I. Signal Fades, Strategy Shifts: The Future of Open Web Targeting
In the early days of digital advertising, marketers generally focused on simple, broad-reach tactics - placing banner ads on popular websites and hoping the right audiences would see them. As the internet matured, the industry shifted toward behavioral targeting, using third-party cookies and historical browsing data to infer user interests and expand their reach by connecting with specific audiences across a multitude of websites. By leveraging the information stored in cookies, brands could effectively engage users in their areas of interest, significantly broadening their advertising footprint. This approach proved more effective than one-size-fits-all media buys, but it also introduced privacy issues, added complexity, and user discomfort.
Today, amid mounting regulatory scrutiny and consumer pushback, the digital ecosystem is in the midst of a profound transformation. The question becomes urgent: How will advertisers adapt in a world where data drips instead of flows? Are we witnessing the end of an era - or the dawn of a smarter, more sustainable approach? And most importantly, is the new approach truly smarter? Let’s find out.
[Accuracy issues]
For years, third-party cookies have been the backbone of digital audience segmentation, providing marketers with various labels: from age and gender, through interest and behavior, to purchase intent. Yet, while some of those identifiers bring results, others can be highly inaccurate. Many brands suspect that the accuracy promised is more illusion than fact. We decided to put this suspicion to the test:
We targeted a set of standard audiences defined by classic cookie-based socio-demographic segments - such as young women, parents, residence owners/renters, married people, etc. - and then served them a multi-step survey. The survey asked each respondent to confirm being a part of these audiences, allowing us to compare how these “official” cookie segments measured up against reality. After collecting 1,323 responses, the results were troubling: only 18% of the supposed “women 18-24” audience matched their assigned demographic, and among the “parents” segment, a mere 34% actually had kids. In fact, 35% of those labeled as “young women” declared to be 55 or older, while 52% of the “moms” segment claimed to be men!
This level of inaccuracy challenges the traditional reliance on cookie-based targeting, raising uncomfortable questions about how much advertisers should trust segments built on third-party data. If these foundational identifiers are crumbling, what does that mean for the rest of our targeting and measurement strategies?
[We’re in a privacy-first era of marketing]
The shift from cookie-based targeting isn’t happening in a vacuum. On one hand, regulations like GDPR and CCPA have heightened consumers’ awareness of data use, while browsers like Safari and Firefox have long blocked third-party cookies. With Chrome soon following suit, many behavioral and retargeting strategies are increasingly difficult to execute. On the other hand, consumers have grown wary of ads that seem to “follow” them across the web. In this new landscape, marketers must find methods that deliver results without encroaching on consumer privacy. Here we face a new hypothesis: as personal data pipelines dry up, can approaches that don’t rely on intrusive identifiers step in and provide both efficiency and reassurance?
[New directions]
Despite 92% of UK marketers expecting no further delays to the end of third-party cookies, 73% admit they’re not fully prepared to operate without them [7]. Many are still developing first-party data strategies, concerned about identity solutions, and planning to spend more on the open web - underscoring the urgency of exploring new targeting and attribution methods [5]. As brands and publishers confront this “signal loss” era they’re weighing different paths forward. Two primary solutions have emerged: alternate ID frameworks and ID-less approaches. Think of it as a crossroads: one route tries to patch the holes in the old system with alternate IDs, while the other path embraces a radically different model that abandons user-level data entirely.
Alternate IDs aim to preserve user-level addressability through privacy-compliant identifiers, such as hashed (anonymised) emails, device IDs, or universal IDs from neutral parties. While these methods uphold a form of personalized targeting, they’re still anchored in user-specific data. Major backers include The Trade Desk (Unified ID 2.0) and LiveRamp (RampID).
Is it a viable fix, or just delaying the inevitable? Let’s weigh the pros and cons:
Advantages:
Challenges:
ID-less approaches, on the other hand, avoid user-level data entirely, relying instead on the context or environment to guide ad delivery. Beyond contextual targeting - focusing on the nature of the content itself - ID-less methods can tap into cohort targeting, geolocation, or timing signals to deliver relevant ads without revealing who the user is. Supporters of this model include Google’s Privacy Sandbox (Topics API), Apple’s SKAdNetwork & ATT, Oracle’s Contextual Intelligence, GumGum’s Verity, and Adlook’s suite of ID-less solutions like Deep Context, Deep Search, Deep Survey, and ContentGPT.
