Members of IAB Europe’s Programmatic Trading Committee Have Come Together to Compile a Q&A Blog on Measurement in a Post Third-Party Cookie Environment
The impending termination of support for third-party cookies on Google Chrome means that many of the metrics that marketers use today to measure their digital media performance will soon become obsolete. So how will we measure in a cookie-less future?
Josephine Howe, Sr. Marketing Manager, MSA, Daniel Godoy, Microsoft Global Programmatic Evangelist, MSA: There is no silver bullet solution for measuring programmatic outcomes. However, industry ID solutions will add a valuable layer to measurement of digital campaigns with authenticated IDs. Advertisers will also need to rely on non-addressable signals, for example how website visits changed during the campaign, and metrics like attention measurement.
Raman Sidhu, SVP at Verve Group: The objective of third-party cookies was to enable cross-channel (multi-touch) tracking and measurement. However, they have not been perfect for a while as the environment in which they run is already limited, most notably for Apple users. So we need to remember that cookies are already not doing what they were designed to accomplish.
As a result, new models must be considered:
- We are already measuring consumer relevance and interaction in today’s cookieless world, as opposed to retargeting tactics which are limited in function.
- When it comes to sheer awareness, we can measure the context that drives affinity and performance towards the relevance of the consumer, in turn understanding incrementality.
- There are also new media metrics–attention is one where we are putting a lot more emphasis begins–which have different inputs currently focused on media placement as opposed to audience, recency and relevance.
When we’re talking about conversions, we need to think about the type of conversion and where it’s taking place. Whether it be ecommerce, leads, sales, app downloads, etc. All of these have alternatives without cookies, IDs or PII which are proven & measurable. One of the main differences is that it’s focused on the last touch or via a data-driven model.
Alex Berger, Senior Product Marketing Director, Adform: The third-party cookie is already a shadow of what it once was. New solutions have been introduced and scaled that are ready to cover the majority of targeting and measurement use cases. At times they’re more precise; in other situations, more complex. But, the key is that advertisers retain the ability to be strategic. Today, they have to navigate a more complex and strategic technology landscape to do that while the industry goes through a period of unbundling. As the third-party cookie further erodes, and technologies continue to come online I think we will see the trendline bottom out and reverse towards a re-bundling which will translate into significant scale 2–5 years from now. Ad performance for branding doesn’t really change radically from third-party cookies. Measurement for Branding campaigns is a matter of quality of exposure with the right reach and frequency. Quality of exposure doesn’t need third-party cookies. To measure reach and frequency, you need cross-identity measurement to unpack the effectiveness of your strategies. But this is generally the same regardless of if they are based on contextual, cohorts, first- or third-party IDs. For Performance you need an approach that still relies on real-time decisioning, but further validates that against longer-term, deeper reference and additional data.
Louise Mbugua, Sr. Product Manager & Go-To-Market Lead, Teads: We must first acknowledge two realities. The first being that with over 50% of web supply today missing 3P cookies (3PCs), cookieless is not near, it’s here. Secondly, not all cookies are equal, with 1st Party cookies (1PCs – created and stored by the website a user visits) surviving into the “post-cookie” future. So how do we define whether ads are working in a post-cookie world? Let’s first define success. The success of an ad can be viewed through two lenses: “engagement” (user interaction with the ad) and “outcomes” (post ad engagement value such as brand uplift & sales). Ad engagement, as measured with traditional media metrics including ad viewability, completion rate and clicks, is minimally impacted by the deprecation of 3PCs. As an aside, however, with up to 20% of programmatic ad spend distributed to Made For Advertising (MFA) supply, it’s increasingly vital for marketers to evolve these metrics to include attention-based measurement – e.g., ‘dwell time’, ‘attentive seconds’, ‘attention CPM’ – with studies proving attention to be strongly correlated to outcomes and 3x more effective than viewability as a predictor of consumer behaviour. Unlike ad engagement, outcome measurement, specifically attribution measurement that relies on 3P consumer-level data, will be challenged by the deprecation of 3PCs. Here the industry is still in an experimentation phase and there is no single replacement solution, but key options marketers are likely to explore include:
- 1P data-driven attribution measurement
- Browser-based solutions such as Google’s Privacy Sandbox Attribution Reporting API, which seeks to connect a given click or view on the publisher site with a conversion on the advertiser site in a privacy-centric manner, providing both event-level reporting on individual conversion events and summary reports with aggregated, but richer data on the nature of said conversions.
- Attribution measurement leveraging universal ID infrastructure. Authenticated IDs are likely to prove more accurate and persistent than 3P cookies and enable cross-device attribution, however challenges remain on the scale of these IDs due to the requirement for user logins.
