Skip to content

guide

Why audiences leave a station or programme

Why listeners switch stations comes down to nine causes, and each one belongs to a different department.

Takwimu editorial team, Editorial team10 min read

Why listeners switch stations comes down to nine causes: the presenter, the music, the adverts, the signal, the language, content that is not for them, something specific on elsewhere, the cost of access, and a change in their own day. Takwimu Intelligence (TI) codes each one separately, because the department that fixes a transmitter is not the department that books the adverts.

Key points:

  • Every exit looks the same in a reach figure, and the causes need different owners.
  • Signal and language are often read as editorial problems when they are engineering and coverage decisions.
  • Moving for a match or a bulletin is a scheduling finding, and it should stay out of the fault count.
  • Reasons for staying are a separate question, and they do not mirror reasons for leaving.
  • An audience held by habit or by one presenter can look strong right up until it goes.

Why do listeners switch stations?

Stations often answer this by instinct. TI's coding frame, why an audience leaves, sorts every reason into one of nine codes.

Reason What it covers Who usually owns it
The presenter Manner, opinions, voice, or a presenter who has moved Programme controller
The music The songs, repeats, genre, or a change of sound Music or programme team
Advertising load Number of adverts, length of breaks, repeats, sponsored talk Sales with programming
Signal and reception Noise, fading, dropping out, no coverage, a stream that buffers Engineering
The language The language used, a change of language, code-switching, an accent Senior management
Content not for me Topics, news or guests that are not what this listener wanted Editorial
Moved for something specific A match, a bulletin or a show elsewhere Scheduling
Cost of access Mobile data, electricity, batteries, a subscription, a cover price Distribution
Their day changed A new job, a new commute, a move, a lost phone Nobody at the station

The first six are the ones a broadcaster can act on this week. Keep them separate even when a respondent gives two.

Why is the instinct usually wrong?

Stations tend to treat every drop as a programming failure. The schedule changes, a presenter moves, and the audience keeps falling because the real cause was somewhere else.

An engineering fault answered as an editorial one costs a year. So the first job is sorting, and the second is acting.

The regulator sees the wider pressure. The Communications Authority of Kenya's April 2026 report says audiences spreading across platforms make it difficult for broadcasters "to retain consistent audience share and loyalty". Its July 2025 report points to short-form content and social media scrolling as a reason broadcasters struggle to hold viewers through long-format programming. Those are market explanations. Your station's reasons have to be measured.

How do you measure switching away?

TI asks one direct question: in the last month, have you switched away from a station or channel part-way through a programme?

That gives switched away mid-programme. It is recalled over a month, so it shows how common the behaviour is. It says nothing about how many times it happened. The question covers any outlet, so a cross-tab by main station narrows it to your own listeners.

Where more than half say yes, the study raises a warning. Look at the reasons before assuming it is content, since adverts and signal both produce this pattern. Our page on audience loyalty and retention covers the wider picture.

How does TI capture the reason?

There are two routes. A closed question asks what made the person switch away the last time, with seven named reasons and "something else". An open question asks, in their own words, what makes them stay with a station and what makes them leave one.

The open answers are coded. This matters more in Kenya than a generic tool assumes, because the same complaint arrives in two languages. "Matangazo mengi" and "too many ads" are one finding. Without a lexicon that groups them, each looks half the size it is.

Cost of access and a changed routine come through the open question. So the question stays in, even on a tight budget.

Is the signal driving listeners away?

Check this first. It is the cheapest to check. Left because of the signal is the share of those who gave a reason who named reception or a failing stream. Above 20%, the study raises a high warning: one in five who left blamed the signal.

This is an engineering figure that arrives dressed as an audience one. Read it by county before reading it at all. A signal fault is almost always concentrated in one or two markets, while an editorial fault is not.

The pack sets a 48-hour clock where reach has fallen sharply in one market and not the others. The first question is whether it is transmission or schedule. Our guide to radio coverage and signal explains how to tell.

Are adverts driving listeners away?

Left because of advertising is the share of those who gave a reason who named the advertising load. Above 30%, the study raises a high warning.

Treat it as a trade-off. The inventory behind that number is earning revenue and costing reach, and usually only the revenue appears on a report. Somebody should make that decision on purpose.

One complaint about one advert's content is a different matter, and it is coded under something else. Our guide to advertising load sets out the options, from break length to repeats.

