Share of audience is the percentage of all people using a medium at a given time who are tuned to one station or programme, calculated as the station's audience divided by the total audience for that medium, multiplied by 100. It measures competition for the audience that exists, and says nothing about how big that audience is.
Key points:
- The formula is your audience ÷ total audience for the medium × 100.
- Share is based on people using the medium, while a rating is based on the whole population.
- Three common bases give three different shares: people at a moment, hours over a week, and the Top 10 only.
- Share can rise while the audience falls.
- A share's margin of error is wider than a reach figure's, because its base is smaller.
What does share of audience mean?
Think of a county at 7am. Some adults are listening to radio, and the rest are not. Share looks only at those listening and asks how they divide between stations.
Nielsen's radio terms define AQH share as the average quarter-hour audience of a station as a percentage of the total radio listening audience. They call it the primary metric programme directors use to judge a station's effectiveness.
For television, the definition carried by the Media Literacy Clearinghouse from Nielsen is the percentage of households or persons using television that are tuned to a specific programme at a specified time.
The share of audience formula
Share (%) = your audience ÷ total audience for the medium × 100
Both audiences must come from the same people, place and time. That rule sounds obvious. It gets broken constantly.
A worked example
This example is invented, with round numbers. A county has 1,000,000 adults. Between 7am and 8am, 300,000 are listening to radio, and 75,000 of them are on Station A.
Station A's share is 75,000 ÷ 300,000 × 100 = 25%. Its rating is 75,000 ÷ 1,000,000 × 100 = 7.5. Same audience, different bottom number.
What is share of audience based on?
The base is the whole story. Three are in common use.
| Base | Example | What it tells you |
|---|---|---|
| People listening at a moment | Nielsen's AQH share | Who is winning this time slot |
| Listening hours over a week | RAJAR's share of listening | Who holds the most time |
| Listening to the Top 10 only | GeoPoll's time-block share | How the leading stations divide their block |
RAJAR defines share as the percentage of total listening time accounted for by a station in an average week. GeoPoll describes its Q1 2026 figure as each station's share of total Top 10 viewing within the time block. Both are clear about their base, which is what makes them usable.
Share of people vs share of time
A station with few loyal listeners can beat a station with many occasional ones. Here is an invented week in a market with 5,000,000 total listening hours.
| Station | Weekly listeners | Hours each | Total hours | Share of hours |
|---|---|---|---|---|
| A | 100,000 | 10 | 1,000,000 | 20% |
| B | 300,000 | 3 | 900,000 | 18% |
Station B has three times the reach. Station A has the larger share. Neither number is wrong, and a station should know which one it is quoting.
Why two Kenyan shares for one station can both be right
GeoPoll, for January to March 2026, gave weekday breakfast (6am to 9am) shares of Top 10 listening as Classic 105 18.8% and Citizen 17.4%. In drive time (4pm to 7pm) it gave Citizen 20.4% and Classic 105 14.5%. Its method is daily mobile panel surveys.
RadioInfo Africa, reporting the Ipsos survey for October to December 2025, gave Citizen a 17% market share. The report does not say how that share was calculated.
So one station has several shares. They cover different quarters, firms, times of day and bases. Read each as that source's finding and compare it only with itself over time.
Why a Top 10 share runs higher
A share of the Top 10 is larger than a share of all radio, by construction. Suppose, in an invented case, the Top 10 stations carry 80% of all listening. A station with 20.4% of Top 10 listening then has 20.4 × 0.8 = 16.3% of all radio.
Neither is better. They answer different questions, and a sales sheet should say which it uses.
Share can rise while the audience falls
Share is relative, so it can flatter a station. Take two invented months at 7am.
| Month | All radio | Station A | Share |
|---|---|---|---|
| March | 300,000 | 75,000 | 25% |
| April | 200,000 | 60,000 | 30% |
Station A lost 20% of its listeners and gained five share points. Its programme team may celebrate. Its advertisers bought 15,000 fewer listeners. Share should always travel with the audience number.
Why share has a wider margin of error
Share is calculated only among people using the medium, so its base is smaller than the sample. Take an invented survey of 1,200 adults in which 360 listened in the block. A 25% share on that base carries a margin of about 4.5 points either way at 95% confidence, on a simple random sample.
The base here is only 360. So 25% means somewhere between about 20.5% and 29.5%. A rival on 23% is not clearly behind you. Report a lead only when the gap is larger than the noise, and read sample size for an audience survey before promising a share by daypart.
How do you calculate share from a recall survey?
Day-after recall gives you the raw material directly.
- Ask who listened to radio yesterday, and in which time blocks.
- For each block, count people who named each station, using weighted counts.
- Divide each station's count by the count of everyone who listened in that block.
- State that this is a share of people in the block. A person who listened to two stations counts for both.
Step four matters. If many people name two stations, shares calculated on people sum to more than 100%. To force them to 100%, divide by total station mentions, and say so. The method is set out in day-after recall.
Never average shares across dayparts
Stations often want one number for the day. A simple average of daypart shares gets it wrong. Here is an invented day.
| Daypart | All radio | Station A | Share |
|---|---|---|---|
| Breakfast | 300,000 | 75,000 | 25% |
| Midday | 100,000 | 15,000 | 15% |
| Drive | 200,000 | 36,000 | 18% |
| Evening | 120,000 | 12,000 | 10% |
| Sum of dayparts | 720,000 | 138,000 | 19.2% |
The average of the four shares is 17%. The true figure is 138,000 ÷ 720,000 × 100 = 19.2%, because breakfast carries more listeners. Sum the audiences first, then divide.
What is a strong share of audience?
It depends on how many stations compete. The CA report of August 2026 counts more than 200 radio stations in Kenya, so a national share will look small beside a county one.
A quick yardstick is fair share: 100 ÷ the number of stations in the market. With eight stations in an invented county, fair share is 12.5%. A station on 25% holds twice its fair share, and a station on 6% holds about half.
Fair share is only a yardstick. Stations differ in signal, language and format, so the kinds of station in Kenya matter when you read it.
Share of audience, market share and share of voice
Three similar phrases mean three different things. Share of audience measures listeners or viewers. Market share usually means a brand's sales. Share of voice measures a brand's advertising against rivals.
RadioInfo's "market share" above means a share of the national audience, so read the words with care. Check before you compare.
What do stations get wrong about share?
They quote share to advertisers. A programme team needs share, but a buyer pays for people, and a 30% share at 2am may be a few hundred listeners.
They also rank themselves on differences smaller than the margin of error. Takwimu Intelligence (TI) would have them do the reverse: quote share by daypart with its base, beside the audience behind it, and call a tie a tie. See breakfast radio for where this matters most.
Read next
- Audience measurement software for Kenyan media: the TI pack page for this work
- Audience metrics explained: reach, share, ratings and cost: all the pages in this section
- Rating vs share
- Dayparts in radio and TV
- Top radio stations in Kenya
- 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
Share belongs with its base and its audience. TI reports station share and reach by daypart from day-after recall, weighted to KNBS census totals at county level, with the margin of error shown on each figure. Its warnings include a breakfast show smaller than the afternoon, so a share shift reaches the programme team while it can still be acted on.
To see it on your own station or title, book a walk-through.