We get this question more than almost any other: How do you actually test the things you write about? It usually arrives in an email from a reader who liked a review but wanted to understand what stood behind it, or occasionally from someone who disagreed with a verdict and wanted to know whether we'd really used the product at all. Both are fair questions. This page is our answer to both of them.
Consumer Finds is a small, independent editorial operation. We are not a laboratory. We do not have a testing floor full of calibrated instruments and white-coated engineers. What we have is a group of people who genuinely enjoy using things and figuring out whether they work the way their makers claim they do — and a set of rules we hold ourselves to so that preference and personal style don't quietly substitute for honest evaluation.
Here is how it all works, from the moment a product lands on our radar to the moment a review goes live.
How we decide what to test
Our testing list is driven by two things: reader questions and our own actual needs. Both are deliberate choices.
Reader questions are the cleaner signal. When several people ask the same thing — "Is the $40 version of this thing meaningfully worse than the $90 version?" or "Has anyone actually tested whether that kitchen gadget works on cast iron?" — it tells us the market has created genuine confusion that a careful review might resolve. We keep a running log of these questions and revisit it every few weeks when we're planning coverage.
Our own needs are the second filter. If someone on the team is genuinely in the market for a particular category of product, that creates a natural testing situation: we're motivated to find the real answer rather than a tidy one, because we're going to live with whatever we buy. Some of our best reviews have started with "I actually need one of these and I'd rather test five of them than guess."
We also deliberately keep our testing list shorter than we probably could fill. Testing something properly takes weeks. We'd rather do fifteen reviews a year that we stand behind than fifty that we rushed.
How we acquire the products we test
In the majority of cases, we buy the products ourselves, at retail, through whatever channel a regular consumer would use. We pay what any buyer would pay. No bulk pricing, no review discounts, no advance samples obtained through back channels. The receipt sits in our files.
Occasionally a brand will offer us a loaner — typically in categories where the retail price is high enough that purchasing six competing units for a comparison would be impractical. When that happens, we have a firm policy:
- The loaner must be disclosed clearly in the article, at the top, before any evaluation language.
- We accept no conditions on the review. If a brand attaches strings — "please send us the draft before you publish" or "we'd appreciate a chance to respond to any negative findings" — we decline the loaner and either purchase the unit or skip the product.
- Loaners are returned when testing is complete. We do not keep them.
- A loaner does not earn a product any extra benefit. It just means the box arrived differently than usual.
In practice, most of what we test is self-purchased. It keeps things simple.
The testing process itself
Testing at Consumer Finds has two phases: a structured protocol phase and a longer unstructured "just use it" phase. Both matter.
The structured phase
During the first week or two with any product, we run it through a fixed set of tasks that we define before we open the box. The exact tasks vary by category — what we measure for a portable speaker is not what we measure for a kitchen tool — but the principle is the same across all of them: define the task in advance, run it the same way multiple times, and write down what happens.
For any category involving speed or output, we time things and we measure things. A charge time gets clocked at least three times and we note the variance. A cleaning radius gets measured with a tape measure, not eyeballed. A battery claim gets checked against our actual runtime under realistic use conditions, not manufacturer-specified ideal conditions. We keep a notes document for each tested product that logs these numbers as we go. The numbers in our reviews come from those documents, not from the product's marketing page.
We also deliberately use products in the conditions they were designed for, plus one or two conditions they weren't. We want to know where the edges are. A product that works perfectly in ideal circumstances but falls apart the moment a real-world variable enters the picture is telling us something important about how it was designed — and about how honest the product page is being with the people who will actually buy it.
The "live with it" phase
The structured phase tells us what a product does. The longer phase tells us what it's like to own one.
We use every product we test for at minimum three weeks as a normal part of daily life — not as a test subject we pick up and put down deliberately, but as the thing we'd reach for if we'd already bought it and just owned it. This is where durability questions start to answer themselves. This is where we notice the design decision that seemed fine on day one and becomes annoying by day twelve. This is where the battery life we measured in the protocol phase meets the charger we keep forgetting is in the wrong room.
The best stress test for any product is six weeks of ordinary use by someone who has stopped trying to evaluate it and is just trying to get things done.
For products that are built around long-term outcomes — anything that's supposed to change a habit, improve a space over time, or hold up through repeated use cycles — we extend this phase to six weeks or longer. Some of our category reviews have involved four-month live-with-it windows. We don't rush these.
