Prior to the invention of the browser cookie in 1994, the internet didn’t have a memory. But this lack of memory made the early web a frustrating user experience – every time you visited a website, it was as if you were a stranger who had never been there before.

Lou Montelli, who was working as an engineer at Netscape at the time, developed browser cookies – a small text file that could be passed back and forth between a person’s computer and a single website – to enable websites to remember visitors. “We designed cookies to exchange information only between users and the website they visited,” Montelli told Quartz. “[Third-party tracking] wasn’t something that we had really anticipated sites doing.”

And yet, very quickly, cookies were embraced by the nascent digital advertising industry as a way to understand people’s behaviour by tracking them across the web. While first-party cookies exist to benefit the website owner and make a more functional and better user experience – such as remembering the items left in a person’s cart – third-party cookies exist to allow advertisers (or third parties who do not own the website) to understand user activity, enabling more targeted advertising.

The free web had to monetise, and to do so meant advertising. The result was an internet ecosystem powered by consumer data. But now, the proverbial cookie is crumbling as Big Tech moves away from cookie technology, leaving marketers and brands unsure of what is to come.

Here we examine why the cookie is going the way of the dinosaurs, which technologies are gearing up to replace it and what it means for digital advertisers and brands looking to gain and retain customers in a cookieless world.

The long, slow and very painful death of the cookie

Time was officially called on the cookie on January 14th 2020. On this day, Google announced that it would be ending support for third-party cookies in 2022. In many ways, Google was a laggard – other browsers such as Apple’s Safari and Mozilla’s Firefox had already blocked third-party cookies by default – but Google’s dominance of the browser market (it captured 68% of the global desktop internet browser market share at the time of the announcement) meant advertisers’ would truly have to find another way to deliver effective advertising without cookies.

On June 24th, 2021, Google announced it would be delaying its cookie deprecation to 2023, due to technical reasons. Many marketers breathed a sigh of relief but there is still a sense of impending doom. Even if they are currently on life support, the cookie’s demise is all but inevitable. Apple’s privacy-first announcement in 2021 that it would make its own identifier for advertisers (IDFA) opt-in emphasises this shift away from ‘tracking by default’ to consumer consent.

Regardless of whether or not browsers’ support cookies, consumers for the most part do not. They may not have strong opinions about the technology itself, but they do not like the underlying sentiment of being tracked. According to a survey by Adswerve 69% of respondents said control over the information collected about them is "extremely" important, with 39% feeling uncomfortable sharing their data with brands. And yet, at the same time, almost two-thirds (63%) of consumers still expect personalisation as a ‘standard of service’.

So while the cookie may be on its way out, marketers still require the means to understand their current and potential customers, to deliver the most effective and relevant marketing to them.

What technologies could replace the cookie?

Currently, there is no clear winner as to which technology might replace the cookie. A few that have been tested over summer 2021 include Google’s FLoC and The Trade Desk’s Unified ID 2.0 –– each of which aim to replace the cookie with a more privacy-forward alternative.

FLoC

Google proposes replacing third-party cookies with its own technology, FLoC (Federated Learning of Cohorts). Federated learning is a machine learning concept that allows many clients, browsers in this case, to collaborate to form a centralised model without exchanging actual data samples. This process groups users into different cohorts.

In essence, Google’s Chrome browser would be able to put its users into different groups, based on behavioural browser data. Each group has its own cohort ID, so that an individual’s personal identity is shielded. Advertisers can then use these groups for targeting or measurement. For retargeting, rather than target every individual ID that’s visited a website, advertisers would retarget all the groups who went to that site. Instead of retargeting a certain behaviour, you target the groups that correlate to that behaviour.

