Skip to content
Key & Crate

How we pick

We do not run a test lab, and we will not pretend otherwise. Here is exactly what our picks are based on instead, how the scoring works, and where our judgment can be wrong.

By Mike M.Last updated

What we have not done

Starting here, because it is the part most buying sites are vague about.

We have not lab-tested these products. There is no test kitchen, no workshop, no thermal camera, no instrumented rig, and no panel of experts. We have not cooked on every cookware set on this site, slept on every mattress topper, or timed how long every towel takes to dry.

Any site claiming to have hands-on tested forty products across nine categories is either a large operation with a real budget, or it is not telling you the truth. We are one person, and the honest thing is to say what we actually do.

So we compete on something else: a published, reproducible method. You can read exactly how a pick was reached and disagree with the reasoning, which you cannot do with “after six weeks of rigorous testing.”

What our picks are actually based on

  1. Published manufacturer specifications. Dimensions, materials, capacity, wattage, oven-safe temperature, whether the base is induction-compatible, whether the basket is dishwasher safe. These are verifiable facts and they decide more purchases than people expect — a 4-quart air fryer either fits your counter or it does not.
  2. The physical design, reasoned about. Whether a cookware set nests. Whether a dish rack drains into the sink or into a tray you have to empty. Whether a shower caddy is mesh on all sides. Whether a clamp jaw opens wide enough for a lofted bed rail. These are design facts with predictable consequences, and reasoning about them is legitimate.
  3. Patterns in long-term owner reports. Not a star rating — the recurring specifics that appear in reports after a year or more. “The latches broke after three years” appearing repeatedly across a product is signal. A five-star rating from someone who has owned it a week is not.
  4. Material behavior that is simply known. Nonstick coatings have a service life of two to three years. Memory foam traps body heat. Coated steel chips at the welds and then rusts. Glass boards are harder than knife steel. None of that requires testing; it requires not pretending otherwise.
  5. The reader’s actual constraint. This is the one that most changes our rankings against other sites. A reader setting up a first apartment is constrained by cabinet space, a security deposit and a budget covering forty items — not by thermal conductivity. A set that nests into one cabinet beats a set that heats 4% more evenly, and we rank accordingly.

The scoring, and why these four metrics

Every product carries four scores out of ten and an overall figure. The four metrics change by section, because what matters in a bathroom is not what matters in a tool kit — and they are chosen for a mover in a small space on a budget rather than for a spec enthusiast.

The metrics by section
SectionThe four metrics
KitchenFits a small kitchen · Value · Durability · Easy to clean
Bath & laundryFits a small bathroom · Value · Durability · Renter-safe
DormFits a dorm · Value · Durability · Moves out easily
Tools & fix-itBeginner-friendly · Value · Durability · Stores small

Two things to note about these. “Renter-safe” appears as a standing criterion in the bathroom section rather than as a footnote, because our reader has a landlord and a deposit, and most buying guides write as though everyone owns their walls. And “Moves out easily” is a dorm metric because everything in that room gets carried out again in eight months.

The scores are our editorial judgment, shown component by component so you can see where a product lost points and decide whether you care about that metric. A set that scores 9.6 on fitting a small kitchen and 5.6 on durability is the right buy for a one-year lease and the wrong buy for a house, and a single aggregate number hides that.

These scores are never published as review markup. They appear on the page as content, and we deliberately do not emit them as product-review or star-rating structured data, because doing so would tell search engines we had conducted product reviews of a kind we have not.

How prices work, and why they expire

Every price on this site comes from a live Amazon lookup and is shown with the date it was checked. None is typed in by hand, because a hand-typed price is a price that will be wrong within a month.

More importantly: prices expire on their own after 48 hours. The daily refresh re-checks every product; if that refresh stops working for any reason, every number on the site disappears and the pages fall back to “check price on Amazon” instead.

That design is deliberate. The failure mode of this site is silence, not a stale claim. We would rather show you nothing than show you a price that was true last month, and the system is built so that nobody has to remember to make that happen.

Where we quote a total — “about $190 for a budget kitchen” — that is derived from the same live prices and is approximate by nature. Treat those as planning figures, not quotations.

How products get onto the site at all

Every product we feature is a real, currently-sold model, identified through Amazon’s catalog and pinned to a specific product identifier. We do not write about products we cannot link to, and we have never invented a product, a model number, a specification or a price.

We generally feature between three and five options per category, because the useful job is narrowing rather than listing. Where we think a category needs only one answer, we say so and explain why.

We also deliberately include known brands where a known brand is the right answer, and lesser-known ones where the specification genuinely wins. A budget set that nests into one cabinet is sometimes a better recommendation for this reader than a respected brand that does not.

Where our judgment can be wrong

Being specific about this is part of the method, not a disclaimer.

  • Unit-to-unit variation. Specifications change silently between production runs, particularly on budget products. A pick that was right six months ago can be made differently today.
  • We cannot assess how something feels. Whether a towel is pleasant against your skin, whether a particular foam suits your back, whether a handle fits your hand. Those are real factors and we can only reason around them.
  • Long-term durability is inference. We read what owners report after a year or more and look for repeated specifics. That is better than guessing and weaker than having owned it.
  • Owner reports are a biased sample. People write about things that failed and things that delighted them, less about the adequate middle. We weight repeated specifics and discount volume.
  • Our weighting is a judgment about you. We assume a small space, a tight budget and a lease. If you have a large kitchen and money, our ranking is wrong for you and the component scores will show you why.

When pages get updated

Prices update daily and automatically. Content gets revisited when a pick is discontinued, when a specification changes, when a reader sends a correction, or when a category changes enough to change the recommendation.

Every page shows its last updated date prominently, because a buying guide without one is asking you to trust that it is current. Corrections to factual errors are made on the page itself rather than quietly removed.

If you think something here is wrong, tell us. The full standards we hold ourselves to are in the editorial policy, and the disclosure page covers how the site is funded.