The podcast audience in Kenya is small and hard to size, with published figures running from 1.4% of platform mentions in the Media Council's April 2026 survey to one in four respondents in a 2021 GeoPoll SMS survey. Those two numbers do not conflict, because they measure different things.
Key points:
- Podcasts rose from 1.0% to 1.4% of platforms mentioned in the Media Council's survey between 2024 and 2025.
- GeoPoll's 2021 survey found a quarter had listened in 30 days, among 1,905 mobile respondents.
- A download means a file was delivered. It does not mean anyone listened.
- The IAB Tech Lab rule filters downloads by one minute of content and a 24-hour window.
- Each platform's dashboard counts only its own listeners.
How many people listen to podcasts in Kenya?
Published figures are few, and they are old or partial.
| Source | Date | Method | Finding |
|---|---|---|---|
| Media Council of Kenya | April 2026 | 3,774 face-to-face interviews | Podcasts 1.4% of platforms mentioned, up from 1.0% |
| GeoPoll | Q1 2021 | SMS survey, 1,905 respondents | About 1 in 4 listened in the last 30 days |
The first is a share of all platform mentions, where each person can name several. It is not a percentage of the population. The second covers people reachable on a mobile panel, in a survey by text.
Neither is a reach figure for Kenya today. Treat both as signals of scale.
Do Kenyans know what a podcast is?
Awareness was a real barrier in 2021. GeoPoll found that among the three in four who had not listened, close to 60% did not recognise what a podcast is.
That finding matters for survey design as well. Asking "do you listen to podcasts?" assumes people share your definition. Many people hear the same show on YouTube or a radio catch-up, and may never use the word.
How are podcast audiences measured?
Four methods exist, and each sees part of the audience.
| Method | What it counts | Blind spot |
|---|---|---|
| Hosting logs | Downloads delivered | Whether anyone played them |
| Platform analytics | Plays inside one app | All other apps |
| Survey of people | Who says they listened | Memory, and small samples |
| On-air or social response | People who reacted | Everyone who stayed silent |
These cannot be added together. The same rule applies here as for radio streaming logs.
What counts as a podcast download?
The global reference is the IAB Tech Lab Podcast Measurement Technical Guidelines. Version 2.2 sets rules that a compliant host filters by.
- Use a download threshold based on one minute of content.
- Count within a 24-hour window, fixed or rolling, so repeats are removed.
- Use IP address and user agent to judge whether a request is unique.
- Exclude pre-downloads, known servers and implausible volumes.
A download under those rules is a cleaner number. It remains a count of delivery. Auto-downloads to a phone can pass the filter and sit unplayed.
Is a download the same as a listener?
A download is an event. A listener is a person.
One person can download on a phone and a laptop. One phone can serve a whole family. The guidelines also note that many people can share one IP address in offices, dorms and similar setups, so IP is only one factor.
So never present downloads as listeners in a sales pack. State what was counted, and over how many days.
What do Spotify and other platforms count?
Each platform sets its own terms. Spotify describes plays as the number of times people have actively watched or listened to any episode in a catalogue.
A Spotify play and a host download are different events, so a show cannot sum them. YouTube sits apart again. Reuters Institute finds YouTube widely used in Kenya, at 84% for any purpose in its online sample, and much podcast listening there would register as video views.
Who listens to podcasts in Kenya?
Reliable national profiles are scarce. The claims in circulation come from small samples or single platforms.
The nearest guide is the shape of internet use. The CA shows 74% of people aged 18 to 24 online in April to June 2026, against 33% of those 45 and over, with 98.9% of internet users on smartphones. Any podcast audience sits inside that group, so it is likely to skew young and connected, as youth media consumption in Kenya would suggest.
A station with a real interest should measure this directly. Rumours about young urban listeners are a poor basis for a sponsorship deal.
How do you compare a podcast with a radio show?
Compare them with care, and only on matching measures. A radio programme's reach is an estimate of people from a survey. A podcast's download count is a log of files sent.
The periods differ as well. Radio reach is usually a day or a week. A podcast episode gathers downloads over months, so a large total can hide a small weekly audience. State the window, such as downloads in the first 30 days.
Are news podcasts big?
They remain a minority habit worldwide. Reuters Institute reports news podcast use unchanged at 11% globally in 2026.
The CA treats podcasts as an opportunity for broadcasters. Its August 2026 report suggests repurposing radio and TV programmes as podcasts, and lists podcast sponsorships among new revenue streams. So the commercial interest is real, even with a small audience.
Can a radio show count its podcast as extra audience?
It should report it separately. A listener to the show on air and a listener to the catch-up may be the same person.
Adding the two inflates your reach. Report broadcast reach from a survey, catch-up downloads from your host, and say that overlap is unknown. Our page on actual radio listenership sets out the same discipline.
How should you ask about podcasts in a survey?
Describe the behaviour, and avoid leaning on the word alone. Try this sequence:
- In the last seven days, did you listen to a recorded talk show or programme on your phone?
- Did you listen on YouTube, Spotify, a station's site or another app?
- Which show?
- Did you hear the same programme on the radio that week?
The last question measures overlap, which no log can do. Read audience survey questions before drafting yours.
What do people get wrong about podcast audiences?
They quote a single impressive figure with no base. They also take global download benchmarks and apply them to a Kenyan show.
The third error is treating a podcast as a replacement for reach. A small and loyal audience can be valuable, and it should be sold as exactly that.
What should a producer or station do?
Keep your host's filtered downloads as your internal trend. Add platform plays in a separate line.
Then run a short audience survey on how people found the show and what they would change. Sponsors usually want to know who listens and whether they acted, and logs cannot answer that. Takwimu Intelligence (TI) asks people directly, and a programme panel can collect appreciation and suggested changes.
Read next
- Audience measurement software for Kenyan media: the TI pack page for this work
- Print and digital audiences in Kenya: all the pages in this section
- Streaming analytics
- Can a radio station know its actual listenership?
- Audience survey questions
- Takwimu Intelligence (TI) home page
- Audience measurement in Kenya: the guide to this library
- The TI platform and how we check our numbers
Where TI fits
A download is a delivery and a listener is a person. TI measures people by asking them, and its programme panel gives appreciation and what they would change, which a hosting log cannot show.
To see it on your own station or title, book a walk-through.