As the digital marketing world prepares for what is estimated to be somewhere between 50-70% of the browser market to go somewhat dark, we take a look at what the current status is and what brands can do to prepare themselves.
OOH! Campaign Breakfast Briefings are back! Join us to discuss all things Out-Of-Home – live and in-person. But register now. Tickets are limited. Click here for more details.
First, a quick recap. Third-party cookies (aka 3PCs) will be gone by late 2023. Those are the things that track a user around the web. They’re implanted on your browser when visiting sites, not by the site you’re on themselves, but by… third parties. Google wants to stop that. Firefox and Safari made the call a while back so it’s not new, except that it affects more than half of the world’s web users in one shot.
Here’s an example of a third party – Say genericmarketingblog.com has a live chatbox. That chat box is implemented by adding a piece of code from another company (a third party) to your website. That company will include its own tracking cookie within that code. So, when you visit genericmarketingblog.com, the chatbox will add their tracker to your browser and begin seeing your behaviour on the web from then on, until you clear your cookies. This helps in customizing your experience and ensuring you get served ads that are relevant to you, among other things.
You can imagine those who built companies around third party cookies are less than happy and think that the playing field will now be heavily in favour of big tech who have oodles of first-party data from their “walled gardens”, like Google, Facebook and Amazon. But that’s for a different conversation.
Problems with losing 3PCs
There are a few major hurdles to overcome with this change. One is that ad targeting will get really hard without having the behavioural data we previously did as well as not being able to retarget, and another is conversion tracking (Keep in mind that a conversion doesn’t only apply to e-commerce). We will no longer be able to get a view of the conversion funnel as we did before if we can’t track a user, and our advertising partners (like Facebook) will not get past the conversion information meaning machine optimization takes a hit.
However, there are workarounds being developed.
Where are we now?
As with any problem a business is faced with, the first port of call should be a question. I.e. what you are trying to achieve. From this point, you’ll be able to start looking for relevant solutions. In this situation we find ourselves in, there is no one size fits all solution and there will be many more developments in the run-up to the 2023 deadline. A deadline extension that was thrust upon Google due to backlash and what they say has “become clear that more time is needed across the ecosystem to get this right.” (1)
One example is the ever-evolving Privacy Sandbox. This is Google getting heads together to try and find a solution that fits business needs while respecting user privacy. Thus far, FLoC (Federated Learning of Cohorts) has been the most noteworthy outcome. FLoC will essentially replace 3PCs by grouping users into “cohorts” based on their behaviours. Sure it sounds a lot like cookies, but there will be privacy built-in by removing individual identifiers, plus you will be re-assigned to a new cohort every week based on your previous week’s browsing. If a cohort does not have enough users, it will be grouped with another cohort to try to ensure there is no way a user can be identified individually by being part of a small, easy to analyze herd.
There are concerns that FLoC will not fix some of the old problems, create new problems and still provide some way of “fingerprinting” for ad tech companies. Those ad tech companies (DSPs) are worried that this move will just strengthen Google’s hold on the web while it prevents them from getting the same access to users for the purpose of personalization. The DSPs who can best personalize content are sure to get their clients a better return on their investment, so it’s no wonder they’re looking for ways to keep things as close to the current status quo as possible.
Additionally, Amazon, Firefox, Brave and DuckDuckGo are all blocking FLoC. For Amazon though, the reason is not about privacy. It’s about competition. They don’t want to feed their sizable and valuable data to a competitor and they don’t need what FLoC will offer. They have their own aspirations of gaining ground in the advertising space with their DSP (because of course, they do).
It’s a messy situation that is changing all the time. There have been other developments in the “Privacy Sandbox” that aim to plug holes in FLoC and provide other functionality. Recently we’ve heard of “FLEDGE” (previously “TURTLEDOVE”) which is said to solve the retargeting dilemma and a host of other bird-themed solutions. A hotly debated topic – All we can advise for certain is that you need to stay tuned!
Some of the large platforms including Facebook, Snapchat and Google themselves, have begun rolling out enhanced conversion APIs which allow a website to pass a user’s conversion information straight back to them in a hashed/encrypted form, instead of via the browser, thereby solving the conversion tracking problem. The issue is that as we have to now bypass the browser, server-side data collection is not just a nice-to-have any more, it’s essential.
Server-side data collection
This will require publishers and brands to own and manage the data they collect. You can arrange this by integrating any number of cloud services, which Google has an offering of their own, so you can set up and pass data through a GTM (Google Tag Manager) tag. However, the server rental will come at a cost depending on the amount of data you collect. There are multiple other cloud server providers, but at this stage, even GTM is in a beta stage and the knowledge base around it is very small. This will make implementation a bit tricky.
The alternative is to use a technology partner that already specializes in this type of data unification, like a Customer Data Platform – CDP (Tealium for example).
The biggest brands are already working with CDPs in most cases or building their own. Those that aren’t, will surely have to in the months to come. The advantages of working with an existing CDP like Tealium are plentiful. You’re not just renting an easy-to-use server that’s managed securely. You will use your data in a meaningful way while automating big chunks of work. You’ll also be able to marry online and offline data to get a holistic and single view of your customers.
CDP’s can collect your data, harmonize it, and make it usable, thereafter send it to your ad server. Your data will get put into real-time use. Instead of having to manually segment your audience based on their on/off-site behaviour, and then passing that on to an ad platform, it will all get done at lightning-quick speed. Using logic rules will allow you to automate the process, so future actions are made based on machine learning.
That’s just a tiny snippet of what they can bring to your business, but we will leave for another article on how a CDP can boost your marketing team’s efficiencies and overall performance.
Companies now have to make choices on what their strategies around data will be. Do you go it alone and build internal capabilities, or do you add complementary services to your tech stack? That depends a lot on who and where you are in your digital maturity journey, but the answer for most is fairly clear.