I was asked this question at a recent seminar I gave.
The answer I gave was "none."
It was an honest answer. I don't like telling attendees I use something that I don't. I really don't use a genealogy database of my own all that much. There are times when I think I should and I have used one in the past. I know there are several very capable programs out there. Because of how I research and the fact that I am constantly researching something for the newsletter
Casefile Clues, I haven't taken the time to perform data entry into an actual genealogy database in several years. However, that doesn't mean that I don't extract data and use a computer to sort and analyze it. I just don't use one of the traditional genealogy databases to do the work.
Why?
I tend to find genealogy databases limiting--at least for how I work. Most of the families I'm working on have numerous conflicting records, missing pieces, inconsistencies, etc. I like to perform analysis and write up as much as possible my thought process as I transcribe. For some, this would seem the ideal reason to use a database. For census records, I have charts I use to extract the data to assist with analysis. I have a variety of charts and tables I use.
Does this occasionally get repetitive? Yes sometimes it does. Could I better manage footnotes and sources? Yes.
But I like the freedom that a word processing document gives me. Would it be helpful to have a software program "generate" the text of a genealogy report from the information already entered in my database? Perhaps. But I'm not writing compiled genealogies of several generations. I don't want to write compiled genealogies. That simply is not my interest. My preference is to write about the analysis of records and what they say and where the research should continue. For me that helps me research and discover more information.
I realize this may sound like heresy, but I don't think it is the end of the world if you don't put all your information in a genealogy database. I've made significant discoveries in the last several months on my great-grandmother's parents who married in Iowa in 1870. They both were part of a migration path that began for both their families in New England in the early 1800s. In the process of that research, I've written up most of what I have found. I've incorporated tables and charts where necessary. I've used maps (both contemporary and modern). I've used record images obtained from a variety of sources. Putting this together in chronologies and writing up my process and reasoning so that it made sense to me (and to others) has been invaluable.
The process of that writing has given me a long list of things to do. That list is in a chart, with goals and priorities attached. I love to use charts and tables for those things (attendees at the recent Tennessee Genealogical Society seminar KNOW that I like tables and charts). I just don't have information put in a "genealogy database" per se. My "to do" list on the family is fairly long as it is and as I discover more and incorporate that into the material I already have, the list gets longer. Are there things I might notice if I used one of the main databases for the information? Perhaps. There might be a time when I start entering the information into a database.
But for now, I'm making enough discoveries and have enough analysis to keep me busy that I don't find the need to stop for data entry. Transcribing, analyzing, and explaining my rationale in my word processing documents on my family keeps me occupied enough.
And that's ok.
. Emma was born in Canada (probably Ontario) about 1840, the daughter of Clark and Mary (Dingman) Sargent. You'd think with that many, there'd be something. 
