BARB News

First of all, she set out a brief history of the last two years when she was engaged to oversee the implementation of a brand new television audience contract.

'Why change a system which was serving an industry perfectly well? Because the revolution in television called for a revolution in techniques for monitoring television audiences. The number of channels runs into two digits, round the clock, and the advent of digital TV is provoking another turn in the revolution. And there have been major shifts in viewers' habits. We needed a new system of capable of absorbing future technologies, one that would have a lifespan of at least 5 – 8 years.'

She then described the process of placing the new contracts, and recruiting a new and larger panel that would be more truly representative of the national patterns.

'28 international bids came in and 4 discrete contracts were awarded, to three contractors – ATR, RSMB, and IPSOS – covering engineering, technical and research expertise. By getting all three on side, we gain a number of benefits – any future changes can be made independently, additional quality control checking between the companies is provided, and BARB gained three times the research expertise.'

'We undertook the largest ever TV survey census in the world, 80,000 at the start, settling down to 50,000 continuously. From that, we are currently drawing the eventual 5100 panel households. The panel will offer more robust ratings but is also better weighted to reflect the viewing behaviour of 24 million TV households in a growing multichannel broadcasting environment. The panel has been fitted with a unique metering system, bespoke to the UK dynamics in broadcasting.'

She then went on to the lessons learned in adjusting to technological changes in such dynamic industry.

'We had to confront and absorb significant and cultural changes right up to the wire as we were designing and implementing the new system. To embargo these changes would have meant BARB launching in January 2002 without critical component parts being measured. We took the decision to ultimately be able to accommodate these changes.'

'This remains an ongoing problem. New technologies in broadcasting, the increasing satellite and cable channels, interactivity, and new gizmos will always be an unwelcome challenge to us if we cannot monitor what is being transmitted.'

'We are reliant on the technology providers and platform owners to work with us. Sky has allowed us access to their set top boxes to extract their viewing data at source. When we have access to the source, we can extract the most comprehensive information from transmission. As broadcasting technology evolves, there will be increasing demand for more of these to be measured.'

She went on to talk about the difficulties of doing this 'if you cannot forge a relationship with the source itself.'

'We are cognisant of the fact that asking these organisations to share commercially sensitive information with BARB is not an easy proposition. This typifies the problems we confronted with the digital terrestrial and cable platforms – let's just say they had a bit more at the top of their agendas than television audience measurement.'

She described the implementation of the alternative, but more complex solution for ITV digital and digital cable – Content Tracking System (CTS) which relies on sound matching. 'This is bespoke metering solution for the UK and we continue to work with ATR to address any of the technical limitations and teething problems that new metering technology faces.'

There were also problems with installation and unexpected technological challenges.

'It takes about three hours, or more, to install the equipment into a panel household. Once it was ascertained that direct service information access was not available, early cable and digital panel households required an additional piece of kit to be installed in the meter.'

'New metering technology not only has to cope with any changes in the broadcast environment but also with the diverse telephone and TV-related equipment in the households. The introduction of cable telephone households and the more sensitive digital phones meant we had to find solutions to prevent phones ringing in the night and security alarms going off when data was polled.'

There were also difficulties in recruiting the new panel, and ensuring that it was representative of the television viewing public.

'On a world-wide basis it is becoming harder to persuade the public to participate in a continuous panel, even harder to recruit the old and the young – one category is more mobile, the other more security-aware.'

'The new panel has been recruited against more stringent selection procedures, designed to be more geographically and demographically representative. This has been achieved by increasing the London and Midlands regions. RSMB conducted extensive tests on the old panel to determine the optimal weighting scheme. We now have more consistency and more reliable audience categories for most trading categories. Even at below full-strength, the panel is still proportionate by region and platform.'

'The level of trading categories continue to multiply; there are more than 1500 targets. The design of the new contract incorporated "compromises" to ensure that it was flexible enough to cope with the demands. The explosion of new channels continues unabated, meaning many channels have a share of viewing less than 0.5%.'

Caroline McDevitt made it clear that BARB can ensure that there is no additional bias against lower rating channels.

'It is a statistical inevitability that the smaller the ratings, the greater the sampling error. What BARB can ensure is that there is no additional bias against lower rating channels. BARB continues to offer the most comprehensive statistical piece of competitive analysis in the broadcasting market, e.g. if you are an EMAP Channel, you may be buying into the data to get a snapshot of your 3 million fickle target market as they sample the product across a week or month.'

'We always give a health warning to any channel that signs up. They need to take an intelligent view on what the data can realistically tell them.'

So, is BARB cutting the mustard?

'The proof will be in the data itself. What is being produced is showing signs of stability and consistency. The latest data for week ending 17 March shows that hours viewed are at last year's levels. All individuals, all adults, and the 16-34 year-olds, are reflecting a steady trend, as each week goes by.'

'We also looked at the trends for the last two years. Comparing the last quarter of 2001 with the last quarter of 2002, taking out the research effect of the new service, the trends were remarkably similar between terrestrial and multi-channel.'

'As the quality control checks continue and the panel members remember to push buttons and register both theirs and their guests' viewing, we start to see positive results. Part of this process is to assess genuine records of viewing, e.g. is the panel member present when he/she says so, have they pressed the buttons? If it is not, and the panel member's participation is poor, then they are removed from the panel. This improves the quality of the data but at the expense of the size of the panel. It needs a proportion of new homes needs to be replaced. This all takes time and is to be expected. The data is paramount and the size of panel at launch at circa 4000 remains robust enough to produce sound data.'

She described the presence of continuous queries.

'There are, of course, queries or anomalies arising from the data and for certain audiences, especially for smaller and regional samples and for certain sub-demographics.'

'The service has been designed to be more transparent which also leaves it wide open to misinterpretation by the uninitiated.'

'BARB data is a sample, not a precise science, and requires sensible interpretation. Television measurement provides the industry with the most reliable "crutch", way beyond other forms of media research.'

'With the recession hitting the service industries hardest, it is not hard to see why there has been a reluctance in some quarters to accept the long-anticipated changes.'

'Overhauling the UK audience metering system has been necessary. We now have a system capable of absorbing future technologies and moving with the times. Our system has always been a blueprint for audience measurement systems across the world – the UK is the only country to measure all six analogue and digital platforms, and the extent of complex broadcasting and regional and macros. It is also the most transparent. BARB figures are objective, and they are credible, and form a central point of reference for the UK television and advertising industries. This system will serve British broadcasting, its programme makers, its advertisers and agencies for many years to come.'