Why Did Google Delay The Death of Third-Party Cookies (Again)?

Google delayed the death of third-party cookies to the end of 2023. Consumers want more online privacy. How can advertisers navigate a cookieless world? Keep watching to find out!

Hi folks, it’s Ari from Sharethrough here with a brand new episode of Behind Programmatic Headlines. In this video, I’ll cover why Google is delaying the disappearance of third-party cookies, what that means for you advertisers and some potential alternatives. Without further ado, let’s get into it. 

Google Delays Depricating Third-Party Cookies

Google’s Chrome isn’t the first browser to stop supporting third-party cookies. Firefox and Safari already did it years ago.

Curt Larson:

So I think the important thing and what I’ve talked about in several contexts in the past, is to remember that third-party cookies today only exist on Chrome. And 40% of the internet exists outside of Chrome. So there’s already a huge audience of people not accessible via third-party cookies and what advertisers have largely done over the past 5 years it’s been evolving, it was 5 years ago that Safari got rid of third-party cookies, is they’ve largely ignored that change and unintentionally biased their spend towards Chrome users. 

Which means biasing away from people who own Apple devices, people who own iPhones and Macs, which is a pretty attractive segment of consumers to market to for many products. So I think advertisers are relying on Google and are maybe feeling a sense of relief, wrongly.  They should still be looking at “how can I reach consumers that don’t have third-party cookies, if nothing else, I want to reach consumers on Safari and Firefox.”

So until they get rid of them, third-party cookies are here for a little while longer. But wait, what are third-party cookies?

What Are Third-Party Cookies?

Third-party cookies are a small piece of code that are created when a user visits a site. The code stores their browsing history, settings, and any other piece of information provided or set by the user. Third-party cookies carry over to the next site the user visits, whereas first-party cookies are used by the initial website. Advertisers use third-party cookies to target and retarget ads, which is why if you left shoes in a site’s cart, ads for those shoes could follow you as you browse different sites. 

First-party cookies are used for things like email marketing, and generally perform better than third-party cookies, but can be difficult to get from users. First-party cookies can be obtained through site forms, purchases, sign-ups, etc. Third-party cookies used to be enabled by default, and users would have to disable them or clear them manually. But now, sites are required to ask consent from users if they can collect third-party cookies.

What Are Cookieless Environments?

Apps don’t use cookies, and instead, use something like Apple’s ID For Advertisers (IDFA). Since the introduction of Apple’s AppTrackingTransparency (ATT) framework, users have to directly enable tracking between apps.

CTVs are also a cookieless environment. CTVs use apps to access content, so no third-party cookies there either.

Curt Larson:

So in CTV environments, you know, you have user-level targeting, it’s typically through an IDFA type solution. The biggest challenge in CTV-type environments is more about conversion tracking because people don’t do anything on CTV. They watch an ad, and they ultimately, potentially, convert via buying your product or going to the car dealership offline. It’s historically, and still, very difficult to track successive CTV. You are seeing some innovation there, for example, Sharethrough launched our QR codes product which is really designed to bridge that CTV to mobile divide. 

And I think there are other ways to do that. Not only are you allowing your user to action against your ad, but you’re now able to track that and understand the effectiveness of your ads. And even if most users aren’t scanning your ad, you’re still able to use that as a signal to know which ads are working better and look at different tactics and different markets and different underlying organic content environments. So you’re seeing some innovation there. 

Most of the innovation you’ve seen on CTV is about gathering audience data so you can target the right user, so there’s a more sophisticated ability to do that than there has been in the past to say “Hey I want to reach these type of users and be able to find that audience on CTV” because people have built audience segments based off CTV IDFAs. What’s still really lacking is the conversion tracking, where’s still early days there but that’s where Sharethrough’s QR codes fit in.

Curt Larson:

So if a user’s scanning it, you’ve immediately gathered one piece of data at least. Which is that your ad was working. A user watched it. And you can use that to compare different creatives, different channels, different times of day, different geos and start to understand which ads are resonating more with users. Even if all users do is scan the code and never take further downstream action. But beyond that, you’ve been able to also link the user’s ID on their web browser to the ID on the TV and help you build a device graph of understanding “oh I know this user, they’ve been on my website before and they just scanned my QR code” so you can start to build that type of data. 

You also understand something about the user like “I now know this user watches this type of content” so I can start to build some information about them: they were watching a sports show, I guess they’re a sports enthusiast.” So it does allow advertisers to build more data about them. But realistically, we know a low percentage of users will ultimately scan a QR code. It’s not necessarily something every user is going to do when they see an ad, so it’s most interesting to use it for comparison data. Comparing one ad to another, one channel to another, one geo to another. 

