Adam Powell, Head of Data, Kinetic UK
What is an audience and how does it behave? It’s a question that many in the digital media and marketing industry haven’t had to think about. If the last decade could neatly be defined as the age of targeting, fuelled by a free flow of personal data, the next decade or so looks uncertain and messy when it comes to just how the individuals which brands would like to reach through digital channels are found, corralled, understood and reached.
Google may have delayed the removal of third-party cookie tracking from Chrome for another year, but the direction of travel is clear. Digital media will likely shift in large part from deterministic to probabilistic audience data analysis. The momentum for change will be nudged and prodded by ever greater regulatory scrutiny and action in Europe and the US.
With huge investment in first party data now building a head of steam within brands as they transform their internal operations, will add another powerful dynamic to how customers are understood at an individual and an audience level.
The idea of audience, an entity created by the pull of content, the need to move or the power of contextual forces in everyday life will very likely become a central defining element of the media industry.
What is changing is our growing ability to understand an audience. Rather than attributing basic demographics aligned with a particular event, we’re beginning to be able to understand at a far more granular level, the type of people who are present, why and what their presence might mean and what the outcomes for brands could be.
The OOH sector has, over the last decade undergone its own journey of transformation and digitisation. We at Kinetic began trading programmatically in 2019 and by last summer all the major agencies and media owners were capable of automated trading of space on digital sites.
Programmatic trading is an important step but given that by its nature OOH is all about audiences not individuals, the really significant aspect of the industry’s evolution is the increasing sophistication of contextual data analysis and our understanding of audiences at a macro and micro level.
OOH advertising has traditionally been a broadcast medium and in many respects this continues to be the case – it delivers brand or public messages to huge numbers of people building awareness, affinity and understanding.
The way OOH media was used during the pandemic to support the NHS by government and brands illustrated powerfully the public role OOH media plays with a mass audience who know the message is being seen by fellow citizens.
However, connected digital screen technology and data analytics has enabled OOH to also become an addressable medium in much the same way as smart TV. Fast, near real-time creative delivery to a multiplicity of screen formats organised by postcode or other geographical definitions combined with the shared audience data for each individual site provided by Route can then be combined with bespoke or open-source contextual data and optimised by bespoke analysis.
Brands can now leverage viewer movements, interests, and habits through location data coupled with demographic information on contextual triggers beyond weather, live news, sports or other significant events and time of day.
The range of contextual data available is huge including everything from maps, the national census, traffic, air travel, shipping, logistics, economic, sales, cultural – this list is endless.
For example, Kinetic’s behavioural planning platform Journeys fuses multiple behavioural and attitudinal data sources, including Adsquare’s mobile SDK data and retail footfall data allowing us to plan very detailed localised campaign plans around micro audiences and up-wards.
This opens creative possibilities for brands to deliver personalised creative based on real-time data backed by customer insights and market research. The application of first party data from brands within this mix of insight means thinking in groups rather than individuals will gain momentum.
Our understanding of the behavioural impact on people of different formats in different contexts is also gaining ground. Kinetic recently published the results of a three-year joint research project with Clear Channel, Global and JCDecaux, Wavemaker and Mediacom that analysed £2.2bn of media investment including 46 brand campaigns to provide planners with data to demonstrate the role different elements in the OOH media mix can play.
We have now scientifically proven which formats build long-term brand memorability, why frequency builds a positive impression of a brand using formats in certain contexts, the impact of extended exposure to messages and the role of high-quality formats. We’ve been able to layer this insight into the planning logic that powers Journeys. The result is even richer context-aware that can act as a substitute for personal data.
Google has slowed the pace of transition but it has been explicit about the need to move on from cookies and IDs.
The new audience infrastructure “Topics”, hot on the heels of their recently scrapped ‘Federated Learning of Cohorts’ (FLoC) it had previously been modelling, is Google’s latest attempt at replicating similar planning capabilities currently enjoyed by OOH.
Topics works by analysing your browsing history to work out the things you’re interested in. If you like cars, for example, Topics will show you adverts for cars on the websites that you visit. To work out that you like cars, each website that uses Google’s Topics API will be assigned an overall category. A website about tattooing, for instance, may fall into the body art category; a city newspaper would likely be assigned to the local news category.
As you move around the web, Chrome will record the categories you visit the most. Then, each week, your five most popular categories will be gathered up—Google says this process is done on your device and not on its servers—and a sixth random topic will be added to add some noise in the system. These six categories are then shared with the websites you visit and are used to target the ads you see. The data is deleted after three weeks.
Under FLoC, people could have been grouped into more than 30,000 different categories, which would allow advertisers to gain specific knowledge of their interests. This information could then be combined with other data to build up an incredibly detailed picture of each and every one of us. This is less likely in Topics, as there are around 350 interest categories that can be assigned to people.
But again, grouping audiences into these broader, non-personally identifiable groups and delivering to audiences at scale sounds awfully similar to OOH’s current model, doesn’t it?
And as a growing number of brands experiment with the programmatic flow of campaigns across media, translating campaigns from channels on private screens on to OOH media in public spaces creating increasingly useful, relevant and pleasurable brand experiences, there’s never been a better time for the digital media community to reconnect with the concept of audience.