How to Start a Podcast in Kenya: Format, Kit and What It Really Takes
Joel Kimani, Content & Strategy·
Most podcasts do not fail because of equipment. They fail at episode six, when the person doing everything runs out of time.
So this guide starts with the decisions that determine whether a show survives, and gets to microphones later.
Decide what the show is for
There are broadly three reasons an organisation starts a podcast, and they produce different shows:
- Authority — you want to be the voice people in your sector listen to. Long-form interviews, guests who bring their own audience.
- Relationships — the real value is the conversation with each guest, who is often a client or partner. Audience size matters less than who is on.
- Reach — you want a lot of listeners. This is the hardest and slowest, and it needs a genuine content angle, not just a brand.
Being honest about which one you are doing changes the format, the length and how you measure it. A show built for relationships that gets judged on download numbers will be declared a failure while it is quietly working.
Audio or video?
Video podcasts take more to produce and reach further. The clips are what actually travel — most people discover a show through a 45-second vertical clip on Instagram or TikTok, not through a podcast app.
If you can only do one thing well, do audio well. If you want the show to grow, record video and cut clips from it.
Consistency beats production quality
An adequate show published every Tuesday will beat a beautiful show published whenever there is time. Listeners build a habit around reliability, and platforms favour it.
The practical implication: batch record. Do not record one episode a week. Record four in a day, then publish weekly. It is the single most useful piece of advice for a new show, and the one most often ignored.
What equipment actually matters
In order of impact:
- The room. Sound bounces off hard walls and glass. An untreated boardroom makes an expensive microphone sound cheap. This is why studios exist.
- Microphone technique. Close to the mic, consistent distance. This matters more than which microphone.
- A dedicated microphone per person. Not one in the middle of the table.
- Recording each person to a separate track. If one guest is quiet or coughs, you can fix one track without touching the others.
- The camera, if you are shooting video. Last, genuinely.
A common and expensive mistake is buying good microphones and recording in a room with a glass wall and a hard ceiling. The room wins.
Studio or record it yourself?
Recording yourself makes sense if you have a suitable room, someone who enjoys the technical side, and a format that does not need multi-camera.
A studio makes sense when you want to turn up and talk. The value is not the equipment — it is that the acoustics are already solved, the cameras are already positioned, and someone else is watching the levels so you can concentrate on the conversation.
Our Kilimani studio is set up for exactly that: multi-cam video, treated acoustics, and post-production, so a session produces a finished episode rather than a folder of files.
Publishing and distribution
The mechanics are straightforward:
- Host the audio somewhere that generates an RSS feed
- Submit that feed once to Apple Podcasts, Spotify and YouTube
- After that, publishing to your host distributes everywhere automatically
Cover art and episode titles do more work than most people expect — they are the entire browsing experience in a podcast app.
A realistic first three months
- Month 1 — decide the format and length, record a pilot you never publish, batch record four real episodes
- Month 2 — launch with three episodes available at once so a new listener can binge, then publish weekly
- Month 3 — review honestly: which episodes travelled, which guests brought their audience, is the schedule sustainable
Most shows that make it past month three keep going. Getting there is the whole challenge.
---
If you would rather press record and let someone else handle the technical side, talk to us about studio time.