Three things we need to do as an industry to make the Privacy Sandbox work for Advertisers

We’ve always believed that there would not be a silver bullet for a cookieless world and that a multi-signal approach would be required by advertisers to navigate effectively in the future.  

Apple, Firefox and others have already removed 3rd Party Cookies (3PC), however we still have access to use these signals within Google Chrome, the world’s largest browser. However the elevated ‘user controls’ approach announced recently by Chrome will lead to more users rejecting 3rd Party Cookies. This means that 3rd Party Cookies can no longer be depended on as the overarching signal for targeting, measuring & optimising digital advertising across all browsers and has cemented our belief that we will be operating in a multi-signal ecosystem moving forward.  

Alongside other signals, browser based technologies will play an increasingly important role; not only for targeting cookieless audiences but also for bridging across the entire web ecosystem regardless of cookies (e.g. Measurement). Google Chrome’s Privacy Sandbox is the most comprehensive set of technologies so far, providing a range of APIs which enable use cases across and beyond advertising for a privacy forward internet.  

We’ve lent heavily into understanding the Privacy Sandbox and we will focus in this article on how the industry can take advantage of the new technologies in Google Chrome to support Advertisers as we shift into the next phase of digital advertising.  

Facing the Challenges Ahead

As part of our readiness programme, in the first half of 2024 we ran tests across all 3 of the advertising focussed APIs within the Privacy Sandbox (Topics, Protected Audience & Attribution Reporting) and regularly share findings with our advertisers and ad tech partners. Our tests so far have concluded that all 3 APIs functionally work broadly as we expected at this early stage, but there is still a lot more work to be done before everyone is confident in the role they can play supporting a vibrant and privacy-first open web economy of the future.

The only way we can accelerate the development and rollout of vital solutions using the Privacy Sandbox as part of a multi-signal landscape is with deeper collaboration from all sides of the industry – advertisers, agencies, ad tech, publishers, browsers & trade bodies. As we navigate through this period ourselves, we have identified three critical challenges stand out above all others in order to support advertisers interests, and should be considered in parallel with the technical feedback being provided by ad tech:

1. Signal Availability – Scalable testing opportunities and faster adoption of the Privacy Sandbox APIs, providing high reach across the open web to meet the needs of advertisers
2. Advanced Addressability – Surfacing advanced retargeting and prospecting solutions that use the Privacy Sandbox, in combination with other signals, to foster addressability throughout Chrome
3. Effective Attribution – Developing new standards and frameworks for using the reporting APIs to meet key use cases, facilitating a smooth transition for advertisers as capabilities and requirements shift.

Navigating The Future: A Path Forward

To address these challenges together, we must first delve into the key learnings that have shaped the first half of this year. Understanding these elements allows us to outline specific recommendations for the industry, focusing our collaborative efforts on areas where they are most needed.

Signal Availability

What have we learned?

We have run functional tests across targeting and measurement use cases for advertisers but they have all been ‘proof of concepts’ or capped at low levels of investment. Primarily this is due to the limited, but increasing, adoption of updates across the ecosystem (DSPs, SSPs and publishers mostly) that are needed to use the APIs. Another factor has been the CMA testing labels Chrome agreed to apply to the 1% users who have had 3P cookies removed. Separate updates are required to pass test labels and enable the privacy sandbox and we see them applied inconsistently. This reduces the % of users available to target during tests even further to a sub 1% number.

Another significant challenge to testing in H1 has been the lack of opportunities outside of the US. Many ad tech partners have chosen to initially avoid markets with more complex regulatory regulations, such as the GDPR in Europe, and focus on a single, large market which is easier to test within. Although this is not true for all solutions, especially those just using Topics signals, it is important for Advertisers to understand and validate how these globally available signals will be used across markets differently.

What needs to happen next?

Breaking through this challenge requires effort from all sides of the industry. First and foremost, it is vital that the updated ad tech which enables the Privacy Sandbox is scaled rapidly across publishers and supply side ad tech partners. This needs to be incentivised and matched by buy side ad tech and partners leaning into the additional scale available, by ramping the size of test budgets and accelerating roadmaps to introduce features into their platforms for all advertisers.  

To continue to support robust testing, we also want to see an updated approach to facilitating cookieless testing from Chrome which allows ad tech to test across greater proportions of Chrome users. This could be supported through the IAB Tech Lab’s Privacy Sandbox working group to streamline and accelerate scaling new solutions with a testing framework that all ad tech can align on. This would help improve consistent application of test/control audiences beyond the Chrome facilitated testing labels and reduce unpredictable auction dynamics. 

