I don't think it's normal that this would be due to differences in the preps. More typically, it would be due to differences in the mammalian cells' growth stage, low-level contamination (mycoplasma), general cell health, and also a large degree of irreducible variance due to chaotic behavior of many transfection methods. So, my first question would be, what evidence do you have that batch differences are actually to blame? Are those specifically designed, statistically meaningful experiments? Or more like a gut feeling? Because if you don't have the evidence, then the problem is likely elsewhere...

I agree. It's far more likely that the variability in transfection is due to variability in your starting cell population, not variability in your DNA prep.

However, just for the sake of discussion I'd also point out that the DNA concentration and purity aren't the only thing you'd want to look at to assess your prep. For example, how sure are you really that you have the same population of plasmid molecules from prep to prep? Plasmids can sometimes undergo recombination (especially if they're kept in an unsuitable bacterial strain or kept in a glycerol stock that's been freeze thawed a bunch). You can also get contamination from bacterial genomic DNA into your preps. Also, preps might have different proportions of supercoiled versus nicked DNA (supercoiled will do better in transfection).

Best practices (especially if you're already suspicious of your plasmid preps) would be to re-transform and colony streak new bacteria for liquid culture each time (instead of using an old glycerol stock or something) and then run your DNA out on a gel to make sure it's high quality supercoiled plasmid. I also like to run a panel of restriction enzymes on the prep that will give you a known banding pattern to fingerprint it. This will let you know if you're getting any major rearrangements in your plasmid or if any irrelevant DNA has hitchhiked into your prep.

Nice answer Govinda

I agree

What is key is growth in medium with your transformants from one prep to another

To keep your cells in log phase for maximum time which will maximise yield subtle ptactices like

Picking a colony and growing during day in 5mls of medium in the absence of selecting antibiotic

Then inoculating an over night culture with say a 1:1000 of that starter but only adding the selectable antibiotic after 1 hour in the orbital shaker and then browing until morning: this encourages your cells to come out of lag phase and into exponential or log phase growth

Remember, your plasmid is a foreign entity in your transformant and so the cells are busy trying to deal with it rather than devote ambient time to mitosis; including recpombination or in some instances spitting your clone out of the cell altogether: this obviously will lead to reduced yield

Try also growing at 33C instead of 37C if more than 12-14 hours of growth once again to ensure that by early morning the culture has not "peaked too soon" and spends too much time in stationary phase

Consider also aeration: make sure that your liquid culture is only 1/3 of your total volume so that eith shaking at 250rpm your culture is kept fully aerobic the whole night