When Google announced it would deprecate third-party cookies, the digital advertising world was thrown into disarray - and it’s been a wild ride since then. But what has really happened, what workable solutions have emerged, and where are we now? Taking our usual approach to big industry issues, we unpack the details, look at the options available to all parties, and make some practical suggestions on the path ahead.
On 14th January 2020, Google announced via its blog that third-party cookies in its Chrome browser were to be phased out and replaced by its Privacy Sandbox. It gave itself two years for the task.
The post prompted shock, alarm and intense debate across every facet of digital advertising as the industry faced up to the demise of this stalwart of the targeting toolbox - although the more pragmatic voices welcomed the focus on privacy and believed it would encourage innovation as alternatives were sought.
But the course of cookie deprecation never did run smooth and in the five plus years since the original declaration, there have been many twists and turns. Google extended its own deadline several times as completion of its Privacy Sandbox was delayed. These postponements were reinforced by industry concerns, regulatory intervention, and testing that did not fill potential users with confidence about its efficacy.
Following a major U-turn in July 2024, on 22 April 2025, Google finally confirmed that it would keep third-party cookies ‘for the foreseeable future’.
The collective sigh from the advertising world denoted relief and frustration in equal measures.
It’s far from a case of ‘as you were’. First up, Safari and Firefox both block third-party cookies by default and, while Chrome commands over 70% of the browser market, their stance keeps a spotlight on privacy. Simultaneously, more and more regulators are pro privacy protection, with legislation requiring explicit consumer consent for the collection and use of their personal data only set to increase in the foreseeable future.
Perhaps even more significantly, five years is a long time in adtech. Mindsets have shifted - and, fueled by what for some time was impending necessity, a lot of work has already taken place to find viable alternatives to the third-party cookie by the time Google reversed its decision.
Alternative identifiers (alt-IDs) take the form of browser-based technology that mirror the role of third-party cookie in a privacy-safe way. By April 2025, these were appearing in 35-40% of open auction traffic and 65% of cookieless requests.
Even more recently, ID5’s 2025 State of Digital Identity report, which surveyed advertisers, publishers and technology vendors, found that 91% of players have already adopted alternatives to third-party cookies or are planning to test them.
The direction of travel seems clear: despite third-party cookies still being available (in Chrome), alternative identity is rapidly becoming the new standard.
So what does this evolving alt-ID landscape currently look like?
As we know, third-party cookies are still on the menu in Chrome. Advertisers can continue to track users across multiple websites and use this information to create targeted ads based on the preferences and interests indicated by peoples’ online behavior. They can also retarget consumers with the aim of upping conversion rates, thanks to having access to their browser history, while cross-site analytics give insight into the broader customer journey and offer the opportunity for marketing strategies to be refined.
However, these activities take place against the backdrop of an increasingly privacy-conscious world, in which regulatory challenges, data security and lack of transparency are continual concerns - not to mention the possibility of another about-turn by Google.
The alt-IDs developed fall broadly into two camps: deterministic and probabilistic.
Based on personally identifiable information (PII), with email addresses (for example) being hashed to protect privacy, deterministic alt-IDs are underpinned by user consent and stated preferences. Requiring first-party audience registration, scale is however a key issue. Current options include:
RampID. Widely adopted, LiveRamp’s alternative ID combines first-party data, third-party behavioral data, and offline PII to create user IDs. The ‘RampID’ refers to the resulting cross-device anonymous identifier assigned to each user.
Unified ID (UID) 2.0. This open-source, non-commercial identity solution is an industry initiative initially developed by the Trade Desk before transferred to Prebid. An identifier is created using an encrypted email (the value of which is different each time it enters the bidstream), but only when a user has provided consent via their email address. EIUD is the UK and EU version, designed to work within GDPR compliance.
netID. Operated by the European netID Foundation, netID is a GDPR compliant single sign-on (SSO) for European users, who can manage their consent and data-sharing preferences centrally. An addressable identifier is generated for users that give permission and managed based on the preferences they have specified.
Publisher first-party data, which is a vital element of the addressability equation in a post (third-party) cookie world, also enables deterministic approaches. Collected directly from users (with their permission), it allows publishers to link peoples’ online behaviors, interests and engagement signals. The benefits include rich audience segments, improved targeting precision and personalization and cross-channel activations.
Probabilistic IDs bypass first-party data, deploying instead data points such as IP address, device type, screen resolution, operating system, etc, to generate a ‘best guess’ as to a user’s identity. This approach enables more scale due to the data available, but there can be inconsistencies within this data and cross-device connection can be challenging. Solutions include:
ID5. The ID5 alternative ID uses signals including IP address, timestamps, user agents and other non-PII attributes to match uses across domains, browsers and devices. Where possible, it combines these probabilistic data points with deterministic signals to produce a more robust alt-ID. In 2025, the company introduced Adaptive Identity, its evolved solution designed to continuously learn from data, behavior, and changing signal environments so that ID matching becomes better over time.
Panorama ID. Lotame’s Panorama ID also takes this hybrid approach, combining probabilistic signals such as device, browser and platform data with deterministic data including hashed emails or authenticated logins when they are available. It leverages the company’s identity graph, which links various disparate identifiers and signals into a single (privacy-compliant) user view.
While not technically part of the addressability stack, clean rooms, which sit downstream of the solutions listed above, facilitate post-campaign measurement and analytics - critical components at any time, and all the more so as the industry adjusts to a new way of working. In these secure, privacy-compliant and neutral environments two or more parties can collaborate to match data without exposing any that is raw or personal. Providing transparency and flexible modelling, clean rooms enable relevant players to undertake exercises such as reach / frequency deduplication, attribution, conversion analysis, and incrementality studies so that activity can be continually optimized.
Since Google’s 2020 declaration, many ways to handle the demise of the cookie have been proffered, with some becoming more mainstream than others. In short, there is a lot of choice.
At BidSwitch, our global and neutral infrastructure means we are ideally placed to provide an overview of what is happening on a day-to-day basis and identify trends. From our vantage point of seeing more than 300 billion bid requests every day, we know that buyers are honing down their selection of alt-ID providers, with RampID and UID 2.0 leading when it comes to US retail and CTV-focused DSPs, for example.
Based on what we see in the bidstream, we are also able to consult with SSPs and encourage adoption of a certain provider. For example, if certain DSPs are targeting RampID makes sense for SSPs to also favor this alt-ID. To make our recommendations credible, it is vital that we adhere to our founding principle of remaining agnostic, while providing performance insights to prove effectiveness, and passing every ID end-to-end.
In reality - as is often the case in this industry - there is no one-size-fits-all answer when it comes to the ‘best’ alternative to the third-party cookie. It’s likely that a combination of the addressability tools discussed above will be needed and potentially more than one alt-ID (51% of publishers are working with one or more ID bridging partners). This makes it important to test, and keep testing, for the ID solutions that are right for each individual business.
Looking ahead the picture is still hazy. It seems likely that we will see consolidation, with around a half-dozen IDs surviving. At the same time publishers will slim down to two or three alt-IDs to avoid the excess data that leads to bidstream bloat.
As referenced previously, there is also the possibility that Google will revisit deprecating cookies in Chrome at some stage in the future.
For now, publishers and supply side platforms should audit their alt-ID footprints to ensure those they have adopted are serving their needs, and those of the DSPs targeting their inventory. Buyers and DSPs meanwhile can analyse the win-rates achieved the cookies and alt-IDs they deploy and tune their decisioning logic accordingly.
For both buy and sell side, BidSwitch can provide an ID health check as well as an alt-ID adoption roadmap. Contact your BidSwitch account team to get started.