Behind The Scenes Editorial Standards

Inside the Lab: How We Actually Test Every Product

Every score on this site comes out of the same process — whether it's a phone, a blender, or a pair of running shoes. We buy the product ourselves, run it through instrumented benchmarks, live with it like a normal owner would, and only then does it get a number. Here's exactly what that looks like, step by step.

OurComparisons Editorial Team Standards & Methodology · Reviewed annually

11 min read Published Jan 14, 2025 Updated Aug 3, 2026

Products laid out on a testing bench with measurement instruments and notebooks — Our testing bench on a typical week — three products mid-review, two more queued up

Most review sites summarize a spec sheet and call it a review. We don't think that's good enough, and we don't think it's honest — a spec sheet can't tell you whether a laptop's fan whines under load, whether a blender actually crushes ice after fifty uses, or whether a pair of trail shoes still grips wet rock after a month of hikes. That only comes from putting the product through the same paces a real owner would, on the same bench every time, so the numbers we publish mean something when you compare them across hundreds of reviews.

This page is the methodology behind every review, buying guide, and comparison on OurComparisons. It's the same process whether the product costs $30 or $3,000, and it's the reason a 9.2 in one review means roughly the same thing as a 9.2 in another.

We don't score what a product is supposed to do. We score what it actually did, in our hands, over weeks of real use.

At a glance

The four pillars of every review

Every product that reaches a published score has cleared all four of these.

Step One

Receipt and shipping box from a retail purchase

Methodology

Our four-stage process, in order

Nothing gets a published score until it's passed through all four stages below.

We buy it ourselves

No press loaners and no manufacturer-selected units. We purchase at full retail price from the same stores you'd shop at, so we're testing what actually ships.

We run it on the bench

Standardized benchmarks and measurement tools appropriate to the category — the same instruments and test scenes on every unit, so results are comparable across reviews.

We live with it

Two to four weeks of ordinary daily use by an actual staff member, tracking anything a lab test wouldn't catch — annoyances, quirks, and things that only show up over time.

We score it against a fixed rubric

Every category is weighted the same way for every product in its class, so a score in one review means the same thing as that score in another.

The labs

What testing looks like, category by category

The four-stage process stays the same, but the tools and checklists change by product type.

Phones and laptops connected to benchmarking equipment on a lab bench

Electronics & Gadgets Lab

Phones · Laptops · Audio · Smart Home

14checks

BenchmarksCamera Shootouts

Performance is measured with industry-standard tools like Geekbench for CPU/GPU load and Speedtest by Ookla for connectivity, while camera systems are shot side by side against reference devices and cross-checked against public scores from DXOMARK. Battery rundowns are looped three times per unit to average out variance.

What we test

Sustained performance under load, not just peak benchmark numbers
Camera output across daylight, low light, and zoom ranges
Battery life across three separate full-charge rundowns

What we don't rely on

Manufacturer-published spec sheets alone
Single-session benchmark runs with no repeat

3Units tested each

60+ hrsBench time

14-ptChecklist

3 wksReal-world carry

Blender and kitchen appliances being tested with measuring equipment

Home & Kitchen Lab

Appliances · Cookware · Cleaning

11checks

Durability RunsNoise Testing

Appliances go through repeated-use cycles that simulate months of ownership compressed into days — ice-crushing counts on blenders, brew-temperature checks on coffee makers, and decibel readings for noise. Energy draw is logged against ENERGY STAR reference figures where a certification exists.

What we test

Performance consistency across 50+ repeated-use cycles
Noise output measured in decibels, not described subjectively
Cleanup time and part durability after regular dishwasher use

What we don't rely on

A single demo run out of the box
Marketing claims about capacity or power

50+Use cycles

4 wksReal-world carry

dBNoise logged

11-ptChecklist

Hiking shoes and fitness gear being tested on an outdoor trail

Outdoor & Fitness Gear Lab

Footwear · Apparel · Trackers

9checks

Trail MilesWeather Exposure

Gear is worn on actual trail miles logged by staff testers, not treadmills — wet rock, loose gravel, and temperature swings included. Wearable trackers are checked against a chest-strap heart-rate monitor for accuracy, and footwear is inspected for sole wear after a fixed mileage benchmark.

