What defines today’s most successful tech companies – Google, Amazon, Meta, representing more than 2.5 trillion in market cap value combined (Statista) – is access to volume, velocity and variety of data, as well as the application of the most sophisticated AI. more importantly, their unique ability to unlock and connect demand and supply data intelligence at scale has driven the creation of highly impactful AI. One could argue that these companies’ entire business model relies on this ability to monetize the demand and supply intelligence of the advertising ecosystem.
We, at Choreograph, define Demand Intelligence as the data and interaction signals that are created when people engage with brands directly. Supply Intelligence is the vast data generated from digital media engagement, where people spend on average more than six and a half hours per day, according to GWI’s Q3 2022 study.
Google, Amazon and Meta – unlike other companies – have access to both demand data, from search, commerce and custom audiences, as well as supply data to feed their AI. For that reason, they have set the gold standard of how to compete and win in the new AI-driven environment for achieving transformational growth.
Brands have learned to identify the demand signal and have invested heavily in understanding it and organizing their response to it. On the other hand, the supply signal has always been elusive, and with data legislation, fragmentation and tech privacy on the rise, it is becoming unattainable, despite brand spending more than $800bn in 2022 on generating it (Statista).
Getting ahead of this challenge requires brands to manage four core misconceptions.
- It is impossible to understand my audience because people’s digital interactions are hyper fragmented.
- I need to move or share my data outside of my own environment.
- Addressability requires the use of highly private individual-level consumer data.
- Measurement is impossible.
Misconception no 1: it is impossible to understand people across their fragmented digital experience
According to Gartner, only 14% of brands believe they have a 360-degree view of their customers. The reality is that, even in those cases, that view only refers to demand data, leaving untouched the vast amount of intelligence generated when people engage with digital content across platforms on the supply side. Most enterprises mention data fragmentation and lack of common identifiers as the reason a complete view of peoples’ digital experience is impossible. This is because they still focus on the narrow view of individuals, rather than a privacy-centric anonymous deep pattern understanding. Brands’ obsession with resolving identity at an individual level is the equivalent of looking at a tree and missing the forest. Most digital signal data, approximately 80% according to the IAB State of Data Report, 2022, is not structured and therefore not addressable. Creating actionable data representations of this vast amount of unstructured data that can fuel model development and drive media decisions represents a real opportunity to create a sustainable competitive advantage.
The solution to uniting the two siloes lies in the application of AI. More specifically, it is in the application of federated learning and the ability to tap into supply-side data – to better understand the complex anonymous patterns that define key audiences – that is the game changer.
At Choreograph, we have brought together for the first time – in one user interface designed for our agency partners and clients – supply, demand, and third-party unique intelligence. Media planners can mix newly available publisher signal data, advertiser first-party data and industry-leading global research, with unique taxonomies like Lifestyles™, Pathways™ or Occasions™ without ever moving any data.
Misconception no 2: I need to surrender my data to activate it
It is true that there are existing tools that attempt to provide methods of privacy-first data-sharing, but they are neither seamless nor futureproof. Clean rooms, where data is aggregated and anonymized, do have the potential for deep insights, but present challenges in terms of scale and ID-match accuracy.
More importantly, they still require the movement of sensitive data, which represents a significant risk that many organizations are not willing to take, and nevertheless potentially falls short in relation to privacy standards.
The solution is a decentralized data model where data can remain in its native environment – never moving – and still connect across partners and supply-side intelligence.
There is precedent for succeeding in peer-topeer knowledge-sharing across industries where the protection and movement of sensitive data is thoroughly scrutinized. Take financial fraud detection, for example. The banking sector depends on competing financial institutions sharing information in a compliant way across regulatory borders to tackle this one trillion-dollar problem. In health, too, companies gain access to patient-level intelligence across hospital systems to formulate individualized cancer treatments. In both cases, data never moves from its native environment.
At Choreograph, we have built a technology solution that deploys the algorithms to where the data is, and connects intelligence to drive AI growth predictions – unlike every other approach in the market today that still require centralizing data to run algorithms over the top.
Misconception no 3: addressability is about IDs
It is old news that identifiers will soon become a thing of the past. Nevertheless, marketers still need to deliver the right message to the right audience at the right time.
The solution is to tap into a scaled partner ecosystem to unlock supply intelligence that goes beyond buying media. Deploying edge computing and using federated learning allows for data to remain in its native environment and, at the same time, contribute to creating algorithms that drive growth predictions at scale using data that was never accessible to advertisers before. This is how demand and supply intelligence come together to power AI predictions that, until today, were impossible with fragmented data.
At Choreograph, we have successfully applied federated learning techniques to optimize AI models using data held at the edge of publishers’ and advertisers’ native environments. Infusing these algorithms with a colossal amount of never-seen-before granular data can drive better audience definitions, planning and activation without ever sharing any data, thereby preserving privacy and data ownership for every participant.
Misconception No 4: measurement is impossible.
Today, the consumer journey looks more like a game of snakes and ladders than a clear linear pathway. But that does not mean that understanding the impact of all interactions and continuously optimizing the customer journey needs to be a guessing game. New ways of thinking about audiences requires a reset of how we think about measurement.
The solutions, again, rely on the ability to share deep granular intelligence across supply and demand data, enabled by technology that preserves privacy at its source, without ever moving or sharing data. At Choreograph, we have developed a measurement protocol, that does exactly that, allowing for a more accurate understanding of the impact of advertising investment to drive continuous optimization.
Unlocking intelligence powered by algorithms that can learn at the edge of advertising demand and supply data represents a meaningful change in the advertising ecosystem. It requires a massive shift in spending, and a complete rewrite of ad tech infrastructure. This fundamental change cannot be accomplished overnight, and the industry will need strong leadership to guide the way. WPP is positioned to lead this transition. With more than 42,000 professionals dedicated to planning and executing campaigns at WPP media agencies alone, we wield significant influence.
By adopting a new approach to deploying algorithms and mining vast volumes of advertising supply data, we can influence a new wave of campaigns and spending. This collective effort will help create a more effective, privacy-compliant advertising landscape that benefits advertisers, publishers and people alike.
Sources:
https://www.statista.com/statistics/209331/largest-us-internet-companies-by-market-cap/
https://www.statista.com/statistics/236943/global-advertising-spending/#statisticContainer
https://www.gartner.com/en/newsroom/press-releases/gartner-marketing-survey-finds-only-14–of-organizations-have-ac
https://www.iab.com/wp-content/uploads/2022/09/IAB_State_of_Data_2022_Preparing_for_the_New_Addressability_Landscape.pdf
Yannis Kotziagkiaouridis
CEO Resolve
Carsten Hyldahl
CPO Resolve
