WAV to MP3 is the standard "studio to distribution" step
WAV is what your DAW writes when you bounce a finished mix. The session was at 24-bit 48 kHz, the master export is the same, and the file is large because it preserves every sample of the recording without compression. That is the right format for archive and re-editing.
It is the wrong format for distribution. Uploading a 350 MB WAV to a podcast host, attaching it to an email, putting it on a USB stick for a car stereo: all hit the same wall. People stop downloading at 50 MB. Email caps at 25 MB. Car stereos run out of memory. MP3 at 128 kbps fits the distribution side of that workflow cleanly.
The studio bounce → publish workflow
The standard pattern: bounce to WAV (the master, archive copy), then encode to MP3 (the distribution copy). Many podcasters do exactly this every episode: Reaper or Logic mixes the episode, exports the master as 24-bit WAV for the archive, then encodes a 128 kbps MP3 for the host upload. The WAV stays on your project drive; the MP3 ships.
Why we default to 128 kbps
For voice content, 128 kbps mono or stereo is the sweet spot between file size and perceived quality. Going higher (192, 256, 320 kbps) makes the file 50-150% bigger with no audible improvement on consumer playback. Going lower (64-96 kbps) starts to introduce artifacts even non-trained listeners notice on music, though voice tolerates it well.
If your source WAV is huge, downsample before upload
A 60-minute stereo 24-bit 48 kHz WAV is around 650 MB. Past our upload cap. The conversion still works if you first shrink the source: ffmpeg -i input.wav -ac 1 -ar 16000 -sample_fmt s16 small.wav produces a roughly 110 MB version with identical voice quality. Or skip the conversion entirely and re-encode the source WAV directly to MP3 with ffmpeg -i input.wav -ab 128k output.mp3.