What we test

Grip and durability across at least 40 logged trail miles
Tracker accuracy against a chest-strap heart-rate reference
Comfort and fit feedback from multiple testers, not one

What we don't rely on

Indoor-only or short-distance testing
A single tester's foot shape or fitness level

40+Trail miles

2Testers minimum

3 wksReal-world carry

9-ptChecklist

Spreadsheet tracking software updates and long-term support history

Software & Long-Term Tracking

Updates · Reliability · Repairability

7checks

Re-TestedRepair Scoring

This isn't a one-time visit. We keep a spreadsheet on every product we've reviewed and revisit it after major software updates, price changes, or a discontinuation. Repairability is scored against public teardown data from iFixit wherever a teardown exists.

What we test

Update cadence and how long a product stays supported
Repairability, using third-party teardown data where available

What we don't rely on

A single launch-day snapshot with no revisit

OngoingRe-test cadence

1xPer major update

7-ptChecklist

PublicData cross-check

The math

How the final score is calculated

Five weighted categories, fixed for every product in a given class before testing starts.

Rank	Category	Weight	Measured by	Score contribution	Notes
1	Performance / Core Function	30%	Bench instruments	0–3.0	Category-specific
2	Build & Durability	20%	Repeated-use cycles	0–2.0	Same rig every unit
3	Real-World Experience	25%	Staff carry period	0–2.5	2–4 weeks minimum
4	Value for Price	15%	Category comparison	0–1.5	Checked at publish + retest
5	Support & Longevity	10%	Ongoing tracking	0–1.0	Revisited after updates

Editorial standards

The principles that keep this honest

Four rules we don't bend, regardless of who makes the product.

We buy what we test

Review units are purchased at retail, not supplied by the manufacturer. If a brand insists on sending a unit, we buy our own anyway and use that one.

We retest after changes

A major software update, a price change, or a hardware revision triggers a fresh look, not a footnote. Scores are updated, not just dates.

We publish the losers too

A product that scores poorly gets the same depth of testing and the same detailed writeup as a category winner. Low scores aren't buried.

We disclose affiliate links

Some outbound links earn us a commission if you buy through them. That relationship never influences a score — the rubric is fixed before we know the outcome.

FAQ

Questions readers ask us most

If yours isn't here, reach out and we'll add it.

Do brands pay for a good review or a higher score?

No. Our scoring rubric is fixed before testing begins and applied the same way regardless of who makes the product. Affiliate commissions on purchase links have no bearing on the score a product receives.

Do you ever accept free review units from manufacturers?

Occasionally a brand sends one unprompted, but we purchase our own test unit at retail regardless and that's the one we score. Any manufacturer-supplied unit is disclosed if used for a secondary check.

How do you decide what products to test next?

A mix of reader requests, new releases in categories we already cover, and gaps in our existing comparisons. We don't accept payment to prioritize a product for review.

How often is a review retested or updated?

Whenever a major software update, price change, or hardware revision ships. We also do a periodic sweep of older reviews to check whether anything has become outdated.

Who is actually on the testing team?

A small group of category editors, each focused on one lab — electronics, home and kitchen, outdoor and fitness, or long-term software tracking — rather than generalists covering everything.

Inside the Lab: How We Actually Test Every Product

The four pillars of every review

Bought at Retail

Instrumented Testing

Real-World Carry

Rubric Scoring

Our four-stage process, in order

We buy it ourselves

We run it on the bench

We live with it

We score it against a fixed rubric

What testing looks like, category by category

Electronics & Gadgets Lab

What we test

What we don't rely on

Home & Kitchen Lab

What we test

What we don't rely on

Outdoor & Fitness Gear Lab

What we test

What we don't rely on

Software & Long-Term Tracking

What we test

What we don't rely on

How the final score is calculated

The principles that keep this honest

We buy what we test

We retest after changes

We publish the losers too

We disclose affiliate links

Questions readers ask us most

Curious how a specific product scored?