672 rows. Reuse it however you like.
Average high, low, mean, precipitation and snowfall for all twelve months across 40,306 US ZIP codes, from the NOAA 1991-2020 Climate Normals. CSV and JSON, plus the station table. Every row carries the source station, its distance, and a confidence field, so you can see which ZIPs are well covered instead of guessing.
40,502 ZIP codes: hardiness zone, frost dates, growing season.
40,502 US ZIP codes with the 2023 USDA hardiness zone, average last spring and first fall frost dates at the 28F, 32F and 36F thresholds, and growing-season length. Built from USDA PRISM and NOAA public data. Ships with a planting-offset table covering 27 vegetables, a 60-row sample, and a README that lists the coverage honestly: 28,213 ZIPs high confidence, 11,621 medium, 296 low or worse. Read the README before redistributing the raw file.
Winter 99% and summer 1% design temps for 32,865 ZIP codes.
Winter heating and summer cooling outdoor design dry-bulb temperatures for 32,865 US ZIP codes, computed from 576 NOAA GHCN-Daily stations over 1991 to 2020. The numbers a load calculation starts from. Checked against 20 published station figures: average error 1.2F on heating and 1.3F on cooling, worst single station 3F. Ships with the methodology, a data dictionary and the validation table, so you can audit the claim instead of trusting it. Not for permit-stamped work. Read LICENSE.txt inside the pack before redistributing anything.
All three datasets. Cheaper than buying two.
All three ZIP-level datasets in one purchase: typical monthly weather, frost dates and planting zones, and HVAC design temperatures. $39 instead of $57.
What each marketplace really takes out of a sale.
One page you can keep open while you price something, showing what eBay, Etsy, Amazon, Mercari, Poshmark, Depop, Facebook Marketplace and Shopify actually take out of a sale once the payment processing cut is counted. PDF.
Print it or type into it. Dose, weight, side effects, questions to ask.
A weekly log for dose, weight, side effects and the questions worth raising at the next appointment. Two PDFs: one to print and write on, one fillable so you can type into it and keep it on your phone.
Checkout is handled by Stripe, and the Stripe account is registered as Ace Systems. So that is the name on the payment page and the name that shows up on your card statement, not The Boring Thing. Same business. Worth knowing before you see it and wonder.
The assembly, not the data. The data is free and always will be. What costs money is that somebody already reconciled 6,450 NOAA stations against 40,306 ZIP centroids, kept the leading zeros, left missing values blank instead of guessing them, and wrote down how far each ZIP sits from the station its numbers came from. Every row tells you how much to trust it.
If you would rather do it yourself, the READMEs inside each pack name every source URL and every variable code. Go ahead. That is the point of public data.
Two of these datasets already run as free lookup tools, so you can check the numbers against a ZIP you know before spending anything:
Same data, same stations. The pack is the whole country in one file.
The Boring Thing, Fredericksburg, Virginia. Questions to hello@whathitmyhouse.com. Last updated 17 August 2026.