Latest news insight

Tracking audiences in times of change

13 July 2011

BARB is 30 years old this summer, and has overseen TV measurement for the UK through some years of dramatic change. The pace of change is getting faster and the challenges of measurement seem to continue to increase. BARB has been actively addressing the future of TV audience measurement and anticipates an evolving role going forward.

BARB’s strategy priorities have been clearly identified in recent years:

  • to protect the standards and credibility of the core service, and
  • to seek to extend reporting (beyond the core service definition of TV sets and 7 days from transmission).

A key role is to enable the bigger picture to be seen – by keeping the measurement of television together.

Our goal is to continue to deliver an industry relevant, quality assured, sufficiently progressive and dependable Gold Standard Television Audience Measurement Service for the UK.

In pursuit of our development priorities, BARB has enjoyed some success in the last year – so that we’re now able to actively offer options for implementation.

Our commencement of what we intend will be a continued rollout of web TV measurement is detailed later in this article.

BARB is also creating the capability to report viewing to stand-alone content assets, as a supplement to the current practice of relating viewing to a date and time of originating broadcast. This will increase the flexibility of what it is possible to monitor and report from the BARB panel.

Currently more viewing is captured by BARB meters than is reported – due to BARB’s definition of timeshift viewing as ‘within 7 days of broadcast’. Whilst this core definition will remain, it’s possible by an additional process to identify stand-alone material and to report it separately.

This capability has been delivered in pilot form and BARB has been able to track viewing across a period of months to individual content assets where the content’s audio signature is included in a reference database.

The expectation is that content providers may wish to place audio signatures for either:

  • what they perceive as their biggest content offerings (to be tracked individually),
  • a whole collection of programming that is being made available in series stacks,
  • or a selection of content which makes up a genre proposition.

Then, via additional BARB reporting, we would expect to report who was watching, for how long, and be able to relate this to other viewing either occurring on the same TV sets or other TV sets.

We are experimenting with what may be sensible from this new type of reporting, and whether or not individual commercials that have been viewed can be identified.

Our pursuit of future priorities includes practical assessment of server and return path data – to what extent it matches metered and actual viewing – and consideration of options in the measurement of Connected TV devices.

We are also testing the use of measurement identifiers to enable different types of distribution to be able to be clearly flagged – to aid alignment with collections of server and return path data held by platform operators and content providers.

For 30 years BARB has developed and enhanced its service – and although the pace of change may be faster, we continue to be forward thinking in addressing future options and possible requirements of the industry.

TV via the Web

BARB has been field testing software meter technology since last summer and is about to roll it out to a number of homes on the BARB panel as a move towards introducing a viewing measurement of PC’s, laptops and tablet devices.

The field test was not an attempt to provide a specific quantification of the amount of viewing – more so because it was conducted on a deliberately skewed sample. The chart below does, however, demonstrate that the project was able to return and make sense of data, in that we have been able to align some viewing data from TV meters (in blue) with PC meters in the same households (the thin yellow line on top).

R&D Project Virtual Meter example data

In the test project we gathered useful understanding about:

  • viewing sessions relating to different people on the same PC
  • monitoring a number of PCs in the same home
  • the logistics and feasibility of delivering software updates to the PC meters
  • the frequency of use of PCs – the varying usage patterns affecting the number of days between data return
  • the kind of data that can be returned, and how to attribute viewing to TV content and its source of origination from the internet
  • levels of participation in combined TV and PC measurement

Whilst there are still some aspects of deployment of the approach to be considered, verified and confirmed, it is BARB’s intention to rollout to 1,100 homes on the BARB panel during the course of 2012. This would anticipate that 2,500 individuals would be included in this measurement. All homes selected would have broadband connectivity.

The initial stage will be to establish the new technique in 100 homes on the BARB panel during 2011, in order to ensure data production and processing are viably delivered from BARB panel homes. There will also be verification of the impact of the technique for the existing panel homes.

Following the initial 100 homes rollout, if successful, a confirmatory decision will be taken before a rollout to a further 500 homes and then another 500. This staged approach will be to ensure that the standards, quality and viability of the core BARB service are not jeopardised.

It is anticipated that, during the course of 2012, the industry will need to determine how data can best be utilised and applied from the planned sub-panel of 1,100 BARB homes. The prospect of a measurement which enables web-TV viewing to be examined alongside TV-set viewing, from a single source, will have a number of applications. It is likely that the greatest initial learnings will come from data aggregated over time.

The rollout of the web-TV meter aligns with other projects for non-linear reporting and assessments of the potential of server data which form part of BARB’s pursuit of the futurefor TV audience measurement.

The move to rollout the web-TV meter benefits from the test project, which continued until May 2011. Developments were identified to improve the viability of the web-TV software meter and its application – some have already been delivered and proven, whilst others are anticipated to be delivered over the rest of this year.

BARB will be utilising software meter technology from Kantar Media, research supplier for the metering and data processing for the core BARB service.