How the HADONE Score works
One number, four inputs, and a rule about when we are allowed to show it at all. This page is the whole calculation — if something here does not match what you see on a product page, the page is wrong and we want to hear about it.
Last updated: September 1, 2026
What goes into it
Every owner answers the same four questions. Their most recent answers are averaged across owners, and the four averages are weighted like this:
| Question | Weight |
|---|---|
| Would you buy it again today? | 40% |
| Does it still work properly? | 25% |
| Do you still use it? | 20% |
| Has it needed a repair? (a repair lowers the score) | 15% |
What the labels mean
The number carries a label so it reads as a judgement rather than a decimal:
- Exceptional 90 and above
- Strong Buy 80 and above
- Good 70 and above
- Mixed 60 and above
- Weak 40 and above
- Avoid under 40
How much data is behind it
The score never appears below 10 rated owners. Above that, it carries a confidence level that says how much is behind it — because 12 owners and 1,200 owners should not look the same:
- Early score from 10 owners
- Moderate confidence from 50 owners
- Strong confidence from 200 owners
- High confidence from 500 owners
One owner, one vote
Only the most recent check-in from each owner counts. Someone who checks in every six months for three years is still one owner, not six — otherwise the most enthusiastic person in the room would quietly outvote everyone else.
Reports flagged during moderation are excluded from the calculation entirely.
Why nothing appears below 10 owners
Under 10 rated owners we publish no score at all — not a provisional one, not a greyed-out one. We show the reports we have and say plainly that there are not enough of them.
A number built on four opinions looks exactly like a number built on four hundred, and the reader has no way to tell them apart. Rather than dress that up with a caveat, we leave the number out.
What is deliberately not in it
Not the price, and not what it costs today. Not affiliate commission — no retailer can pay to move a score, a ranking or a product's place in a list, and no such offer would be accepted. Not the brand. Not our own editorial research, which sits in its own section on the page precisely so it cannot be mistaken for owner data. Not reviews from anywhere else: we do not scrape or import ratings from other sites into the score.
It cannot be set by hand
There is no administrator screen for changing a score, and the database rejects any direct write to the score table: a score can only ever be the output of the calculation above, run over the reports.
So a correction means changing the reports — removing a fraudulent one, unflagging a real one — in a way that leaves a trace, rather than editing the number.
What this score is not
It is not a measure of how good a product is when new, and it is not a lab test. It says how a specific group of owners felt about a purchase after living with it — which is the thing that is hardest to find out before you buy, and the only thing this site tries to measure.
It also cannot predict your individual unit. A product with a high score still has owners whose one broke.
When this changes
If we change the weights, the threshold or the bands, we will say so on this page and date it. A methodology you can only read after it has already been changed is not a methodology.
Reports we transcribed
Owners sometimes send us a report by email instead of filling the form in. We enter those, and they count toward the score exactly like a report submitted directly — but they are marked as ours on the product page, with a count, so you can weigh them yourself.
We only ever transcribe what an owner actually told us. We do not write reports.