Manager

Audience Research

TVOntario

Toronto

One of the big issues in my mind is the ability to measure increasingly fragmented audiences and to do so at affordable cost.

There are technical complications in measuring audiences that are divided across many broadcasters, where each broadcaster is capturing only a small portion of the people who are viewing television at any given moment.

And if we take seriously the reports in the press, scores of other services launched in the next little while will complicate things further.

More difficult

One thinks that if the audience is smaller, it will be easier to measure when, in fact, it is very much more difficult and more costly.

Typically, measurement services would need to deal with larger samples, and with all the viewing choices, be very careful to construct a system that was capable of identifying exactly what it was that was viewed.

Even in a 40- or 50-channel universe, people aren’t always clear what channel it is they are watching.

Larger samples

We have to figure out ways to make samples larger, without sending costs out of reach, especially if specialty broadcasters are looking for detailed profiles of the composition of their audience, as opposed to top-level numbers.

For heavily viewed stations and programs, existing samples tend to be quite adequate in terms of providing basic information.

Certain corners of the industry are saying, ‘What about consumption patterns? What about income?’ -things which are not being picked up at the moment.

Basic characteristics

But if you want to know the basic characteristics of the audience, and you have a heavily viewed channel or program, things aren’t so bad.

But for specialty broadcasters of one kind or another, there are times when the data run dry before our questions about the audience do. We don’t have a sufficient number of cases to give a reliable answer to those questions, and that’s just because viewers are distributed across so many stations.

Current testing

We do some [qualitative research] ourselves.

The kind of testing we do is to try to get reactions to different programs in the people’s own words. It’s not a marketing approach at all that we take, at least in the sense of selling some kind of goods or service, it’s service delivery we’re interested in.

We don’t sell advertising, but if we’re going to deliver a service, the more we know about the people at the other end of that system, the better we can produce or schedule programs that are suited to the people that tune in.