Estimating a large audience

Millions and billions

Like many people who have worked in TV audience measurement, I’ve learned to sigh heavily when hearing audience estimates for big events.   How many people will actually watch the Olympics opening ceremony? Well, a LOT, especially if people in some really big countries such as India and China tune in, but it’s not my job to measure it, and I REFUSE to be a kill-joy about the number.  Why would I – I’d rather celebrate TV, rather than nit-pick about over-enthusiastic PR.

And besides without decent data, it’s not easy to estimate.

In 1997 while working for Star TV in Hong Kong, I was asked to forecast the audience in China for the historic handover of Hong Kong on June 30th.

My boss at the time. I used to know Wendi Deng. (Lame name-drop, but it’s all I’ve got)

Don’t be nervous, but there’s a LOT of people looking at us.
How many?
It’s too complex to fit into a picture caption

A load of exciting television was planned, and people like Tony Blair, Prince Charles and Jiang Zemin, the President of China had turned up for the raising and lowering of flags, fireworks and whatnot for the evening of the 30th, culminating in the legal hand-over at midnight.  This was regarded as a historic event of great pride to China, reflecting its new economic and political status, and its emerging pragmatism (Hong Kong would be governed under by a new arrangement called ‘one country, two systems’).

In a sign of its growing commercialism, China’s broadcasting authorities decided to raise money from TV advertising during the event, and had asked Star TV (part of News Corp) to help sell advertising during the commercial breaks.  They wanted Western multi-nationals to snap up the chance to reach so many Chinese tuning in to this historic broadcast, especially during the 30 minutes or so immediately before and after midnight.  As part of this process, I was asked to estimate how many people might watch such a one-off programme.
In 1997 China did not have a national electronic TV audience measurement system.  Like most countries in Asia at the time, it relied on people in recruited households to complete a special diary showing what programmes they had watched.  While this may have been the only possible method in 1997, it wasn’t very accurate.  Apart from the usual measurement issues concerning fallible memories, it was clear that the authorities believed that its billion population should watch.

The biggest audience for TV programmes according to the national diary system was for the nightly news at 7pm.  This wasn’t any news – it was the Chinese government’s news – and its reported audience of 400 million, was wildly exaggerated because the people in the survey felt that they should report watching it.


It’s the cover of ‘Olympian’ by Gene. I went to their gig in Hong Kong just before the handover. Martin Rossiter – the man for the big occasion.

Anyway, asked to predict the likely audience for a late night programme, on a Monday night, I estimated that around 15% would be watching – 165 million people.  That’s a massive audience – 50% bigger than the audience in the US to the Superbowl.    The Star TV management decided that the programme would get more viewers than the national 7pm news, and estimated that half the population would tune in – 550 million.

What was the actual audience?  No idea.  I actually like their estimate, but reckon mine was nearer.


I’m saying this is around 25 people watching TV, but respect estimates of up to 100

Speak Your Mind