In the 1980s, television listings and newspaper reports regularly announced that millions of people had watched a particular programme. A Friday episode of EastEnders, for example, might be credited with an audience of several million. But how could anyone possibly know what was happening inside millions of living rooms? Was somebody monitoring every television set in the country?
The answer is both simpler and more ingenious: they didn’t monitor every home. They monitored a carefully selected sample of homes and used statistics to estimate the viewing habits of the country as a whole. Unless your household had been recruited into that sample, nobody at the ratings office knew whether your television was switched on, let alone what you were watching.
The beginnings of BARB
In Britain, the organisation responsible for producing the industry’s standard television audience figures is the Broadcasters’ Audience Research Board, better known as BARB. It began reporting audiences in August 1981, replacing separate estimates used by the BBC and ITV with a common system.
BARB commissioned research from specialist contractors and built a panel of households intended to represent television-owning homes across the UK. The BBC’s 1985 Year Book described a panel of about 3,000 homes. These were not simply the first 3,000 people who volunteered: the sample was designed to reflect the wider population and its different types of household.

How the television meter worked
Before the major upgrade in 1984, a small electronic meter was attached to each television set in a participating home. It recorded when the set was switched on and which channel it was tuned to. The equipment stored this information, rather than sending somebody to watch the household.
But there was an obvious problem. A television being switched on does not prove that anybody is watching it. It might be playing to an empty room while the family is in the garden, cooking dinner or out of the house.
To find out who was actually viewing, households also kept paper viewing diaries. Members aged four and over recorded when they watched television, generally in quarter-hour intervals. The electronic meter established what was on the screen; the diaries supplied information about the people watching.
The 1984 revolution: the peoplemeter
In December 1984, BARB’s enhanced measurement system introduced the peoplemeter, replacing the combination of set meters and paper diaries. This was a significant step because it made it possible to link the television’s channel information with the viewing activity of individual household members.
Each participating person was assigned a button on a small handset. When they started watching, they pressed their personal button; when they stopped, they indicated that they had left. The meter could therefore record not just that a television was tuned to BBC1 or ITV, but which registered members of the household said they were watching.

The information from the television meter and the household handset was collected and linked. The BBC’s 1986 Year Book explains that the data were sent overnight by telephone line to a central computer. The computer combined the records to establish the viewing of each panel member.
It was a remarkably advanced arrangement for the time. A household could be watching television in the evening, and the detailed viewing record could be transmitted automatically while the family slept.
How a few thousand homes represented millions
The crucial step was statistical weighting. Each participating household represented more than itself. The sample was designed so that different kinds of households and viewers stood in for their equivalents in the national population. Researchers applied weights to the viewing records to produce estimates for the country.
Imagine, in simplified terms, that a particular type of household was under-represented in the panel. The viewing records from the households in that group could be given greater statistical weight. A group that was over-represented could receive less. The real calculations were much more detailed than a simple multiplication, taking account of the composition of the sample and the population it was intended to represent.
That is how a programme could be reported as having an audience of 3.2 million even though only a small fraction of that number had been directly measured. The figure was an estimate derived from the panel, not a headcount of every viewer in Britain.
Could the system tell if somebody was really watching?
Not perfectly. The peoplemeter depended on people remembering to press their buttons when they began or stopped watching. Someone might leave the room and forget to register it, or sit down and forget to identify themselves. A television meter could identify the channel, but it could not read a viewer’s mind or guarantee that their attention was on the screen.
That is an important distinction. Audience figures were never a literal count of every pair of eyes fixed on a programme. They were carefully constructed estimates based on measured television use and information supplied by a representative panel.
What about today?
BARB still provides the UK’s standard television audience measurement service, but the media landscape and the technology have changed enormously. Its current measurement covers viewing across television platforms and includes more than traditional live, linear broadcasts. The panel remains central, supported by additional data and measurement techniques that reflect modern viewing habits.
BARB also distinguishes between overnight figures and later consolidated figures. The overnight numbers are not necessarily the final audience: exact programme timings and viewing after the original broadcast can affect the eventual total.
So, was your television being watched?
Unless you lived in one of the selected panel households, almost certainly not. Nobody at BARB had a secret switch that could reveal whether your set was on, which channel you had chosen or whether you were enjoying Friday’s EastEnders.
Your viewing habits were represented by the behaviour of a relatively small group of carefully selected households. Their meters, diaries and button presses became the evidence from which national audiences were estimated.
So the next time an old television listing announces that 15 million people watched a programme, remember the unlikely machinery behind that number: a few thousand homes, little electronic boxes, personal buttons, telephone lines and a great deal of statistics.
Sources and further reading
- BARB – About us and the history of audience measurement.
- BARB – Television history: 1984 and the introduction of the peoplemeter.
- BARB – Frequently asked questions, including overnight and consolidated viewing.
- BBC Year Book 1985 – contemporary description of the meter-and-diary system and the planned change.
- BBC Year Book 1986 – contemporary explanation of the peoplemeter, household handsets and overnight telephone data transmission.
Image credits: Older television set by Rfc1394, Wikimedia Commons, public domain. Television audience photograph by Alan Light, Wikimedia Commons, CC BY 2.0. The photograph is reproduced with attribution; the licence permits sharing and adaptation with appropriate credit.
