Since the early 1980s, European networks and advertisers have benefited from audience currencies that provide the reliable measurement of the vast majority of television programming and advertising.
Although many channels have lamented the number of zero-rated spots, the audience for these is only a small proposition of overall viewing, and the vast majority of TV viewing has been well measured providing a stable currency that underpins TV ad sales.
This stable ecosystem is now under attack from two sides. Three of the most significant video providers, Amazon, Netflix, and Google, are not participating in the joint industry committees, and are actively resisting independent measurement. Neither has to sell advertising, and both believe that their audience data gives them a competitive advantage that they don’t want to share with their competitors; this is significantly reducing the coverage of TV viewing that the measurement systems cover. Even on the free-to-air platform Freesat, people spend more time watching Netflix than they do on the iPlayer. This weakens Barb in the UK much more than the launch of pay-TV did.
On the other side, advertising is become much more targeted; this, in turn, means that fewer people see each advert and it can no longer be efficiently measured using an audience panel. Agencies are subsequently developing new methods to measure the effectiveness of advertising. Going beyond the reach and frequency measures provided by measurement systems, closed-loop attribution is now commonplace in the United States and now starting in Europe. These systems fuse TV viewing data with first-party data from the brands to attribute individual sales or actions to specific views of TV advertising. Should this become mainstream, this “last click” approach to TV advertising effectiveness will diminish the need for traditional currency for swathes of inventory, particularly for advertisers that are seeking some direct, measurable response such as a software download, website visit, phone call, etc.
Now an audience currency only retains its value if it covers enough of the audience for it to be representative. With two of the most significant OTT providers not playing ball, the situation is tricky. Furthermore, industry measurement committees are actively having to remove addressable advertising from their panels, so the proposition of advertising measured is decreasing even more than the amount of content.
With less and less being measured, what does the future hold for these currencies?
The UK’s Channel 4 has followed the Netflix model to develop its in-house data service, combining multiple datasets to create its own view of the audience. Many US networks are adopting the same approach. Brands and agencies who are also building in-house data infrastructures revel in the opportunity to combine their datasets with those of the broadcasters to provider bespoke views of campaign effectiveness.
At the same time, increasingly large datasets are making their way to market from pay-TV platforms and Smart TV vendors. Advertisers are embracing them as an alternative to the traditional currencies. Set-top-box data can already provide census level coverage for pay-TV channels in many markets, and Smart TV data covers free-to-air channels for millions of households.
It is likely as these datasets grow in coverage of the market that they will overtake those provided by the measurement panels which currently act as the as the only source of truth for exactly how many devices and homes have seen any individual show or advertisement.
With multiple ways to measure advertising and numerous sources of viewing data, it is likely that the joint industry committees will evolve from being the single source of truth for TV viewing to one of many sources of truth. As one of many sources of data, they will only remain relevant if they open up their datasets for fusion with third parties and remain the authoritative source of data on individual viewing.
The days where a single platform can look to provide comprehensive measurement are over. The future will be all about delivering datasets that can work well with the other datasets in the market.