One of the big concerns for people using VoIP-based interviews in their podcasts using Skype is the variance in quality between participants. It's my experience in having conducted doing several interviews based on this setup that typically the person doing the recording will be a lot louder than the person being interviewed.
For most people that don't have sophisticated home setups with condenser mics, complex mixers and avanced tools, we normally get one shot at gettting this right, lest we have to embarrasingly call the person again and re-record. With podcasts being not-live time-shifted digital audio, we're of course able to go in and make post-production fixes and tweaks to improve the sound of our stuff. The main goal should be to balance out the volume between you and the other person(s).
In the freeware app Audacity, this is best done by leveraging the Amplify and Normalize tools, both of which are found under the Effects menu. The former alters the volume setting (either positively or negatively) of a clip, and the latter lets you make spot corrections to an DC offset displacement. What this means to you in non-technical terms s that you can level out the audio. By highlighting a clip within an audio track and then using Effects ==> Amplify and Effects ==> Normalize (in that order and in reverse), you can find a good balance so they're isn't a huge disparity between your voice and someone else's. (Tip: if the person on the other end is coming in a lot lower than I am, I typically only normalize my audio, to bring it down, and then play with both tools on the guest's levels).
The downside is that this is a lot of post-production effort. I typically spend 90 minutes editing each 30-minute blocl of a Skype converation. So, I get around this by having managing my Skype interviews so that they're focused, quick, and to the point. But more importantly, I make sure that when I ask a question, I let the paticipant answer and I try and stay silent. This makes the editing mentioned above easier, without two voices blending in together.
This results in a better-sounding podcast that people will want to listen to.
Here are some good related tips: