There are several possibilities, we're currently testing some.
As this is an open topic, let me write down the tests we made and some thoughts on the matter. Asusual, I think discussion will lead to the better solution.
No doubt a perfect match will be impossible, unless we start asking billion new data while configuring stuffs.
There are users already going nuts for the "very few" options we have already... can't imagin what would happen if I'd add even more!
The latest test we performed was about trying to reverse the concept.
The main facts are essentially these:
1) carriers are calculating no more on the real weight but on the box size
2) carriers still uses real weight if it is bigger than volumetric size
So, our test was about ignoring the max carton size you currently set in the RTD, and try to create a fake box size that would actually match the real weight plus a littme margin.
This worked essentually well, but the drawback as clear and immediate: light items in big boxes would return very wrong results (you would calculate against weight, so showing users a probablu low price: but whey you'll ship, you will use your big box and the extra expense for the shipment will be all on you)
Other idea was similar to what you proposed: having size for all products. But again, there are a number of BUTs:
Essentially, we would need to add 3 new fields for the max box size everywhere we're currently allowing to enter a weight value.
This because, in a fixed size product, it is not uncommon that different sizes are REALLY different (sometimes even twice as much), even if they're the same product.
Also htere are accessories from the configurator, in case you're using them: a banner stand can be part of the order, but it's normally bigger than the banner itself.
Finally, there are ranged products: a whole different story, especially if they can even be rolled.
So I am looking with exreme caution at ths solution, because adding 3 new fields in so many places could end up in being just too complex for normal users. In simple terms, if I have to find a solution, it must be something that will actually be used. Would be a nonsense to add so many new fields to the system if I'm expecting just couple users using them, and still with no optimal results
So, after all of this, the brainstorming is moving into yet another possible direction.
You know we currently suggested to place, as the carton sizes, the biggest box you actually have and the max weight per carton.
Now what if:
1) instead of the max, you enter the SMALLEST box size you have
2) you keep the max weight per carton as is
What would happen here is that, if wieght is very low, the smallest box size would be used. Otherwise, real weight would.
The idea could be to, instead of adding the box sizes in several new places, just enter the weight itself (which is already in there) but thinking dimensional instead of phisical.
To say it simple, let's say you have a product you know would go in a box size returning a dimensional weight of 20 Lbs: what you'd do is so to simply assign 20 as the weight for that product, instead of it's real one.
I know this might sound silly, but at the end UPS and Fedex clearly stated this: the billable weight will not be the real weight anymore, in vast majority of cases. So, what if we simply discard thinking in physical pounds but we start thinking everything in billable pounds?
