I agree with other responders that records are really important. US Census records that list children with their ages can open doors to finding extended, non-direct family members. This helped me trace the descendants of my gg-grandfather's sister (who never left Germany) and I am now in contact with her living descendants. I knew the sister's name from another database I found online of the Jews from their town. When you start finding the names of siblings of your direct ancestors, you can go further in your research.
Google "Jews of ....." and look for hits for the towns your ancestors came from. You will be amazed at what you can find. Smaller towns with small Jewish populations can lead to major discoveries because family names are often mentioned. The other day, I found my gg-grandfather's German obituary from 1892 by Googling "Jews of Kulsheim."
I have also been successful on Ancestry with searching public member trees for my ancestral names. It's fascinating to find the names of ancestors that I knew of on other people's trees, and then to find their siblings. There could be a slew of name changes, but now you have a strong clue to follow. You have to be very careful of errors, however, because many trees are sloppy and not based on documentation. That said, I've been able to connect DNA matches that way and discovered siblings of great-grandparents and gg-grandparents and then traced their descendants to a DNA match who may not even have their own tree, but appeared on someone else's. Many years ago, I was told that genealogists should go from the known to the unknown, so the key is nailing down a fact, even if obscure or very distantly past, and using it to move forward.
Sandy Hahn Lanman