NEWS5 July 2018

Channel 4 to test AI targeting for ad placement

Data analytics FMCG Innovations Media News Technology UK

UK – Channel 4 is launching an AI-driven technology that will allow adverts to be placed next to relevant TV programme scenes during breaks, backed by research that finds contextual ads are more effective.

The technology automatically pinpoints specific moments within a linear TV programme that present opportunities for advertisers. The launch is backed by research Channel 4 conducted with BDRC Continental, which suggests viewers are twice ( 101%) as likely to remember an advert when it is viewed contextually.

Channel 4 will bring together contextual moments from shows such as hot drinks (for example, a character drinking a cup of tea), food, relationships or mobile phones, with marketers able to purchase ‘bundles’ relevant to them.

The lab test research with BDRC simulated a real viewing experience. Viewers were shown nine different brands’ adverts throughout three different programmes, either in a regular TV ad placement, in a contextual placement or were not exposed to the brand ads at all.

The study found contextual ads led to an improvement in spontaneous awareness (+34%), positive brand perception (+12%) and purchase intent (+13%) compared to regular ad placements. Additionally, 72% of respondents preferred the contextual method of advertising, according to the research.

Audio and visual data will be combined to automatically identify positive scenes, while negative moments will be discarded, the broadcaster said. The technology will still have human oversight to make sure placements are ‘brand safe'. 

The broadcaster is seeking brands to partner with to trial the technology, with effectiveness results to be presented to the industry later in the year.

Jonathan Allan, chief commercial officer at Channel 4, said: “Our ‘Contextual Moments’ trial offers advertisers the ability to buy bundles of relevant advertising within a quality and safe broadcast environment for the first time ever.”

@RESEARCH LIVE

0 Comments