Does this signal a paradigm shift, where content and context become the new currency of ad relevance, dethroning user-based profiling?
Advantages:
Challenges:
While alternate ID frameworks try to reconstruct traditional targeting models through new identifiers, ID-less approaches embrace a world without personal signals. Instead, they tap into what’s happening in the moment - on the page, in the environment, or at a given time - to determine which ads make sense. By eliminating unique identifiers, these strategies can function seamlessly across platforms and publishers without eroding trust.
Let’s deep-dive into one particularly promising ID-less strategy poised to transcend these challenges: contextual targeting. Can it rise above the chaos and claim its place as the new standard in a privacy-first era?
II. The Contextual Renaissance: Redefining relevance without personal data
As the value of third-party cookies diminishes, contextual targeting - an approach that predates advanced data-tracking methods - is enjoying a renaissance on the open web. According to IAB Europe and Xandr, nearly three-quarters (74%) of European advertisers plan to leverage contextual targeting once third-party cookies and device identifiers are gone [3]. Is this just nostalgia for simpler times, or could contextual targeting offer a genuinely superior way to match messages with moments?
[Definition]
At its core, contextual targeting aligns advertisements with the content of a given webpage or app environment rather than with a specific user’s profile. Instead of deducing who the user is or what their past behavior might imply, it pinpoints where they are now - showing, for example, an SUV ad on a page about road trip destinations. Here, relevance flows naturally from the environment, enhancing brand impact, and capturing attention at precisely the right moment.
[Relevance backed by consumers - Consumers demand better user experience]
Today’s users, fatigued by relentless tracking and irrelevant ads, welcome contextually relevant messaging. According to WARC, 72% of consumers consider contextual relevance an essential factor in the ads they encounter. In the UK, 81% prefer online ads that complement the content they’re viewing, and 65% feel more favorably toward brands that get the context right [1]. By tapping into the moment and the meaning of the page, advertisers can forge stronger emotional connections, meeting audiences on their own terms.
[Brand suitability solution]
Contextual targeting isn’t just about respecting privacy or serving more personalized ads - it’s also a powerful safeguard for brand integrity. Traditional audience-based methods can place ads alongside content that’s inappropriate or off-brand. By contrast, contextual targeting ensures that ads appear in brand-safe environments. On the open web’s vast and varied landscape, this capability is invaluable. Advertisers can achieve scale without compromising contextual fit, reducing reputational risk, and bolstering consumer trust.
[New tech = New level of contextual]
This contextual renaissance is not just a nostalgic return to older methods; it’s fueled by powerful machine learning, natural language processing (NLP), and cognitive technologies that can interpret webpage content at scale. Leading industry players have developed sophisticated algorithms to battle the challenge of ad relevance and targeting precision. These tools understand language nuances, detect sentiment, and identify brand-suitable environments.
Advertisers can now target not only simple keywords but also entire concepts and categories, ensuring that the message fits the contextual frame and tone of the content. The result: richer, more meaningful ad placements that feel less intrusive and more like a natural extension of the user’s browsing experience.
As we move deeper into the capabilities, case studies, and performance metrics of contextual targeting, we’ll assess how effectively it can stand alongside - and potentially surpass - other targeting strategies. We’ll examine where it excels, where it may face limitations, and how this new prominence is reshaping a more privacy-safe, brand-friendly, and user-centric programmatic landscape.
III. The Central Assumption: Where Contextual Shines - and Where It Doesn’t
The growing body of evidence supporting contextual targeting underscores its potential to drive stronger engagement and performance. Campaigns that employ contextual strategies often report higher click-through rates (CTR), improved brand lift, and enhanced attention metrics. As the advertising industry prepares for a post-cookie ecosystem, the central assumption that contextual targeting can replace cookie-based targeting is being tested in real time. Early results and market momentum suggest that contextual is not only a viable successor but may become the preferred standard.