Google began testing FLoCs in March 2021, but has since announced it will be replacing them with a new Topics API for Chrome. Why? There were a lot of question marks hovering over FLoCs –– the main concern being that they are actually worse for privacy. This is because bad actors could use FLoCs and combine them with digital fingerprinting, considered by many to be the most invasive form of tracking, where a browser gathers enough information about a user to create a unique digital fingerprint –– making each person actually more identifiable. Prior to Google's announcement, browsers such as DuckDuckGo and Brave had already disabled FLoC from their browsers, while others like Apple and Mozilla stated they never had any plans to implement it. Amazon had also disabled FLoC from use on its websites.

Unified ID 2.0 (UID 2.0)

Spearheaded by The Trade Desk as a non-commercial, open-source initiative to replace third-party cookies, Unified ID 2.0 offers cross-site tracking based on a user signing in with an email address. Essentially, a user signs in one time to a publisher, using their email and offers their consent to using UID 2.0. They are effectively informed of the value exchange of the free internet – free content is delivered thanks to advertising, and advertising needs data to run efficiently. That consent would then be valid across the UID2 network, meaning a user does not have to consent multiple times. The email is then hashed and encrypted – essentially anonymised, and is mapped to the browser page. The identifier regularly regenerates itself, ensuring security, according to The Trade Desk, meaning the danger of digital fingerprinting is not present.

But UID2 has its own problems. One question is how many publishers will use it –– the New York Times for instance, has announced it will not take part. Most importantly though, is the fact that UID 2.0 relies on consumer consent and an opt-in mechanism. While that can be considered good for privacy and consumer choice, there is currently no telling how many users will choose to opt-in. Apple’s change in its IDFA offers an interesting parallel. In its iOS 14.5 update, Apple began to allow its users to opt-in to being identified for advertising purposes –– as of August 2021, that opt-in rate stands at 13%, according to Flurry. Many analysts believe that does not bode well for UID 2.0.

What does all of this mean for brands and agencies?

While alternatives to third-party cookies will continue to emerge and be tested, advertisers and brands cannot simply sit and wait for alternatives.

In this cookieless landscape, several things will be critical. First of all –– gathering as much consented first-party data as possible is vital. But more importantly, in order to deliver relevant advertising to new and potential customers, you need to build a robust, thorough understanding of your audience. Privacy-first insight gathering is crucial.

At Fifty, we are building the future of digital advertising, not through alternative tracking technology, but next-generation audience insights that puts people and privacy first. Over the last five years, we have developed a means of identifying naturally-occurring communities that exist on the web, what we call Tribes. These are groups of people clustered together, based on social connection, shared interests and passions.

We then use FiftyAurora, our supercharged contextual engine, to reach these audiences without tracking them. We analyse each tribe’s granular behaviours, shared interests and content consumption through our entity classification engine to reveal unique topics, affinities and demographics. We use this granular and organic audience understanding to build a map of where these users can be reached uniquely. This data also allows us to create subsets of a desired audience and understand the nuances of their interest – tailoring messaging to an audience drives far higher engagements than treating them as a single group.

By analysing audience affinities we can understand who might be reading a page, not just what topic it is about, without needing identifiers or cookies. This approach means we can serve campaigns based on the breadth of a Tribe’s interests rather than individual keyword targeting, making FiftyAurora a more efficient and effective solution.

The Cookieless Future

We at Fifty are not in the business of predicting the future. With Google pushing off its deprecation of third-party cookies until 2023, and Topics API and UID 2.0 still undergoing testing, it is hard to guess what the future digital advertising landscape will look like. What is clear is that consumers’ privacy concerns are not going anywhere, and neither is their thirst for unobtrusive, relevant advertising. The free internet is based on an implicit exchange between users and publishers, where data is discreetly exchanged for free information. That implicit exchange will continue to become a lot more explicit in the coming years. Whether or not consumers shy away from opting in from being identified for advertising purposes, the need to understand your audience’s interests and motivations will remain. For brands and advertisers looking to survive a cookieless future, human understanding and privacy-first solutions will be paramount.

To discover how Fifty can deliver audience intelligence and effective, privacy-first advertising, get in touch with our Sales Team or book a free demo today.