Our scoring framework
We don't publish a numbered score for every product. We find that decimal scores often give false precision to what is genuinely a judgment call. Instead, we evaluate products against five questions, and our written verdict reflects how a product does across all five:
- Does it do the job it claims to do? This is the fundamental question and it carries the most weight. A product that fails its core promise doesn't get redeemed by nice packaging.
- How does it hold up over time? Performance on day one and performance on day forty-five are often two different numbers. We try to report both.
- Is the value proposition honest? Price is relative, but we think about whether a product delivers meaningfully more than cheaper alternatives in the same category — or meaningfully less than more expensive ones.
- How difficult is it to actually use? Ease of use is not about whether something is "simple" in the abstract. It's about whether a normal person, without reading the full manual, can get it working correctly and keep it working correctly.
- Does the marketing accurately describe what the product is? We read the product page and the box copy before we test anything, and we compare what was claimed against what we found. Products that overpromise go into our review that way.
We do not give special credit for aesthetics, for how good the unboxing experience was, or for how impressive the product looks in a photograph. These things may be genuinely pleasant, but they are not what determines whether a product earns its place in someone's home.
When we don't publish
We skip products more often than we publish reviews of them. This is worth explaining because it's a meaningful part of how we think about coverage.
If a product is mediocre — not broken, not a scam, just thoroughly unremarkable — we often decide not to write about it at all. The internet does not need another article that says "it's fine, we guess." A reader who arrives at a review looking for a real answer deserves a real answer, not a shrug with a buy button attached.
We also skip products when we can't find a meaningful angle that would help a reader decide. If the answer to every question about a product is "basically the same as three other things we've already covered," a new review doesn't add value.
The exception: if a product is actively bad in a way that could waste a reader's money or disappoint them in a predictable way, we will publish that finding even if we'd otherwise skip the category. Negative reviews serve readers. We write them when we've earned the verdict through actual use.
Our corrections policy
We get things wrong. Not often, but it happens — a specification we quoted was later updated by the manufacturer, a measurement we took under one set of conditions doesn't hold in another, or a product we reviewed was quietly revised after our testing window closed.
When we find an error, or when a reader points one out, we correct it promptly and note the correction at the bottom of the affected article with the date it was made. We do not silently edit articles and pretend they always said the corrected version. The original finding and the correction both appear, so a reader who remembers the prior version can see exactly what changed.
If a correction is significant enough to change our recommendation, we say so explicitly. A correction that changes our verdict gets a prominent note at the top of the article, not just a quiet line at the bottom.
Manufacturers sometimes contact us after publication to dispute our findings or provide additional context. We read all of these carefully. If they present new factual information that changes our understanding of the product, we investigate and update accordingly. If they are asking us to soften a negative finding without presenting new evidence, we decline.
How affiliate links work — and why they don't change our verdicts
Some links in our articles are affiliate links. When you click one and make a purchase, we receive a small commission from the retailer. This is how most independent editorial sites like ours cover their costs, and we want to be straightforward about it.
A few things worth understanding about how this works in practice:
- Your price doesn't change. Affiliate commissions are paid by the retailer from their own margin. You pay exactly the same price whether you arrive through an affiliate link or by typing a web address directly.
- We do not have revenue-share agreements with specific brands. Our affiliate relationships are with large general retailers, not with individual product manufacturers. This means we earn roughly the same commission whether you buy the product we recommended or the one we said wasn't worth it — which removes most of the financial incentive to push any particular product over another.
- Brands have no approval over our copy. No brand we cover has ever seen a review before it published, and none ever will. Commission income does not buy editorial influence here.
- We don't add affiliate links to products we wouldn't recommend. If a product doesn't earn a positive verdict, we don't add a purchase link to it. There is no financial reason for us to send you somewhere we think you'd be wasting your money.
Affiliate income is real and we're not pretending otherwise. But the structure of how it works — same rate across categories, no brand-level relationships, no copy approval — means the incentive to manipulate a verdict is much smaller than many readers assume.
A standing invitation
If you think we got something wrong, we want to hear about it. Not a polite "have you considered" — an actual counter-argument, a link to a spec sheet we missed, a note that your experience with a product was meaningfully different from ours. We read every email that arrives at support@consumer-finds.org and we respond to most of them.
If you have a product you think we should look at, we're interested in those tips too. Especially if it's something you bought yourself, used for a while, and have real opinions about — that's exactly the kind of signal that ends up on our testing list.
We built Consumer Finds because we wanted product coverage that treated readers like adults: specific enough to be useful, honest enough to be trusted, and independent enough to say something true even when the true thing isn't what anyone's paid to say. This page is our attempt to show our work on that. Hold us to it.