Comparing data sets with two very different samples sizes - stats newbie

#1
Hi, I realize this question might have been asked before, or is very rudimentary, but I'm having trouble figuring it out. I have several data sets (~50) comparing outputs from two different brands of product from different areas of the country. The issue I'm coming across is that for Brand A I have about 100 data points for each area of the country, while for Brand B I only have 10-20. I'm trying to assess the viability of comparing the mean output across these two brands, unique to each specific region. Searching online, some suggestions I came across are a Welsch test, and also a Chi-square. I'm not sure which, if either, are right.

Any and all advice will be most appreciated. Thank you, and please also ask for clarification if anything I've stated is unclear!