PoE, PoE+ or PoE++?

How this verdict was reached

1. What the published standard settles
2. What it costs to install
3. What it costs to specify wrongly
4. Whether both sides can be supplied from this catalogue

How every verdict on these six pages was reached, in the order the criteria were weighed. Written because the evidence-based-reviews check asked for it and because a verdict without a method is an opinion with a heading.

1. Does the published standard settle it? Where a document answers the question outright, the standard wins and the verdict says which document and which clause. Cat 6A carries 10 Gigabit to 100 metres because ANSI/TIA-568.2-D says so; nothing here is entitled to a second opinion about that.

2. What does it cost to install? Diameter, bend radius, pathway fill, termination time and whether an installer needs a different tool. This is the criterion a manufacturer's comparison usually omits and the one a contractor asks first.

3. What does it cost to get wrong? Asymmetric almost every time. Specifying Cat 6A where Cat 6 would have done wastes money once. Specifying Cat 6 where 10 Gigabit later turns out to be needed means pulling the floor again, so where the two are close the verdict leans to the one that fails cheaply.

4. Can both sides actually be supplied? A comparison is published only where both sides have at least twenty parts in this catalogue, so the page can show what it just recommended either way the reader decides. This is why there is no OM3 against OM4 page.

Not used, deliberately: margin. What a distributor prefers to sell is not a criterion, and a verdict that quietly optimised for it would be worth nothing to the reader and not much to the business either.