Here are the questions that we’re asked most often about our plans for the future.
What is BARB's scope for audience measurement?
BARB’s first priority is that we must measure viewing to all programme and commercial content that is distributed by our established base of broadcast customers.
We need to do this across all platforms and all devices as this represents the vast majority of current monies at risk. We measure viewing live at the point of broadcast and timeshift viewing up to four weeks after broadcast. We can also measure viewing to non-linear content that hasn’t been broadcast in the previous four weeks.
Beyond this, our customers’ primary requirement is to assess programmes and commercials viewed through online platforms, such as Amazon Video, Netflix and YouTube.
How can BARB deliver total reach across all devices?
Calculating total reach needs to recognise the challenges posed by a fragmenting market. This naturally demands larger samples to deliver robust results across the whole market.
Our solution for robust measurement of total reach relies on combining two different data sources: people data from our established panel and data sourced directly from the devices that are used to watch online TV.
Our representative panel of homes is critical. Without this, it’s not possible to understand the reach of a programme/campaign, demographic viewing profiles and the number of viewers per screen.
BARB also collects census data for online TV viewing. We’ve been publishing the UK’s only fully-audited, joint industry measure of online viewing since September 2015. The viewing figures are generated from software code that’s been added to TV player apps such as All 4, BBC iPlayer, ITV Hub, My5, SkyGo and UKTV Play.
Fusing these data sources will deliver robust cross-device reach for the television and advertising industry. We call this Project Dovetail.
Will cross-platform reach be available for commercials as well as programmes?
Yes, BARB will publish data so that advertisers know how online VOD campaigns increase the reach and frequency of linear advertising campaigns. Data will be available at the advertiser/brand level and by device type.
By way of an example, Advertiser A could see that its campaign for Brand B achieved the following.
- Linear campaign: 66% reach at an average frequency of 3.4 on TV sets.
- On-demand campaign: 12% reach at an average frequency of 1.6 on TV sets and computer devices.
- Linear + on-demand campaign: 68% reach at a frequency of 3.6 all devices.
Will BARB provide the trading currency for online VOD?
Online VOD data will be published for media planning purposes. This reflects other parts of BARB reporting, such as 8-28 day timeshift and dynamically inserted advertising, which are not part of the industry-accepted trading currency.
How does BARB deal with dynamic ad insertion?
BARB is the first television industry currency in the world that is delivering a measurement of dynamic ad insertion. Our service is designed to achieve two objectives.
- We report an aggregate level of commercial impacts that are attributable to dynamically inserted ads.
- We ensure that these commercial impacts are not attributed to advertisers whose linear ads have been swapped out for dynamically inserted ads.
We currently deliver this solution for Sky’s Adsmart service, although we don’t provide a verification service for individual Adsmart campaigns.
Can BARB measure audiences for Amazon Video and Netflix?
Yes we can, in the same way that we report our panel members’ viewing to non-linear programmes.
The solution, launched early in 2016, relies on the rights’ owner providing a copy of its programme assets to Kantar Media. This material is added to the audio reference library that’s used to determine which programmes have been watched by our panel members. BBC had provided over 300 non-linear assets for inclusion in the library by December 2016.
This solution could be used by the owners of content distributed through any SVOD service.
We are also exploring how router meters can be deployed in panel homes to deliver an aggregate level of viewing to SVOD services.
Can BARB measure viewing on YouTube and Facebook?
BARB delivers a joint industry measurement for online viewing from a census count; this is possible because software code has been added to online player apps. This code can be implemented by any online platform that wishes to be part of the BARB currency.
BARB is also open to receiving server data direct from an online platform with two provisos. BARB needs to report data based on average duration audience principles, while we also require independent auditing of the process by the platform’s servers to capture, clean and deliver data to BARB.
What are BARB’s metrics for online TV viewing?
Our metrics for online TV viewing are comparable with established currency units such as average audience and TVRs. These are based on the principle of an average duration audience.
There are two metrics at the heart of our plans for reporting online TV viewing; average programme streams and average ad streams. The calculation for each metric divides the total amount of time spent viewing by the duration of the programme/ad.
