From Privacy-Safe To Privacy-Sensitive: A Better Way To Approach Data
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Femi Taiwo, head of consultancy, Europe, at Assembly Global.
It’s no secret that marketing is transforming. Everybody wants to own or build their own ad network (we see you, Apple … and the resurrection of Quattro Wireless). But a concurrent push toward privacy complicates those efforts.
A 25-year-long tradition of data mining will not change overnight. So, the question is: What will data look like in a privacy-first world?
Through decades of research and experiments, we know consumer behavior is rarely binary. The same will hold for data in a privacy-first world. Marketers will need to scale their data usage based on contextual privacy sensitivity and specific use cases. This sliding scale will consist of three classifications of data: features, services and adjacent data sources.
Features: Broadening points of interest
Instead of relying on detailed information about the user, we will move toward “features” that enhance our understanding of the marketing context.
If you opened your favorite map application to find “nearest supermarket to me,” you would see a series of icons that look like a shopping cart to signify how close one is to you. Unlike a point of interest, a feature would create an area around that pin, then assign it a random value, ensuring privacy is further protected.
Features can help us approach the principle that led our industry to a hyper-focus on data, while better understanding the contextual opportunity for stronger advertising.
Everything from ethnicity, religion, cultural mores or regional regulations might raise privacy sensitivity. Where sensitivity is high and deemed to be beneficial to the advertising context, advertisers can rely on “featurized” elements, based on location, to target potential customers without exact locations. One example would be a financial organization trying to attract customers for its Sharia-compliant funds.
Services: scaling categorical behavior
Where privacy sensitivity is deemed medium-high, scaled “services” can be used for marketing activities to replace or supplement existing data sources. Think of “services” as an externally held data source that can be queried to deliver information back to the requester.
For example, instead of trying to gain a detailed understanding of a user’s purchasing behavior by scanning their receipt, we might leverage a scalable source like Open Banking. This can provide category-level classifications while being aggregated to meet tighter privacy standards.
Adjacent sources: expanding with unorthodox data
This is where every brand gets an opportunity to put its stamp on its marketing activities. The nuances an organization provides, or industry-specific data points, fall in the realm of first- and second-party data.
Consider a consumer electronics brand selling TV units, augmenting its marketing activities with low privacy-sensitive data, such as housing development plans or broadband connectivity data. The same brand may benefit from access to open banking data for transaction volumes/sources (not amounts).
Moving from privacy-safe to privacy-sensitive
Much focus is given to privacy-safe methodologies, but the true challenge is having data that can be classified appropriately and used to build models for privacy sensitivity. Data usage in a privacy-first world should focus on treating privacy as a series of sensitivities on a scale, reflecting the position of most active legislation.
So who will be the winner of marketing’s next privacy-centric age? Those who can find data sources that are not directly related to marketing activities but yield valuable information in privacy-sensitive ways. Expect future marketing to supplement each phase with unorthodox data sources with great penetration among the common populace that can give the brand a competitive edge.