The ‘premium internet’ promise and the case for owning your ad IDs

Welcome to the May edition of Spotlight, your monthly digest of what’s making news in the world of digital marketing and adtech.

What we have for you today: OpenAI has a shocker of a month, publishers explore alternatives to the Privacy Sandbox, and the Colossus case shines a light on the need for publishers to own their IDs.

Not so smart

OpenAI and ChatGPT have been in the headlines again this month, but not always for the right reasons. The launch of?ChatGPT-4o?was marred by the?resignation of a senior safety?officer who is concerned that the company has neglected safety guidelines and is too focused on launching “eye-catching” products. And then came accusations that the company has modelled its chatbot’s voice?on Scarlet Johansson’s?against her wishes.

Privacy concerns also continue to mount. In this case, OpenAI is?facing litigation in the EU?for failing to comply with an individual’s request to rectify an incorrect birth date generated on ChatGPT, placing it in potential breach of GDPR. The case once again raises the question of how far Large Language Models can comply with privacy laws and the principles of consent. Marketers beware.

The sinking sandbox

After the not-so-shocking news last month that Google was once again delaying cookie deprecation, comes the equally unsurprising news that publishers are getting tired of waiting around for Google to get its Sandbox in order.

As Digiday found, while publishers?will not pause testing?of the Privacy Sandbox completely, it no longer has their full attention. Importantly, publishers?appear set to continue?the journey away from third-party tracking cookies, regardless of what Google does or doesn’t do in the months ahead. It seems that publishers are no longer waiting on Google and are instead proactively looking to alternative approaches to privacy-centric advertising. Our recommendation is for publishers to test as many alternatives as possible to ensure optimum coverage.

Privacy, US style

In the US, publishers have a choice of complying with privacy laws on a state-by-state basis or choosing the strictest set of regulations and applying them across the board. Google has?come out in favour of the latter approach, announcing that it supports “the MSPA US National Privacy Technical Specification (National String) as part of the IAB Tech Lab’s Global Privacy Platform (GPP).”

This is a logical approach. Ultimately privacy is about individuals and their ability to control how their data is used and it makes sense to apply this in a uniform way. What’s more, given that privacy laws are tightening up the whole time, applying the most stringent practices using the most compliant technologies helps future-proof organisations.

Walled-in advertising

Walled gardens provide useful data on authenticated users, but outside the large players such as Google, Meta and TikTok, they lack the scale that advertisers really need. Now, a?proliferation of small retail media networks?is making this situation much worse.

This approach should be reconsidered. Retail data is significant for advertisers but in the current environment, brands would have to stitch together many scores of retail networks to achieve the reach and scale they need – something that few advertisers will want to do. Instead, retail networks should look to?privacy-first solutions?that enable them to reach beyond the walled garden environment to be interrogable with the open web. This would enable brands to reach retail audiences at scale.

Live from New York

This month saw The Drum Live symposium take place in New York. Regular readers of Spotlight will recognise many of the themes raised by marketing professionals during the event, such as the importance of first-party data, the proliferation of regulation as a driver for change in the industry, and the critical role of data collaboration in the post-cookie world (we would add that collaboration? with telcos for their network intelligence is the perfect example of just such collaboration). Check out the?event summary here.

The rise of the “premium” internet

There was?an interesting take?from The Trade Desk on how the internet is splitting into one characterised by spammy content on the open web, distributed via the walled gardens, and “the best of the open internet, which they are calling the “premium internet.”

This is well observed, and in fact extends beyond the main content of webpages to advertising. On one hand, spam advertising that has no relevance to the end user, and on the other “Advertising as a Service” – content that matches the user’s needs and wants perfectly, which is presented in the right context, and reaches them at the right time. Advertising should be about providing a service to web users, not just enhancing cheap inventory.

A colossal mess

Finally, there’s?the Colossus debacle. The supply side vendor has been accused by Adalytics of misrepresenting user IDs to attract high bids. The accusation is vehemently denied by Colossus, but regardless of whether it is true or not, one thing is clear: brands and publishers should not rely on using other people’s IDs.

If you retain control of your ID, you are directly attaching your audience or an event that you have selected against the new user, providing a level of protection that you can’t get when you’re ceding control to a third party. The story also reveals the weakness of probabilistic identifiers, which are based on best-guesses rather than solid data and which cannot scale.?Our advice: own your ID and choose deterministic data every time.

More from Novatiq

We are attending Telecom World Middle East in Dubai 28-29th May. Get in touch if you’re attending the event and would like to meet up, or if you would like a call after the show. As always, you can also learn more by reading our latest blogs. This month we’re looking at the latest trends and solutions in privacy-conscious?digital marketing in the US.