Advanced Addressability

What have we learned?

There are 2 things that everyone should be aware of for the future of addressability, and they are not always obvious from the conversations we have with those who are not focussed on this topic every day.  

Firstly, the Topics and Protected Audience (PA) API’s are not like-for-like replacements for cookies and we should expect some capabilities will not be replicable as privacy is enhanced. Additionally the APIs are not meant to be ready made solutions for advertisers to pick up and use; they are built for the ad tech industry. 

The APIs in the Privacy Sandbox are tools and signals which ad tech partners can choose to utilise, in combination with any other signals or intelligence available at that moment, to build their own solutions fit for a privacy enhanced future. Understanding this has allowed our teams to focus their efforts on better understanding impact and opportunity around the 2 key targeting use cases for advertisers – prospecting for new audiences and retargeting known audiences. 

Retargeting solutions can use the PA API to build and re-engage audiences that have visited advertiser sites. This use case is fairly straightforward and has significant value for advertisers, but the testing opportunities are restricted to DSPs and rely on them making testing available. 

 For prospecting, both APIs can be used to develop innovative solutions with more flexible opportunities to test, as ad tech partners can develop solutions which we can deploy across any DSP. While this has driven several tests for advertisers, even where a DSP is not actively supporting the Privacy Sandbox, a lot of the solutions available in H1 have been foundational at best – often targeting topic signals directly without any additional signal or intelligence applied. 

Most ad tech partners with advanced solutions have started enabling cookieless, and sometimes ID-less, approaches to support the future of the open web. Some of them have been testing how the Privacy Sandbox signals could enhance their products, but many appear to have ignored them or chosen to disregard the signals completely.   

We often hear that ad tech partners are instead focussing on signals which have already achieved significant scale and can be trusted to deliver results in the short term. This approach undervalues the longer term opportunity and leaves the door open for walled gardens and those with globally scaled platforms to benefit. It ignores the fact that ID-less inventory is already at a significant scale and will continue to gain share over the next few years. Advertisers want to see more test opportunities now so they can be better prepared to reach audiences beyond IDs and are interested in bringing in new partners to test. Those who can move ahead, before the Privacy Sandbox has reached critical scale, can take the most advantage from it.  

What needs to happen next?

For retargeting, our asks are simple. DSPs should bring PA API testing online faster. Ad tech partners should continue to engage with the IAB Tech Lab’s existing workstreams for the Privacy Sandbox, and provide constructive feedback to the Chrome team along the way. Similarly, we expect the Chrome team to keep working with Ad tech partners to resolve issues and blockers within their privacy focussed framework.

Prospecting is the area where we are asking for renewed attention and focus from industry partners over the next few months. We need to see more innovation and testing opportunities available this year from DSPs, SSPs and audience partners who are using the privacy sandbox signals to enhance advanced prospecting solutions that work across the open web, not just within authenticated environments. 

Below are some example areas of exploration for ad tech partners that our Advertisers are keen to see

• DSPs, SSPs and curation platforms utilising Topics as part of multi-signal predictive, AI modelled and interest-based audiences
• Publishers using Topics signals to enrich their PMP or Direct deals
• DSPs and Retail Media Networks using PA API to scale their 2PD audiences across the open web
• DSPs and data partners collaborating to rebuild audience marketplaces fit for a privacy-first world, using PA API
• DSPs, CDPs or data partners extending data collaboration opportunities for advertisers and publishers beyond clean rooms and authenticated inventory

 

Effective Attribution 

What have we learned? 

While addressability post cookie is improved through a multi-signal approach, attribution use cases for advertisers are improved through a single source of truth for a complete, de-duplicated view of campaign performance. 

The Attribution Reporting API (ARA), in combination with tools within the Privacy Sandbox, is the only option for deterministic, cross-partner and impression level attribution in Chrome post 3PCs. Regardless of differing addressability strategies from ad tech partners, integrations with browser based reporting APIs will be a non-negotiable expectation from advertisers measuring performance on their own, or retail partner, sites.

For Attribution, our leanings have fallen into 2 different categories of impact, Ad tech and Advertisers 

Ad Tech

Most ad tech partners have prioritised targeting functionality first but we have also managed to run some initial attribution tests with 2 DSPs. Ad tech partners are still in the very early stages of testing ARA and they are actively providing feedback to Chrome on the essential functionality needed before 3PCs disappear. 

One example of feedback resulting in real change to the Privacy Sandbox is ‘attribution scopes’, which allow for finer-grained control of which campaigns and events are included in each attribution model. The public development of this feature can be tracked on Github and should be available for testing in the coming months.

