Is TV measurement measuring up?

All measurement is flawed. That's the tough reality TV buyers, sellers and researchers have been dealing with for years. But that doesn't mean TV measurement couldn't be better - some say much better - than it is right now.

All measurement is flawed. That’s the tough reality TV buyers, sellers and researchers have been dealing with for years. But that doesn’t mean TV measurement couldn’t be better – some say much better – than it is right now.

The current methods have three main shortcomings, say advertisers and their buyers: small panel sizes can result in misleading data, especially when it comes to the new digitals; not enough markets are being measured; and systems still only measure audience for the shows, rather than how many people actually watch the ads.

None of these complaints are new. What is new are technological advances that may finally allow BBM Canada and Nielsen Media Research to address these age-old concerns in a cost-effective manner. How well the new methods will satisfy the buyers, though, remains to be seen.

Ron Bremner, VP of television for BBM Canada, says the introduction of Arbitron’s Portable People Meter (PPM) – a pager-sized device worn by panelists to measure TV viewing and radio listening wherever the subject travels – will enable the company to increase sample size and also move into more markets, because it is more cost-effective.

‘As researchers, we recognize the value of meters and therefore want to introduce them into as many markets as we can. We want sample sizes that will produce reliable data day in and day out,’ he says. ‘It’s a very expensive proposition to introduce meters, so we need both buyers and sellers committed to the process and prepared to support it financially.’

In addition to BBM’s move into francophone Montreal with the PPMs (see ‘BBM begins setting up Montreal People Meters’ on page 8), Bremner says Calgary and Edmonton are two markets high on the list for expansion.

Nielsen Media Research also has a number of research projects underway designed to improve the compliance of its panels, and thus the quality of the resulting data.

One of the studies is on heuristics, a method of identifying patterns in TV viewing. Mike Leahy, president of Nielsen Media Research Canada, says that right now, the meter prompts viewers with a flashing red light if they haven’t pushed the appropriate buttons to indicate that they’ve changed channels or left the room. Reducing the prompts, he says, would result in less panel turnover by making the system less demanding on existing panel members.

‘We’re also looking at Biometrics, where we learn how people interact with TV. For instance, I might know that’s Suzy because we’ve learned her pattern of viewing so I don’t have to prompt her anymore because I know she’s there. This will reduce the burden on the panel, which will allow us to do bigger panel sizes.’

The development of software panels is also being investigated, says Leahy. This truly passive method involves downloading measurement software right into set-top boxes required for viewing digital channels. The software could automatically record which channels were watched for how long – without any interaction from the viewer – and transmit that information back to a central database.

‘This is a concept we’re working on south of the border with some technology partners,’ Leahy says. ‘When we get it perfected, we’ll take it to [cable and satellite partners] and say, ‘Here’s a way we can increase these panels.’ But we’re still going to need the traditional panel because moving to a software panel alone would only give us access to 30% of the homes in Canada.’

Such a system, while giving a more accurate reading when it comes to channel surfing, still can’t tell when the viewer steps out to grab a snack or picks up a magazine during a commercial break. And that’s a measurement quandary that won’t be solved even with the introduction of Arbitron’s PPMs.

David Shiffman, research director at Starcom Worldwide in Toronto, says that’s a real problem. The PPMs are designed to record viewing by picking up signals encoded in TV programming audio. But if they’re not encoding commercials, he says, the PPM is not picking up actual commercial viewing – a shortfall he calls ‘ridiculous.’

‘The other fear I have with PPMs is that with a set-top meter the panelists press buttons so you know that at least for some of the time, people are in front of the TV,’ says Shiffman. ‘I think with the PPMs, you go from real listening or viewing to ‘you’re in the vicinity’ and that’s what’s being captured. We have questions about attentiveness.’

That’s a concern shared by Bob Reaume, Toronto-based VP of media and research for the Association of Canadian Advertisers.

‘We would hate to see this technology end up in Canada as it’s being tested,’ says Reaume. ‘We’re not saying it is bad technology. Let’s just be sure we’re going to get reliable numbers and cost-efficient technology before we go ahead with it.’

In fact, right now Reaume doesn’t feel advertisers are getting what they need from either system.