With a broad range of post-cookie measurement initiatives – complexified by nuances in the way different types of media and devices are measured – it will be key for the industry to collaborate to facilitate cohesive, effective and efficient measurement in a post-cookie world.
How do we think about using post-cookie solutions for campaign targeting and optimisation and looking across the consumer journey? Not just display, but mobile + CTV as well?
Josephine Howe, Sr. Marketing Manager, MSA, Daniel Godoy, Microsoft Global Programmatic Evangelist, MSA: In the post-cookie world, there will be two groups of users on the internet: identifiable and unidentifiable. For identifiable users, first-party data combined with consent is key to precise measurement and modeling. For unidentifiable, or anonymous users, contextual targeting through keywords, topics, and moments is optimal. Curated deals are a powerful way for buyers to benefit from these solutions at scale. With the population of identifiable users shrinking, there will be more emphasis on first-party data, especially in channels like Connected TV. There are innovative tools to address this shift, for example advanced frequency management. It is crucial for marketers to balance identifiable and unidentifiable groups and understand the opportunity to target and optimize towards each.
Raman Sidhu, SVP at Verve Group: At Verve Group, we often talk about how moments are the new cookies. Brands know the moments that matter to them, especially those which drive interest, intent and sales. We help brands deliver omnichannel campaigns which include web, app, CTV and DooH in meaningful moments across the consumer journey. Furthermore, through a real-time feedback loop we can optimise performance towards specific outcome goals. That forms the basis of our strategy. With the availability of first- party data or cohort data, we can further improve performance.
Alex Berger, Senior Product Marketing Director, Adform: Marketers need to embrace dual thought streams; real-time and long-term. Nothing new here, but the focus and importance has been amplified by third-party cookie deprecation. Real-time decisioning and strong algos don’t replace the need for a robust planning phase or deep post-campaign effectiveness analysis. You’ll still need strong media mix modeling to study the effects of omnichannel advertising. This hasn’t been replaced – quite the contrary, a hybrid approach is now more important than ever. Skip the real-time and you miss out on the value of programmatic. Skip the analysis, and you break the flywheel, which is driving incremental value creation.
Louise Mbugua, Sr. Product Manager & Go-To-Market Lead, Teads: There are many ways to answer this question and they are timing dependent. Any targeting powered by addressable identity will extend targeting reach capabilities across multiple mediums (web, CTV etc). Some of these solutions, however, will take a long time to create significant scale. As an alternative, there are other forms of targeting using contextual and panel based intelligence and those solutions will perform across several mediums.
Other alternatives include:
- Exploratory solutions such as Google’s browser-based Topics API – that assigns a shortlist of human-readable topics to users based on content consumption, which buyers can use to facilitate interest-based marketing – and the IAB’s Seller Defined Audiences framework, which relies on publishers segmenting audiences into a standardised audience taxonomy, whose segment IDs (instead of user identifiers) can be passed in the bid stream to DSPs to decide whether to bid.
- Predictive AI models, which leverage hundreds of signals not reliant on 3P cookies, to compute the probability of an impression achieving a given outcome. At Teads, we operate Smart Traffic Filters on Deal ID powered by our Predictive AI to optimise towards ad engagement metrics incl. MOAT Viewability, Completion Rate & CTR.
- Advanced contextual intelligence using tools such as Media Barometers, that identify media consumption patterns across premium editorial in order to propose optimal communication moments and facilitate audience discovery.
- SSP-side audience targeting. According to the 2022 IAB Europe Attitudes to Programmatic survey, all stakeholders from agencies to advertisers and publishers are ‘putting more emphasis on SSPs as a source of data’. SSPs, such as Teads, with robust data infrastructure including scalable & diverse 1P proprietary audience and contextual segments, are well positioned to support buyers with data-enabled PMPs.
Digital advertising campaigns in environments devoid of third-party cookies will run on a mix of emerging alternatives, including first-party data, consent, contextual approaches, cohorts, identifiers, and more. How can we ensure that these solutions still align with user preferences and future regulations? With data privacy legislation becoming more complex, how can we future-proof?
Josephine Howe, Sr. Marketing Manager, MSA, Daniel Godoy, Microsoft Global Programmatic Evangelist, MSA: Microsoft Advertising is focused on an audience first approach and iteratively building to prepare for the next chapter of digital advertising. Technology platforms must enable the industry to build and connect solutions in a compliant way. Looking ahead, we think that there will be a shift from client-side solutions to more server-side solutions, increased employment of data clean rooms and consolidation of Identity solutions. Multiple identity solutions will be needed as long as they are interoperable. The limited number of ID’s, based on the direct relationship with the user, means that there will be less reach but better authenticated traffic, better controls on consumer data, and maintenance of targeting and measurement use cases.
