How We Research and Recommend
This page explains how our guides are researched, what we refuse to claim, and the stricter standard a hands-on review would have to meet. If a guide on this site ever fails the standard below, it is a bug and we want to hear about it.
- Sources cited, not implied
- No tests we have not done
- Corrections published, not buried

Two evidence standards, and the one we use
Every guide on HomeNerdy is research-led: the how-to guides and the buying guides in Nerdy Reviews alike. The hands-on review standard sits beside it so you know what a product review would have to show before it could claim a test. No product review meets it yet.
| Research-led guide | Hands-on product review (none yet) | |
|---|---|---|
| Built from | Manufacturer documentation, industry standards, professional consensus, documented user outcomes | Hands-on use of a product we bought ourselves |
| Claims hands-on testing | No | Yes, and says how long for |
| Typical language | “per manufacturer guidance”, “commonly recommended”, “widely reported” | “we ran this for six weeks and…” |
| Label | Category badge | Hands-on review label |
When one of our writers runs a test on a method inside a research-led guide, or has direct hands-on experience with it, we say so in the guide, at the point it is relevant, and describe what we did. We never imply it site-wide. Two guides report a test we ran ourselves: six methods on yellowed plastic and five stain methods on a Ninja blender jar.
How a guide gets made
- Research, before a word is written
Product manuals, safety data sheets and official manufacturer recommendations first. Then the standards: EPA, OSHA and the relevant trade associations. Then what licensed professionals (plumbers, electricians, cleaning specialists) actually recommend. Last, the documented real-world outcomes, including the common mistakes.
- Diagnose before you prescribe
Every guide opens by helping you identify which problem you actually have. A limescale mark and an etched surface look similar and need opposite treatments; a guide that skips the diagnosis will damage something.
- Order methods gentlest first
You should never open with the nuclear option. Methods escalate, and each one names the materials it is safe on and the ones it is not.
- State the limits
Every method carries what it cannot do. If a popular hack has limited effectiveness, that goes in before you waste an afternoon on it.
- Editorial review against the checklist
Nothing publishes until it clears the checklist below.
The publishing checklist
Every guide clears all seven before it goes live. This is the same list our editor works from.
- Safety warnings present, prominent, and above the method they apply to
- No absolute claims: no “always works”, no “guaranteed”
- Sources cited for every technical and safety claim
- Limitations stated for each method, not just the article as a whole
- A “when to call a professional” section, with a real threshold
- Materials compatibility verified against manufacturer guidance
- The evidence basis is labelled: research-led, or hands-on
How a buying guide chooses its picks
The buying guides in Nerdy Reviews are research-led like everything else here. None of them claims that we used the products.
| Step | What it means |
|---|---|
| Criteria first | The guide sets out what matters for that product, and why, before it names a pick. |
| A minimum bar | A product that fails it is out, however popular it is. |
| Sources, weakest labelled | Manufacturer specifications and manuals, independent certifications where the category has them, and verified owner reviews, which show what owners commonly report rather than what is established. |
| What we did not do | Stated in the guide’s How We Chose section. |
Six of the eight buying guides carry a How We Chose section today; the nursery purifier and fireplace grate guides predate that format.
What a hands-on review would have to show
We have not published a hands-on product review yet. If we do, it has to meet all of this, and it will be labelled Hands-on review.
- Bought at retail price from an ordinary shop: no free products, manufacturer samples or press units
- Used in the conditions you would use it in, not one demo run
- Judged over weeks, and the review says how long
- No sponsorship and no copy approval
What we will not do
- Accept payment, free products or samples in exchange for coverage
- Let a brand see, approve or influence a guide before it publishes
- Claim hands-on testing we have not done
- Publish an absolute claim, however good it would look in a search result
- Recommend a method we would not use on our own home
- Quietly re-date an old guide to make it look current
Corrections
We will get things wrong. When we do, the correction goes on the guide itself, dated, saying what changed. We do not delete the error quietly and we do not bury the note at the bottom of an unrelated page.
If you have found something wrong, especially anything with a safety consequence, tell us and we will look at it. That is the fastest way to make this site better and we treat it as a favour, not a complaint.
Found something on this site that fails the standard above?Safety problems go to the top of the queue.
Report it to the editor