The outlets we cite
Every placement on this site rests on cited sources. This page counts them by outlet and creator, and shows how often each source's stance sits close to the tier we published. It describes agreement. It does not grade anyone.
In numbers
- 20,133 named outlets and creators, across 64,639 citations that carry a stance.
- 53% of those stances sit within one tier of the tier we published, and 18% match it exactly.
- A source reads a product more favourably than our tier in 49% of citations, and less favourably in 33%.
How close the sources sit
Of the 839 outlets we cite on at least 10 products, this is how many fall in each band of agreement.
| Stances within one tier | Outlets |
|---|---|
| Under 40% | 138 |
| 40% to 59% | 436 |
| 60% to 79% | 235 |
| 80% or more | 30 |
The 40 outlets we cite on the most products
| TechRadar | 386 | 91 | 52% | 49% | 32% |
| Tom's Guide | 318 | 92 | 59% | 46% | 29% |
| Pcmag | 272 | 77 | 56% | 46% | 34% |
| CNET | 242 | 82 | 56% | 45% | 36% |
| WIRED | 195 | 84 | 56% | 36% | 43% |
| Verified Reviews | 169 | 67 | 56% | 47% | 36% |
| Tom's Hardware | 168 | 32 | 59% | 52% | 30% |
| Reviewsinside | 161 | 50 | 46% | 63% | 21% |
| Forbes | 159 | 71 | 54% | 40% | 39% |
| Trusted Reviews | 153 | 56 | 54% | 63% | 20% |
| Good Housekeeping | 139 | 80 | 50% | 43% | 36% |
| Wirecutter | 134 | 77 | 55% | 42% | 44% |
| PC Gamer | 123 | 34 | 54% | 53% | 31% |
| Notebookcheck | 119 | 41 | 46% | 61% | 25% |
| TechGearLab | 117 | 34 | 56% | 39% | 37% |
| Popular Mechanics | 114 | 49 | 66% | 35% | 40% |
| Consumer Reports | 114 | 56 | 59% | 44% | 38% |
| Business Insider | 113 | 60 | 58% | 35% | 48% |
| What Hi-Fi? | 110 | 18 | 60% | 52% | 27% |
| IGN | 108 | 32 | 55% | 49% | 36% |
| TechPowerUp | 107 | 36 | 49% | 58% | 24% |
| PCWorld | 107 | 31 | 47% | 66% | 18% |
| PCMag Middle East | 106 | 53 | 56% | 51% | 31% |
| PCMag UK | 103 | 51 | 54% | 45% | 41% |
| Reviewed | 98 | 48 | 49% | 55% | 25% |
| 98 | 62 | 48% | 54% | 33% | |
| Wizard Review | 95 | 56 | 39% | 68% | 22% |
| Brutal Reviews | 95 | 46 | 53% | 26% | 59% |
| Digital Camera World | 94 | 23 | 46% | 56% | 31% |
| Critical Reviews | 92 | 35 | 55% | 23% | 55% |
| Engadget | 91 | 37 | 51% | 46% | 42% |
| SoundGuys | 88 | 10 | 57% | 46% | 31% |
| T3 | 87 | 49 | 55% | 65% | 16% |
| The Product Critiques | 87 | 61 | 45% | 61% | 22% |
| BGR | 84 | 47 | 51% | 41% | 42% |
| GearLab | 83 | 22 | 57% | 37% | 46% |
| GamesRadar+ | 83 | 23 | 47% | 59% | 28% |
| Gizmodo | 83 | 35 | 49% | 52% | 31% |
| Yahoo Tech | 80 | 39 | 57% | 45% | 36% |
| ZDNET | 77 | 32 | 55% | 51% | 26% |
How we count
A stance is where a source's view of a product sits on our own scale, from glowing to negative. Most stances are read from the source's own words by a model. A few are ratings the outlet published itself.
Agreement means the stance sits within one tier of the tier the product holds on its board. The tier is built from these same sources, so agreement shows how close one source sits to the consensus. It does not show whether that source is right.
We count a citation only on a published product, and only when it carries a stance. Names that differ only in case or punctuation count as one outlet. When a citation is stored under its platform name, such as Facebook or TikTok, we read the account from the link itself and count it under that account. A platform-named citation whose link names no account is left out, because it does not say who.
We do not list the sources that agree least. At this floor, the lowest agreement belongs to sources whose every stance we read as the same tier. That pattern comes from how we read the stance, so it would be wrong to print it under their name.
The figures come from the catalog this site was built from, and they change with every update. For how the tiers themselves are built, see how we rank.