Alongside functionality updates, ad tech partners still need to develop their approaches to ingesting, transforming and modelling ARA data into usable optimisation signals and advertiser friendly reports. Ad tech will need to manage event and summary outputs, statistical noise, and the constraints of a ‘privacy budget’. We expect conversion reporting outputs may look very different between partners depending on their priorities (optimisation vs attribution reporting etc). 

While we don’t know yet how each ad tech partner will approach this, we suspect that advertisers could face less flexible and configurable reporting systems, creating challenges for advanced use cases such as reconciling attribution and business data at a transaction level. 

Advertisers

For advertisers, there will need to be a shift to a purposeful mindset of ‘tracking what matters’. Continuing to track and attribute every single interaction on advertiser sites could mean hitting the privacy constraints in ARA before a user has taken a meaningful action or converted,  resulting in missing out on conversions being attributed to their campaigns. This could have an impact on the number or configuration of tags our teams recommend are placed on Advertiser sites.

Advertisers also need to be auditing what media and conversion data points are currently used in data-driven and multi-touch attribution models within their organisations. The ARA does not support advanced attribution models at launch and there is no clear roadmap for introducing them. The privacy constraints of the ARA may also make it difficult to extract unique values, such as order IDs, alongside conversion reporting. Advertisers should be proactive in bringing their priority requirements to conversations with their Trade Bodies, and AdTech as they continue to build new measurement solutions.  

What needs to happen next?

Although the flexibility and granularity of attribution will be degraded, it is still required to play a key role in advertisers’ future measurement frameworks, alongside incrementality studies and media mix models. Effective attribution will continue to drive business outcomes by providing an immediate source of feedback for media planning and campaign optimisation.

It does not serve the industry, advertisers or individual ad tech businesses to have everyone develop wildly different approaches to handling ARA outputs for advertisers. This risks increased fragmentation and divergence of reporting capabilities across partners, which will only add to the confusion facing advertisers and increase the difficulty of demonstrating advertising performance within their businesses. 

Leading DSPs and ad servers should work together, in partnership with the IAB Tech Lab and advertiser trade bodies like ANA and ISBA, to develop and set new standards which address the major reporting and attribution use cases that need to be maintained. 

Advertisers also need to see clear roadmaps and timelines from ad servers and DSPs in advance of ARA based data flowing into reporting systems, alongside changes in performance benchmarks due to technological changes, to manage their own business and stakeholder expectations. Advertisers will also need support and guidance on changes to tagging or event collection needed to fit within the constraints of the Privacy Sandbox and recommendations on adapting processes to account for new limits in data availability.

How GroupM is leaning in

In November 2023, we announced the first and largest global post-cookie technology readiness program, in partnership with Google Chrome. This global initiative is bringing together GroupM clients to accelerate their understanding of Google Privacy Sandbox APIs and their use in advertising. 

Key features of the program include:

• Unified Framework: GroupM has designed a unified framework for testing to work collaboratively with advertisers and ad tech partners to understand plans to integrate privacy technologies, including Privacy Sandbox APIs, and overlay relevant tests wherever possible.
• Anonymised Data: To facilitate the fullest learning opportunity for each participating advertiser, GroupM will conduct an anonymised meta-analysis across individual brand tests, in addition to a brand’s own Privacy Sandbox program. This information will be anonymised and aggregated as part of collective learnings and eventually a meta-study.
• Accelerated Learning: Participating advertisers will get direct access to testing across GroupM alpha/betas with our global ad-tech partners. As first-to-market testers, participating clients will learn at the forefront of this industry-altering change, gaining access to GroupM’s collective knowledge.

We are actively working with ad tech and publishing partners to share perspectives and develop new tests that could accelerate our advertisers understanding of the Privacy Sandbox. We would welcome any ad tech exploring the privacy sandbox to reach out to our team to discuss opportunities. 

What advertisers should be doing

Whilst we’ve outlined what the industry should be doing to enable advertisers, we also know that advertisers need to be part of this ecosystem shift, and we believe that all advertisers should be doing three things right now:

• • Evaluate your exposure – understand where your marketing plans are reliant on 3rd Party Cookies to assess your risk, and identify your most important use cases.
  • Interrogate your partners – work with your ad tech partners to understand how they are evolving their technology and what this means for how you buy and measure media. 
  • Test early and start ahead – we believe those who test into new technology (of which Privacy Sandbox is just one) will get a head start in a cookieless world.