ID bridging emerged as a controversial solution in digital advertising. The technique aims to address deprecated third-party cookies with deterministic and probabilistic methods to connect user identities. Yet, it raises significant concerns about transparency, privacy, and potential fraud.
Unfortunately, the buy side and sell sides are at odds again – what else is new in ad tech? The buy side called out publishers and their tech partners for using deceptive practices to identify audiences. The practice in question is a technique called ID bridging. ID bridging has become a contentious issue as digital advertising grapples with the deprecation of third-party cookies in Chrome.
As Paul Bannister, CSO and co-founder of Raptive said, “The rise of ID Bridging over the last year is almost directly correlated to cookies going away. This technology could have appeared five years ago, but there wasn’t a pressing need. Now, with 3PC going away, buyers and sellers are looking for more ways to reach addressable audiences and bridging definitely can work for that.”
While many on the sell side rely on these techniques to monetize audiences that would otherwise be inaccessible, some warn that DSPs should be aware of and prepared for such methods because of the potential lack of transparency.
Despite ongoing discussions within industry bodies like the IAB Tech Lab, transparency around ID bridging practices remains a significant concern. The legitimacy of these methods varies widely, with some using ethical approaches while others border on deceptive practices.
As discrepancies between bid requests and actual ad delivery become more apparent, DSPs find tracking conversions and managing ad frequency increasingly difficult. Some platforms, such as Quantcast, had detected these issues, but many others only became aware through recent industry conversations.
ID Bridging: Deterministic and Probabilistic Matching
ID bridging allows publishers to package and target segments of their audience in a privacy-conscious manner, making their inventory more attractive to advertisers even in a cookieless environment. Essentially, ID bridging connects the dots between different user identities without relying entirely on third-party cookies. By leveraging first-party data, such as email addresses obtained with user consent or login information, publishers can create a valuable dataset that serves as the foundation for their targeting strategies.
There are two primary methods of ID bridging: deterministic and probabilistic matching. Deterministic matching is the most accurate, relying on direct, persistent identifiers like a hashed email address that remains consistent across devices. This method requires users to log in on multiple browsers, ensuring higher accuracy.
On the other hand, probabilistic matching is more common in ID bridging. It involves using complex algorithms to analyze signals, such as IP address, device type, and browsing behavior. While it offers a wider scale, it is less precise than deterministic methods, relying on smart guesswork to link different browsing profiles to the same individual.
The Potential Publisher Benefit
First and foremost, ID Bridging can help publishers keep their inventory valuable without third-party cookies, allowing them to maintain addressable audience segments and preserve the value of their ad impressions. Additionally, some argue that by adopting ID bridging, publishers can attract top-tier advertisers who are increasingly hesitant about cookie-based targeting and are reassured by privacy-conscious solutions.
As Yang Han, CTO and cofounder of StackAdapt said, “If publishers can reliably indicate that it’s the same user across different devices, it’s a valuable signal. However, there must be consistency and standardization. It’s not useful for a DSP to know if it’s the same user within a single publisher; we need to identify the same user across multiple publishers.”
Han warns that to achieve this at scale, publishers need to use a Universal ID. A publisher can assign their user ID, even without cookies, and share it with the DSP. However, different publishers generate different IDs for the same user, creating a fragmented and sparse data pool. To make the data useful, a universal user ID across all publishers is necessary.
This approach also supposedly ensures compliance with regulations like GDPR and CCPA, demonstrating a commitment to respecting user choices while safeguarding your business. Moreover, some ID bridging solutions open access to unique demand pools, potentially expanding publisher revenue opportunities beyond traditional cookie-based advertising.
The Not So Great Side of ID Bridging
After a year+ of talking to dozens of companies about ID bridging, I can confidently say that all ID bridging conversations are a series of miscommunications and misunderstandings.
— Paul Bannister (@pbannist) June 5, 2024
Not all that glitters is gold, and the same is true for ID Bridging. One of the primary concerns with ID Bridging is its potential to exacerbate the digital divide that already exists between the sell and buy sides. Advertisers who have long relied on third-party cookies may find the transition to ID bridging daunting and resource-intensive. These advertisers might be skeptical of new solutions, perceiving them as risky or unproven. The shift requires a significant change in infrastructure and a rethinking of strategies that have been developed and optimized over the years.
The reliance on ID bridging demands robust first-party data, which can be challenging for smaller publishers or those who have not built strong direct user relationships. This transition phase can cause friction, with some advertisers potentially experiencing a decline in campaign performance during the adaptation period.
Moreover, privacy concerns remain a significant issue. While ID bridging aims to enhance transparency and compliance with data protection regulations, the method of using hashed or anonymized identifiers might raise alarms among privacy advocates and users. The challenge is ensuring that these measures are sufficiently robust to protect user privacy without compromising the effectiveness of targeted advertising.
In addition, Bannister warns about the ease of taking advantage of ID Bridging. Shady publishers or ad tech firms can bridge IDs that don’t represent the user, so buyers waste their budgets. He adds, “Even for cases where the buyer has consented to the use of Bridging, it can be challenging to ensure that it is being done correctly. ID Bridging can be a good thing but has to be done responsibly.”
The scale and reach of ID bridging solutions also have limitations. These solutions generally have a smaller reach compared to cookie-based systems. This reduced reach can limit advertisers’ ability to deliver personalized ads at scale, potentially impacting campaign outcomes.
There is still plenty of headway before the industry reaches a consensus on ID Bridging. But this is a vast industry with many intermediaries in between, and there’s a chance the industry may never agree. Yet, if the conversations around ID Bridging are, as Bannister characterized them, “A series of miscommunications and misunderstandings,” then we won’t get anywhere.