Safety-wise you'll probably find relatively little - unless you're doing some more unusual things with vegetable juices, you'll be working with products whose nature, particularly with respect to pH, generally makes it far easier to kill pathogens than it is to kill spoilage organisms. Typical testing for finished fruit juices may often just be TVC/Yeasts/Moulds.
Limits will depend a bit on your pack format, but if these are ambient-stable then you'll effectively need to find zero in the product, as yeasts/moulds/various bacteria will happily grow in a nice warm sugary environment if they do survive your process.
Nonetheless for specification purposes it's not uncommon to see limits set as e.g. <10cfu/ml for some product formats, partly to reflect that most processors are using a pasteurisation rather than a sterilisation process, so they're not actually theoretically expecting to kill everything. There certainly are some bacteria that can survive pasteurisation and not pose any spoilage or safety problem in the end product, but it's not something you generally want to gamble on.
One group of bacteria to which you may want to give some thought is Alicyclobacillus - if present, this can survive pasteurisation in a spore state, and is very capable of generating some unappealing off-notes in finished product. There isn't a great deal you can do about it in your own process unless you are doing something fairly unusual in terms of high temperatures - it's one that really needs to be controlled at the raw material processing stage, either via fruit washing and/or filtration (latter only viable for clear concentrates). Testing for this isn't offered by all that many labs, and whilst there are a few rapid methods around, I'd recommend IFU12 if you can find someone who offers it / are able to set up your internal lab to do it. Talk to your raw material supplier(s) about this too, as you may want to get some sort of parameter added to specs to confirm a limit of <1cfu/10g. (Some big processors may not want to agree to this though...)