These new currency units have been ratified by JICWEBS, the industry-owned body responsible for the independent development of standards for online measurement.
Is BARB’s online measurement vulnerable to click fraud?
BARB commissions an independent auditing process to reinforce the trustworthiness of our published viewing data.
The audit process verifies that data we receive has removed three types of activity: a) viewing from IP addresses that are located on the business premises of the broadcasters whose programmes are being measured, b) non-human traffic, and c) traffic from outside the UK.
The audit process is conducted by ABC (Audit Bureau of Circulations).
Can BARB use registration data that is collected by TV players?
BARB is interested in the demographic information that is supplied by people when they register to use a TV player app. The benefits of online TV census data are amplified when we know something about the home/person the data are coming from.
We are currently investigating what registration data are available and the associated privacy implications.
What is Project Dovetail and when will it be delivered?
BARB is an independent body that measures viewing of television programme and commercial content, regardless of delivery platform. Project Dovetail delivers our strategy of dealing with fragmentation by combining the strengths of the data that we collect from our representative panel of people with data collected from the devices that are used to watch this content.
There are four deliverables.
- Generate census data for online TV viewing: BARB achieves this through software code that’s embedded in TV player apps used by viewers. Over 30 different platforms have implemented this software and been audited by ABC to ensure the data meets BARB’s standards. The results are reported each week in The TV Player Report.
- Determine how people watch on personal computers, tablets and smartphones: Software meters are installed on the personal computers and tablets of our panel members. These software meters deliver information about what has been watched on these devices and who was watching. Currently, over half of eligible homes on the panel have these software meters installed; we are exploring how this might also be installed on our panel members’ smartphones.
- Establish a fusion methodology: Dovetail Fusion will blend the data generated in the previous two steps. Kantar Media was appointed to deliver Dovetail Fusion on February 2017. It’s scheduled to launch in March 2018.
- Work with return path data: Set-top box date offers similar benefits to the online TV viewing census data that BARB is collecting. BARB has conducted successful pilot projects with data from Sky homes and expects to start further pilot projects in 2017.
What is the TV Player Report?
The TV Player Report is the first joint-industry, audited measure of online TV viewing in the UK. It’s the first deliverable of Project Dovetail.
It reports on the consumption of online TV content in TV player apps. The TV Player Report relies on census-level information that’s generated whenever anybody, not just a BARB panel member, watches programmes on a TV player app. Over 30 diﬀerent platforms have implemented the necessary software to be included in the report.
How will Dovetail Fusion work?
Dovetail Fusion will enable our customers to analyse the relationship between online and offline viewing behaviours by fusing two data sources.
- BARB panel: Our representative panel of 5,100 homes delivers evidence of who has watched what, and on which device. Software meters are installed that track how panel members watch on personal computers and tablets.
- Online TV census data: BARB reports the level of viewing through TV player apps across the whole population. These viewing levels are based on device usage and don’t represent the number of people that are watching.
In simple terms, Dovetail Fusion will align audience levels to the census data, while audience profiles are aligned to the panel data.
The process starts by converting the census data into people-based audiences for programmes and commercial events.
- How much do people watch online TV together? An all individuals number is calculated by applying a viewers per view factor to the average programme streams.
- Which types of people watch online? Demographic profiles for online viewing of programmes and commercial events are based on how BARB panel members have viewed the programme on their computer devices.
- Does this work for all programmes? Dovetail Fusion uses contextually appropriate variables to generate audience profiles. This is because the number of panel members viewing a programme on a given day will not be enough to generate a robust profile across the board.
- What about for online VOD commercials? The process is no different, in principle, to that of programmes: online census data creates an accurate assessment of delivery, while the panel provides the demographic context.
Having converted census data into people-based audiences, Dovetail Fusion will deliver two output files.
- Every online event that is viewed each day will be reported against an agreed set of audience categories.
- A respondent level file will allow detailed analysis of behaviour trends, including the assessment of how much online platforms are adding to the audience reached through a TV set.
A fuller description of the Dovetail Fusion methodology is available on request.
We expect Dovetail Fusion to launch in March 2018.