I'm sure the answer will depend on the specific species, but I'm looking for a ballpark figure for bacteria.
Yes indeed it will. Bacteria are a huge domain of life; you might as well ask for variability between all plants or between all animals. Bacteria can live in very different environments, and the whole things is made more complex by widespread horizontal transfer.
Still, for bacteria, you can't do much better than E. coli, and boy are they a mess. Here's a paper from 2010 that blew my mind. The authors sequenced 61 E. coli genomes and found:
...that any given E. coli genome sequenced will have only roughly 20% of its genes part of the E. coli core, and the remaining 80% are not found in all other E. coli genomes.
20% variability within a species. 20%, one species. E.coli is an extreme example, but still, species is a stupid concept. Here's a 2006 paper finding that plenty of unique Campylobacter proteins simply aren't in plenty of Campylobacter species.
On a more broad scale, here's an excellent (and free) paper you should look at. It examines islands of "microdiversity" among various species, finding that some species are easily 50%-75% different. A simple way to estimate this is too look at gene number for bacterial species, which should make it clear that some of these bugs vary by up to an order of magnitude.
In short: A lot.