The Holy Trinity Of First-Party Data Collection: Transparency, Compliance And Consent

Malcolm Treanor is the COO of Aussie media technology company WINR. In this piece, he shares his top tips on first-party data collection.

The death of the third-party cookie model will crumble online advertising as we know it.

For many years these trackable identifiers have underpinned programmatic advertising, helping brands determine which readers, on which sites, will see which ads, when and how frequently.

With no alternative identifiers on offer, the impact of a cookie-free world is expected to be sudden and dramatic.

We’ve already started getting a glimpse at what that world could look like.

In Germany, publishers have experienced a 45 per cent decrease in revenue and a 23 per cent drop in CPM on Firefox traffic where cookies are now automatically blocked. Similarly, a Google report in Australia showed impression-based revenue has decreased 52 per cent for publishers without third-party cookies.

First-party data has emerged as the best alternative and mission-critical to publishers’ survival. This model is based on data collected directly from readers including inferred data from their behaviours, actions or interests alongside declared data such as age, gender, self-reported preferences and intent.

The general advice given to publishers is to start collecting, analysing and activating their owned data from reader interactions with content and the like. But there is very little guidance on where to begin without compromising the trust of audiences.

The key to implementing a successful and long-lasting first-party data model is a holy trinity of transparency, compliance and consent.

Transparency, compliance and consent

Consumers and regulators are increasingly concerned over the collection and use of data by big technology companies. In Europe we’ve seen GDPR, in the US CCPA, and in Australia the ACCC’s Digital Advertising Services Inquiry and the impending Privacy Act review.

Publishers must future-proof their first-party data collection programs to ensure they create value for both brands and consumers, while meeting rising regulatory compliance requirements.

It starts with three guiding principles:

• Transparency – making it easy for readers to understand what information is being collected, how it will be used, and who it will be shared with
• Compliance – depending on geographic location, developing an easy-to-read privacy policy that addresses the requirements of current privacy legislation and where possible plan ahead for what is likely to come
• Consent – making sure readers understand what they’re getting into when they share their data as well as how to withdraw consent at any time

But the journey doesn’t stop there.

Maximising the value of consent

With more scrutiny from the public about how their data is being used, publishers are poised to provide custom audiences to advertisers with a price tag to match. This represents a big opportunity for publishers to increase yield and revert to an age of hand-sold deals.

Pseudonymous identifiers, also known as Publisher Provided Identifiers (PPIDs), will allow publishers to start maximising the value of their audiences.

With PPIDs in place, publishers can segment their readers into custom audience lists based on declared first-party data like age, gender, along with inferred data attributes based on their interactions with particular content.

These custom audiences of PPIDs can be ingested into ad management platforms to control frequency capping, audience segmentation and targeting. Publishers can include the PPID value in an ad request sent to an ad server or SSP to match ad calls to specific audience segments.

Scaling data-driven revenue with lookalike audiences

Custom audiences are the first step for publishers to successfully meet the demands of advertisers. To then go onto actually winning budgets from advertisers, the next step is achieving scale.

Lookalike modelling allows publishers to find readers who are similar to the members of a particular audience and extend the reach and scale of deals. These audiences are built via algorithms that finds similar people based on anything from age to intent. Lookalike models are used to build larger audiences from smaller segments to create scale for advertisers. The larger audience reflects the benchmark characteristics of the original audience.

For the modelling techniques to be effective it requires that a database has breadth and depth of data across common variables, starting with gender, age and postcode. The more information a publisher has about its readers, the bigger and more effective the lookalike segments that can be built. Once publishers have this type of data they can look to append privacy compliant third-party data attributes like WINR’s Kaptiv Audience Segments (KAS) to provide descriptive personas to readers and feed their data hungry lookalike models.

Getting started with data collection

The clock is running out for Australian publishers trying to figure out their first-party data strategies. To be in with a chance of surviving the next 12 – 24 months, publishers must prioritise first-party data collection. Starting with basic information – like gender, age and postcode – is a simple and effective way to build the foundations for success.

Smart publishers have already started to build out their first-party data strategies. Those that will thrive in the next digital advertising age will incorporate transparency compliance and consent into their processes from the beginning. Success will lie in creating an owned system of identifiers and implementing a cookie consent management system to match the needs of advertisers today, and tomorrow.

Image: iStock/MirasWonderland