Raman Sidhu, SVP at Verve Group: As digital advertising campaigns move away from third-party cookies and towards alternative solutions using first-party data, consent, contextual approaches, and cohorts, it is important to ensure that these solutions align with user preferences and future regulations. Very often, I hear the same feedback from expert buyers, product and strategy leaders at brands and agencies. The future will be a mixture of first-party data and context. That means mapping 1st party data, for example to advertising ecosystems/walled gardens. Whilst creating audience lookalikes using contextual signals and predictive models in the instances where no identifiers exist. Consent must be explicit in all cases where any form of identifier or PII is involved. Ensuring double opt-in is imperative for brands and advertisers. For sellers, the ecosystems with the largest addressable consented audiences will benefit from buyers looking to engage with current or prospective customers.
For buyers, capturing of consent should include the channels where they want to activate.
Clean rooms aim to help here with the movement of data. However, there is still some way to go in scaling the network.
Alex Berger, Senior Product Marketing Director, Adform: While this is a major concern for the industry, it’s much less of an issue for technologies that were already thinking in a consented / localized way. We started pivoting our tech stack to support this approach as GDPR was starting to gain momentum and found that a focus on transparency, interoperability, and clear consent signals really paves the way for scaling post-cookie without major issues. At the end of the day, it’s about choosing technology that gives you a lot of transparency, ownership of your data, and that clearly supports delivering consented value to the end user. If you work back from that, you’ll consistently find these changes are more a case of fine-tuning vs ground-up reinvention with exorbitant switching costs.
Louise Mbugua, Sr. Product Manager & Go-To-Market Lead, Teads: The impending deprecation of 3P Cookies, set against a backdrop of increasing data privacy legislation, is rightfully forcing a re-examination & redress of the balance between user privacy and ad performance, which has often tilted in favour of the latter. The most important variable here is that the consumer must be involved in the consent process and permission obtained to use their personally identifiable data. Any form of identity or targeting/measurement that relies on inference of identity will be subject to scrutiny for the simple reason that consumers can’t opt-out of said mechanism.For contextual based and technographic based solutions, these have longevity as they don’t target any individual or PII. In doing so they provide a privacy safe scale that we – at Teads – have found performs as well as, if not better. The key will be the balance of finding a strong enough ‘truth set’ of known information from consumers who have opted-in to their data being used to power models and AI to find scale into identity-less environments.
Finally, what conversations are you having with media buyers and brands? What advice would you give them to get ready for programmatic outcomes without the cookie?
Josephine Howe, Sr. Marketing Manager, MSA, Daniel Godoy, Microsoft Global Programmatic Evangelist, MSA: As an end-to-end technology provider, we advise buyers and brands that relevant advertising will increasingly rely on contextual solutions for unidentifiable users. Investment in first-party data assets and partnerships will be needed. Advertisers should seek out data platforms and tech providers equipped to work with authenticated user IDs. It is important to think about how to leverage proprietary data assets while ensuring privacy compliance.
Raman Sidhu, SVP at Verve Group: We encourage brands and media buyers to consider the moments that are important to their business. We help them plan media to capture the largest share of these moments, whilst always identifying interest and intent. It is crucial for companies to comprehend their target markets and pinpoint the meaningful moments when those audiences are most likely to interact with their goods or services. Avoid focusing on cookies and instead focus on audiences and moments to drive demand to your products and services. Redesign your media planning strategy and concentrate on the moments that matter most to your business. This will support your efforts to increase sales and succeed in your marketing campaigns.
Alex Berger, Senior Product Marketing Director, Adform: What’s needed for media buyers and brands is more control, more transparency and ownership, and solutions that can bridge the gap in a consented way. Your activation on CTV or DOOH is going to be a bit different from premium display or mobile. You’re also going to have different IDs that are better suited to certain geographies or channel types. We’re already seeing and working with 40+ IDs on a regular basis, and I think it’s the job of the technology to create tools and solutions that really elevate and solve this complexity for buyers. Of course, the buyers also have to embrace these new approaches, and will need to re-train and re-prioritize workflows somewhat to make this happen.
Louise Mbugua, Sr. Product Manager & Go-To-Market Lead, Teads: Firstly, don’t wait to act! Delays to the deprecation of 3PCs on Chrome belie the fact that a predominantly cookieless environment is a reality today across many major markets. Secondly, understand your exposure by identifying the volume of media budgets dependent on marketing capabilities such as cross-site tracking that rely on 3P cookies. Thirdly, develop a plan to test and integrate cookieless-ready solutions into your marketing strategy.