ad tech

Welcome to the data party

Where’s the data party right now?

The difference between zero-, first-, second-, and third-party data

Data is incredibly valuable to marketers, agencies, and publishers, since it is used to power advertising models and generate revenue. And the amount of data we produce is accelerating. The global volume of data reached 64.2 zettabytes in 2020, and this is predicted to double by 2025.

However, privacy regulations, such as GDPR and CCPA, have made data an increasingly difficult resource to tap. In response, the online advertising industry has developed different classifications of data to help with the handling process. So, what do the terms zero-, first-, second-, and third-party data mean?

Four-party state

The ad tech industry uses four different classifications of data, depending on the degree of separation from the source of information.

Zero-party data

This relatively new term describes data which is willingly shared directly with the recipient, e.g., a customer who bought a sneaker submitting a survey to the brand. Arguably, this is a type of first-party data, the distinction being that it’s intentionally handed over rather than automatically collected.

Zero-party data also refers to edge computing ad solutions, where data is processed on-device rather than transmitted elsewhere online. This protects user privacy by minimising the data collected, while still allowing behavioural targeting and personalised ads, thereby circumventing legal restrictions on the amount of data allowed to be sent to the bidstream.

First-party data

Data which is gathered directly, for instance from customers, visitors, or social media followers. This includes information from email interactions and behaviours on a website or app. First-party data differs from second- or third-party data in that there is no intermediary between the supplier and the recipient of the information.

First-party data includes first-party cookies and advertising IDs which identify individual visitors, such as GAID, IDFA, and UID. Identity solutions use these IDs to track user behaviour while they interact with a website or app, in order to serve behavioural and contextual ads. However, targeted ads relying on first-party data alone can be sub-optimal, since their reach is limited to information from a single platform or site.

Second-party data

This refers to data collected from a partner who has in turn gathered it directly from visitors or customers. Although the data is supplied by another agent, the distinction made from its close third-party relative is that second-party data is sourced from a trusted partner, with whom there is a direct agreement on how the information is to be exchanged and used. The transaction takes place in a closed environment, such as a data cooperative, or a data “clean room” which provides the utmost security and privacy. Second-party data is usually used to periodically enrich first-party data to gain insights.

As with zero-party, there is some contention over the term ‘second-party data’, and the disagreement makes its definition a little woolly. In fact, when the Winterberry Group asked industry experts for the definition, responses ranged from “someone else’s first-party data”, to “there is no such thing”. After pinning down some specifics, the researchers proposed that any commercialisation or licensing of second-party data transforms its state into third-party. But this is by no means a hard-and-fast rule.

Third-party data

This pertains to any data received indirectly, via an intermediary agent who is not the original source. In ad tech, this data is usually gathered, aggregated, and packaged to be sold to companies to build advertising strategies.

Third-party data has historically been the most sought after in personalised advertising, allowing one-to-one retargeting of an online user. This can be achieved via cross-tracking cookies, where a user’s profile follows them online and is appended with data from each site they visit. Knowing so much about a user enables hyper-targeted ads to be served to them. However, third-party data now sees the most legal scrutiny due to the privacy implications of tracking users so closely across different sites, hence the phasing-out of third-party cookies.

So… where’s the party?

If all this has left you confused about the classification of data, you’re not alone. The rise of zero-party data, and the explosion of second-party data, are a result of initiatives by the industry to avoid falling foul of privacy regulations while retaining access to sources of data which make advertising models effective. In the age of big data, just make sure you know the legal requirements when handling it, so you can avoid having your party busted.

Image courtesy of Joshua Sortino

Life beyond the cookie

The third-party cookie is crumbling: what’s next for publishers?

The third-party cookie has enabled marketers to serve targeted online ads for the last two decades, allowing websites to remain free, while ensuring content publishers are paid for their work.

But now, in response to privacy concerns and new regulations, leading web browsers are quickly deprecating these snippets of code. Apple has already blocked third-party cookies by default in Safari, while in January, Google announced a complete replacement in Chrome by 2022.

Even with the death of third-party cookies looming, the majority of digital advertising still relies on them. In response to this, we recently spoke to leading industry experts about the future of the ad tech ecosystem, and asked how smaller publishers in particular can adapt.

Here are the key takeaways we found:

Dare to diversify

There isn’t a predominant replacement for third-party cookies yet, so we don’t know where (or when) the industry will settle. In the meantime, there’ll be a fracturing of ad tech, and the key strategy for publishers during this transition phase will be to sell ad inventory in multiple different ways.

The most promising solutions avoid falling foul of both data laws and privacy-conscious tech developments:

Subscriptions. While this may be supplemental to ad revenue for smaller publishers, it’s crucial to renew focus on building a first-party subscription base. The advice from industry leaders is to offer a value exchange and make users feel like an exclusive member of a club.

Contextual intelligence. Contextual targeting serves ads based on the content of the webpage (e.g., training shoes on a fitness forum), whereas behavioural targeting uses individuals’ browsing activity. Although behavioural targeting has come to dominate the web, there’s little hard evidence that it actually improves revenues. Meanwhile, AI and machine learning have drastically improved contextual methodology, earning it the moniker ‘contextual intelligence’. Many industry leaders think it’s worth betting on this supercharged comeback.

Data clean rooms. These are a legally compliant and accurate way for publishers to continue using behavioural targeting. They can compare their first-party visitor data with those in ‘walled gardens’, e.g., Google and Facebook, to optimise ad matching. However, clean rooms can be expensive and therefore not necessarily viable for publishers with smaller datasets.

Edge Computing. Marking the age of “zero party” data, this data-conservative approach is becoming more popular as a way for publishers to sell remaining inventory. Data is collected and analysed directly on the user’s device, rather than on a server, which allows publishers to serve behavioural ads while completely respecting the user’s privacy.

Test and test again

With so many options available, validating their effectiveness will be just as important. Chris Hogg, EMEA Managing Director at data management platform, Lotame, stresses that now is the time to start testing.

“Test solutions and strategies while third-party cookies are still around to compare against. Ask for proofs of concept around Safari and Firefox inventory. The fact that we don’t have cookies in some of the other browsers presents a good opportunity for publishers to test out solutions and tactics today rather than later.”

Consequently, publishers will need to be far more involved when it comes to their audience data. Mattia Fosci, Founder and CEO of the edge computing solution, ID Ward, urges smaller content providers to approach data analysis with both partnerships and off-the-shelf technology.

“While it may not be viable to hire a full-time data analyst, don’t underestimate the importance of analysis on your bottom line. Publishers should have more control of their audience data, but they do not need to build their own in-house solutions… Instead, publishers should work with partners that protect and enhance their relationship with their own audiences.”

Turn obstacles into opportunity

While ad tech’s brave new world is an uncertain place, one thing’s for sure – it puts publishers in a much stronger position than before. As David Reischer, Head of Product at edge computing solution, Permutive, explains, “The death of the cookie is a huge opportunity for publishers to course correct on what has happened, with their data being aggregated at scale, repacked and sold as audiences or models.”

Now, publishers can use their first-party data to bring brands even closer to audiences. By being prepared, respecting privacy, and fostering user loyalty, the entire industry stands to benefit from the change.

Head over to What’s New in Publishing to download the full report.


Image courtesy of Pezibear