Because even if you don’t have a high scan rate, you can still compare relative scan rates, scan rate being how often a user scans a QR code, you can still compare those relative scan rates to understand how your campaigns are working. 

What’s The Big Deal With Third-Party Cookies?

Third-party cookies don’t preserve a user’s privacy and demands for greater online privacy have consumer and regulatory bodies like the GDPR and CCPA, pushing the ad tech industry for stricter measures and controls. This isn’t fundamentally a bad thing, and the ad tech industry is wholly onboard, it’s just hard to fill that void left behind.

Curt Larson

The reason advertisers haven’t necessarily adapted to the lack of cookies on these other browsers is because their campaigns are still delivering, and they’re still potentially hitting their campaign goals, they’re just unintentionally doing it outside of Safari and Firefox. And so they’re limiting their reach, and by the way, it’s making those environments less ad dense and less competitive. So if a marketer can find a way to reach consumers on Safari and Firefox, they’re going to reach them at a lower CPM and less ad load on a page because many marketers are concentrating their spend on Chrome. 

Why is Google Delaying The Cookiepocalypse?

Creating a viable alternative that still respects the user’s privacy while providing performance for advertisers is rather difficult. That’s why Google had to delay the cookiepocalypse… again. The original date was the beginning of 2022, then the end of 2022. Now the death of third-party cookies is delayed to the end of 2023. The reason stated is there isn’t a viable alternative yet.

Curt Larson

And Google’s really stuck in a difficult position here. On the one hand they want to support consumers and regulators are looking for in terms of increased privacy and controls but on the other hand, removing cookies diminishes the ad tech ecosystem in its current state and potentially advantages Google, which creates an anti-trust concern. So they’re really stuck in this strange position and I think we all expected them to delay. 

I think there’s a bit of a question mark as to whether they even ultimately do eliminate third-party cookies or do they move around a different direction giving consumers more controls and transparency but still maintaining third-party cookies, I’m not sure where that will go over time. 

Regardless though of what Google does, there’s still an imperative for advertisers to adapt to be able to succeed in reaching consumers who aren’t accessible through third-party cookies, whatever Google chooses to do.

And they want more time testing out their Privacy Sandbox and Topics API.

How is the ad tech industry reacting?

Curt Larson:

So I don’t think Google delaying was a surprise to anyone in the industry. If you look at what’s happening with the regulatory environment, particularly in the UK with the CMA, the Competition & Markets Authority, it seemed pretty clear that they weren’t going to allow Google to do this in their original timeline.

Justin Shelby:

It did not come as a surprise and I think that advertisers are still looking to adopt some form of alternatives to cookie tracking because privacy regulations are not going to change, it’s just going to expand and increase. We’ve observed that in EMEA specifically but North America is taking more measures towards that. 

Yet, a few advertisers already started testing out cookieless alternatives.

What Are Some Alternatives To Third-Party Cookies?

Universal Identifiers

There’s UIDs. Which are similar to third-party cookies but more privacy-centric.

Justin Shelby:

The most similar to third-party cookies are called Universal Identifiers. They allow advertisers to maintain those use cases mentioned like targeting, frequency management, some form of measurement in a privacy safe way. But the good thing is Universal Identifiers they also increase the value of inventory on the publisher’s side. 

Because there’s that sense, the idea of the users, the consumers, who’s browsing on the publisher’s website to find content of interest is going to give his consent like, “I’m willing to share my email to access content” and this is the new value behind universal identifiers because they’re techinally cookieless and don’t necessarily require the presence of third-party cookie to accomplish those advertising use cases.

Some examples of UIDs are UID2, ID5, IdentityLink.

Seller-Defined Audiences

There’s also Seller-Defined Audiences, which are much like the customer persona that advertisers use to guide their decision making,

Curt Larson:

The idea of seller-defined audiences being first-party data that a publisher has acquired and can offer to a marketer to to help increase spend on their site. And we’re facing a couple of headwinds to make that really effective, and there’s some good solutions. One is the actual seller-defined audiences spec from the IAB which at least now defines a standard by which sellers can transmit this data to their buyers. The other thing is sellers need to get better at accumulating this data. 