Is the presenter the reason?

Above 25%, the study raises a medium warning. It needs care, because a station figure hides individuals.

Read it beside held by the presenters. A station can be losing audience to one presenter and holding it with another, and a single figure averages the two into nothing. Cut it by programme where the base allows, and read the coded comments before anyone is called into a meeting.

Coding keeps this fair. A complaint about the content of an interview goes under content, and a complaint about the person goes under presenter.

What if the language is the reason?

Language is a coverage decision that somebody made. Left because of the language should be read beside the languages respondents say they follow.

The market context is clear. The CA's April 2026 report finds Swahili the primary language of radio consumption across all quarters, with vernacular stations second and English-language stations the smallest. Our page on vernacular radio stations sets out what that means for a format.

Some losses here are the cost of a sound choice. Lose them knowingly.

When is leaving no fault at all?

Two codes describe people who left with no complaint. Mixing them with the rest overstates dissatisfaction.

Moved for something specific means something they wanted was on elsewhere: a match, a bulletin, a particular show. That is a scheduling finding. If many people leave for the same bulletin, you have learned what your competitor owns at that hour.

Their day changed means nothing about the outlet changed. A new job, a new commute or a lost phone broke the habit. Code it here even if the person frames it as losing interest, provided the reason given is about their own circumstances.

Does cost stop people listening or watching?

Cost of access covers mobile data, electricity, batteries, a pay-television subscription and the price of a newspaper. The signal code excludes it on purpose, so a buffering stream and an empty data bundle stay apart.

The question bank can go further. One optional item asks whether the cost of mobile data has stopped the person listening or watching online in the last month. Add it if your streaming audience matters to your plan.

What holds an audience?

A second frame covers the positive half. Reasons for staying are no mirror image of reasons for leaving, so TI keeps them apart.

  • Attachment to a presenter: the listener names a person or a relationship.
  • Company and habit: the outlet is on in the background, or part of a routine.
  • Trust in the news: they believe what it reports, or it covers their area.
  • It speaks for people like me: its language, region, callers or humour.
  • Something they need: traffic, weather, market prices, obituaries, job or school notices.
  • It always comes through: the signal is dependable where they are.

Why can a strong audience be fragile?

An audience held by habit and an audience held by a presenter look identical in a reach figure. They behave completely differently when that presenter resigns.

Held by the presenters is real audience, and it is portable. It is the share of your figure that can leave with one resignation. Held by habit alone is more fragile still: nothing has to go wrong, and something else only has to become easier.

Neither figure is a fault. Both tell you where to spend. A station with a high presenter hold should work on contracts and on building the brand beyond one voice.

What does a worked reading look like?

Here is an invented wave. Of 1,000 people interviewed, 400 switched away in the last month and gave a reason.

Reason People Share
Poor signal or reception 88 22%
Too many adverts 80 20%
The presenter 60 15%
The music 52 13%
The language 40 10%
The news or talk was not for me 40 10%
Moved for something specific 40 10%

The signal warning fires at 22%. Adverts and the presenter stay under their lines. Cut by county, suppose the signal complaints sit almost entirely in two markets. The station now has one call to make, to engineering, before any talk of changing the schedule.

What should a station do in the first week?

Work through the causes in cost order.

  1. Read left because of the signal by county, and check the transmitter and stream.
  2. Put left because of advertising in front of sales and programming together.
  3. Read presenter figures beside held by the presenters.
  4. Separate moved for something specific from the faults.
  5. Check who can place the programme, then promote before changing it.

The programme performance report carries the same figures show by show, and the radio listener survey questions page shows how to ask them.

Where TI fits

Find the reason before changing the schedule. TI codes each reason, raises a warning when one crosses its declared rule, and reads signal complaints by county so that a transmitter fault reaches engineering.

To see it on your own station or title, book a walk-through.

Takwimu Intelligence (TI) is audience measurement software for Kenyan media.

Sources

  1. 1

    Audience Measurement and Industry Trends Report, abridged version, April 2026 · Communications Authority of Kenya, April 2026Seven-day and daily media use to Q3 2025/26, radio genre and language, advertising window, industry challenges. Opened 5 November 2026.

  2. 2

    Audience Measurement and Industry Trends Report, July 2025 · Communications Authority of Kenya, July 2025Challenges for the broadcasting industry, including attention spans and long-format programming. Opened 5 November 2026.

More from the library