It started because a review lied to us
In 2016 one of us bought a laptop on the strength of a glowing review, and it throttled to half its advertised clock speed after eleven minutes of sustained load. The review had run a three-minute benchmark. Nothing in it was technically false; all of it was useless.
PC MAW exists because that gap — between a test that is easy to run and a test that tells you what owning the thing will feel like — is where most technology writing lives. We have spent ten years trying to stay on the right side of it.
Today four full-time editors and a rotating group of contributors publish long-form guides across six areas, and a small services arm does the same work for clients. The two halves feed each other: the client work keeps us honest about what actually breaks in the field, and the publishing keeps the client work from drifting into whatever is fashionable.
Four people, four bylines, one bench
Every article carries the name of the person who ran the tests. Every one of those names has an email address that reaches a human.
Mary R. Hargis
Mary has spent fourteen years writing about consumer technology, eight of them running the PC MAW test bench. She believes a benchmark you cannot reproduce is just an opinion with a chart attached.
Daniel Okonkwo
Daniel builds, breaks and re-builds around forty systems a year. He handles our thermal testing, power-supply teardowns and the long-term reliability logs that most reviews never bother to keep.
Priya Raghunathan
Priya audits WordPress and headless sites for a living. She has taken more than three hundred sites through Core Web Vitals remediation and still keeps a folder of the worst plugin code she has ever seen.
Marco Lindqvist
Marco came to publishing from twelve years of network operations. He writes the home-security and router coverage, and maintains the lab network that everything else here gets tested on.
The rules we hold ourselves to
These are not aspirations. They are the conditions under which we will publish, and we have turned down work rather than bend them.
Read the full disclaimerNo advertiser, sponsor or client sees a draft or influences a verdict. Sponsored content is labelled at the top of the article and in every listing. We have lost at least two five-figure deals over this and would do it again.
Substantive errors get a dated correction note at the foot of the article and, where the error changed a conclusion, a note in the newsletter. We do not silently edit numbers or quietly delete posts.
We use them sparingly and never let them affect a recommendation. Any article containing one says so at the top. Affiliate income is under 4% of revenue and we would rather it stayed there.
We do not sell it, share it or enrich it. The newsletter list is not rented. If you unsubscribe, your record is deleted within thirty days rather than suppressed indefinitely.
We use language models for transcript cleanup and as a research assistant. No article is generated by one, every published sentence is written or rewritten by a named editor, and no benchmark figure ever comes from a model.
Where the numbers come from
Our test room is climate-logged and holds ambient temperature within one degree across a run. Machines sit on the same open bench, fed by the same metered PSU, with power draw logged at the wall. Thermal runs last a minimum of thirty minutes because that is when interesting things start happening.
The network lab is a separate VLAN with its own upstream, so a router under test cannot take the office offline — something we learned the expensive way in 2018.
- Climate-logged bench, ±1 °C across a test run
- Wall-metered power logging on every system
- 30-minute minimum sustained-load thermal runs
- Isolated network lab with independent upstream
- Every raw result archived and published
Get the Monday Byte
Every Monday morning: one deep-dive guide, one tool that earned its place on our bench, and one benchmark number that surprised us. Around 9,400 readers, zero fluff.
By subscribing you agree to our Privacy Policy. One click unsubscribes you for good.