I think there’s been a bit of a block where sellers thinking around “I’m not a publisher who can ask for a user login and password and maybe a credit card, then I don’t have first-party data.” Which isn’t true. You do accumulate data about users as they use your site, even if you never proactively ask them for anything. If you can understand this user “who I can track with a first-party cookie looked at the following articles on my site, I can learn something about them from that data. And furthermore, I can build lightweight ways to ask them for more information over time as they use my site and use that request to unlock free content on my site.” 

So I think we can publishers can really look at the value exchange they have with users and understand how they can get on an appropriate level of data from users for the value they’re delivering. Not every publisher is the New York Times that can ask for a credit card to use their site but even if you’re a smaller publisher that maybe can’t get users over that hurdle, you can start to ask users for that data. 

So that’s one piece, publishers accumulating more data. Then how do you transmit it to the buy side and that’s handled through the seller-defined audiences spec, but then the last missing piece is right now if you put that data in the bidstream to DSPs, it really gets ignored. It’s not something that buyers are actioning on today and that’s where the exchange role comes in, in packaging and marketing that data and being able to say “we have a deal ID that consists of sports enthusiasts based on first-party data across our publisher footprint of thousands of publishers.” and go and market that to buyers, to enable publishers to see increased monetization from that first-party data. Which of course, then incents them to invest further in building more first-party data and creates that cycle. 

And I think that cycle hasn’t really started to get created yet except for publishers who have really valuable first-party data. If you’re WebMD.com, you have some data that people are really interested in understanding what kind of health solutions someone is looking for, and pharmaceutical companies are going to go beat down WebMDs door to find that. But other than those kind of exceptions, I don’t think we’re seeing a lot of money transact in first-party data. 

But we have the pieces in place to make it happen, and it’s something at Sharethrough we’re investing a lot in and we have some cool products in the works there, and other folks in the industry are investing in that as well.

Seller-defined audience are created by the publisher or SSP. Some examples of seller defined audiences are Sports Fans, Financial Readers, Mobile Gamers, the possibilities are endless.

Cohorts

Groups of individual users that share common attributes. Based off non-private identifying signals. For example, Sports Fans = 3k users, Financial Readers = 5k users, Mobile Gamers = 10k users.

Contextual Targeting

Contextual Targeting is also an option.

Justin Shelby:

Another one that has resurfaced and is not new is contextual targeting because on the publisher side you could leverage contextual signals based off how your user is browsing through your website and identifying patterns and categorizing them under a few buckets, is basically what contextual targeting is. It doesn’t rely on any third-party cookies, and it’s great to expand your reach. It’s applicable on any browser, including Chrome, but also Firefox and Safari.

For example, a user is on a page about tropical destinations, AI can understand the content, and deliver ads for travel, resorts, hotels, and more.

Curated Inventory

There’s curated inventory. SSPs sift through their publisher sites, SDAs, cohorts bundling those together for advertisers to meet their campaign objectives. For example: If advertiser wants to reach sports fans, SSPs can sift through their inventory and bundles sites that sports fans are likely to visit.

How Can Advertisers Succeed Without Third-Party Cookies?

Implement a privacy-first approach and use it as a guide for all their campaign efforts.

Curt Larson:

Imperatives that are moving away from third-party cookies is simply the fact that users use more and more devices every day, which essentially diminishes the effectiveness of third-party cookies. If you’re doing intent and brand marketing on your desktop, the user ultimately converts on their mobile, they viewed the product on the tablet, they saw an ad on the CTV. Those are not necessarily able to be linked, there are some solutions that attempt to link those, but those kinds of things diminish the effectiveness of third-party cookies. So even were Google to announce tomorrow “hey we’re never getting rid of third-party cookies,” if I was a marketer, I would still be looking to implement cookieless tactics to reach consumers that aren’t on Chrome.

You can begin A/B testing out different browsers like Safari and Firefox. Safari represents a lucrative audience of Apple users and advertisers also have the chance to lower their CPM since most advertisers are focused on Chrome.

Justin Shelby:

So the fact that Google Chrome is further delaying the depreciation of third-party cookies is opening a window of opportunity for advertisers to further explore how they can pinpoint and find their strategy and quality they saw on other browsers that don’t rely on cookies. 

You can also combine different tactics like contextual targeting and seller-defined audiences, and testing out multiple combinations to find what works best for your campaigns. 

Well that’s all for this episode. Will Google actually remove third-party cookies? And will ad tech find a viable replacement? What will advertising look like in a cookieless world? Let me know your thoughts in the comments below! And don’t forget to like, subscribe and click that little bell too for more Behind Programmatic Headlines. This has been Ari from Sharethrough, until next time!