‘We would like commercial audience measurement,’ he says. ‘We would like technology that would give us second-by-second viewership, as opposed to broad, average program audiences.’

The first measurement service to address this concern could get a real leg up in the ongoing battle between BBM and Nielsen. It’s a battle that escalated recently when CanWest Global cancelled it’s subscription to Nielsen as of Sept. 1, which showed that the services could no longer rely on all media owners shelling out for both (right now, CTV and CBC still use both services but generally base their rates on Nielsen data).

The subject has been a touchy one since 1995 when the industry thought it had a solution in place: BBM Canada would handle local market measurement through diaries and Nielsen Media Research would provide national audience data via people meters.

Eighteen months later, Nielsen announced it would measure Toronto with meters and BBM countered by partnering with Arbitron and getting into the people meter business. The industry has been in an uproar ever since.

In Leslie Ingham’s opinion, neither service is better, they’re just different – and both have their flaws. Ingham is research manager at Optimedia Canada, which has been a big supporter of BBM over the years because it is tripartite and non-profit.

‘There are benefits to tripartite – and equal amounts of frustration,’ says Ingham. ‘In the committee world, it takes longer to get things done; however it is done in the interest of the whole industry. And let’s face it, BBM is not-for-profit and you can see that when you see the difference in cost between the services.’

Ingham says Nielsen wins hands-down when it comes to customer service and support, but BBM members are not willing to pay for more service.

‘As tripartite, I have the ability to effect some change. With BBM, it’s a matter of how the members vote. If we’re cheap, we get what we pay for.’

Bruce Baumann, media director for TBWAChiatDay in Toronto, says he’s always sat on the fence when it comes to measurement systems because he believes BBM and Nielsen have different strengths.

‘I don’t think we’re doing our clients a disservice by using either of the two services. They’re both good and legitimate and try to run themselves honorably and above board. But there are going to be differences just because there are different people on the panels.’

Rob Dilworth, VP of research for CTV, adds that, in fact, the strength of having two services lies in those differences.

‘They give different answers. When you do two surveys of different people, you’re going to get different answers. It stops us from making decisions based on a survey that’s wonky – and that can easily happen.’

BBM begins setting up Montreal people meters

In just a few weeks, BBM Canada will start setting up the infrastructure – household and station participants – for its first Portable People Meter market in Canada, which will be francophone Montreal. Ron Bremner, VP of television for BBM, says the PPM program is expected to be online and ready for commercial use about a year from now.

The delay in the rollout of the PPM test in the U.S. is not an issue for Canada says Bremner, and BBM is convinced the pager-sized meter is ready for prime time.

The PPM reads signals encoded in radio and the audio component of TV broadcasts, transmitting the results to a central databank daily through its docking/recharging station.

BBM originally claimed the rights to the PPM for Canada in 1992 when Arbitron of New York first presented the new technology. Since then, Arbitron has conducted more than 25 separate tests on the PPM, the latest being in Philadelphia, with its U.S. partner, Nielsen Media Research.

The rollout of the U.S. trial was recently delayed so that more research could be done on response rates and the implications of ‘ambient’ signal pick-up.

Thom Mocarsky, VP of communications for Arbitron, says the company still has to explore whether the increase seen in viewing and listening during testing is due to the units picking up ambient noise, or whether people really are watching and listening more than expected.

New survey looks at TV satisfaction

Canadian Media Research of Ottawa has a new national survey on Canadian attitudes toward television in the field. Barry Kiefl of CMRI says the main purpose of the research is to chart consumer awareness and satisfaction with the growing number of TV channels available today via cable TV, DTH or over-the-air – including the new, struggling diginets.

The TV Quality Survey (TVQS) will also look at consumer awareness of and satisfaction with a large number of Canadian and American TV series, as well as evaluating the brand strengths of various conventional and specialty channels.

The survey is being conducted this month by Comquest, a subsidiary of BBM, using a mail questionnaire sent to 1,500 anglophones aged 2+. The second wave of the TVQS, with a sample of 750, is planned for January through February 2003.

CMRI will release a written report on the top-line results in November 2002 with a cleaned and edited version of the raw data file for all TVQS subscribers.