[Case studies]
As noted by Analytic Partners in their ‘ROI Genome’ project, when done well, contextual targeting can be anywhere from 1.2 to 2.5 times more effective than audience-based or person-level targeting - a substantial performance leap attributed to advanced AI and cognitive technologies that now interpret page sentiment and nuanced language at scale [2].
This effectiveness also translates into attention and brand favorability: a review of 76 campaigns across eight verticals showed that contextually targeted solutions consistently delivered nearly 70% more attention for skippable ads, surpassing industry benchmarks 95% of the time for skippable formats and 97% of the time for non-skippable formats [4]. Such results suggest that contextual relevance encourages users to spend more time with the ad, increases message retention, and often leads to stronger purchase intent.
If contextual targeting operates on the principle that a consumer’s current environment is the most reliable signal for relevance, then our central assumption becomes straightforward: Contextual targeting is a stable source of quality audiences, enabling marketers to effectively deliver relevant ads that resonate with users’ immediate interest and intent.
As third-party cookies dwindle, this seems increasingly credible. We decided to challenge this assumption and check whether context actually plays such a big role in reaching the desired audience.
[Adlook’s internal research]
To evaluate the effectiveness of contextual targeting, we conducted an internal study surveying 3,600 users across diverse verticals, including Automotive, Electronics, Pets, Healthy Living, Travel, and Finance. Our test groups consisted of user profiles targeted through specific contextual categories, while the control group received no targeting - simulating a common scenario of a broad campaign or one attempting to target specific demographics using third-party cookies, which, as previously illustrated, resulted in near-random precision. This scenario is particularly relevant for CPG brands whose primary goal is to build brand awareness and reach wider audiences rather than targeting highly-specific niche groups.
In addition to contextual, we tested other ID-less solutions, including Google’s Topics API and our proprietary content-based targeting tool, ContentGPT, to reach the same relevant audiences.
Participants completed a multi-step survey where they first indicated their interest in the presented categories and their purchase intent, followed by questions about their preferences and demographics.
Our main goal was to examine the lift in interest and purchase intent between contextual targeting and the broad, no-context control group. Subsequently, we compared the performance of the alternative ID-less methods to determine how other solutions fare in reaching audiences with genuine interest and purchase intent.
[Results]
Our research revealed that contextual targeting significantly outperformed the broad, non-contextual placements in driving genuine user interest within high-involvement categories like Automotive, Travel, Electronics, and Pets - the lift in interest (users who answered “A lot”) varied between 16% and 82%.
Specifically in Automotive and Travel categories, contextual targeting also achieved a 136% and 38-99% lift in purchase intent (such as planning to buy a car or planning a trip), compared to the control group.
When looking into other ID-less solutions we found that contextual targeting surpassed Google’s Topics API in reaching people with enhanced purchase intent, demonstrating its superior ability to engage users effectively.
Our proprietary content-based targeting solution, ContentGPT, delivered even greater results, driving 2x and 3x higher purchase intent than contextual targeting and Topics API accordingly, underscoring its advanced precision and capability to connect with users with strong involvement.
[Conclusions]
Key insights from our survey:
So, where contextual shines - and where it doesn’t
The advantage of contextual targeting is more pronounced across several scenarios:
Despite its strengths, contextual targeting isn’t a universal remedy. In some scenarios, it only slightly outperforms broad placements, especially when user attention isn’t closely tied to the content. Situations where contextual targeting might fall short:
[Speaking of advanced technology…]
Scaling campaigns to achieve more complex brand goals and reach niche audiences requires sophisticated signal processing and deep semantic analysis, posing significant challenges for many players in Programmatic advertising. This is why we firmly believe that advanced technology is pivotal: as consumers’ online activities become increasingly complex and diverse, advancements in AI enable brands to quickly process billions of data points across multiple languages, networks, and countries. Instead of merely reacting, advertisers can swiftly identify ideal ad placements and even anticipate future consumer patterns, ensuring campaigns are effective and forward-thinking.[8]
At Adlook, we have harnessed Deep Learning to breathe new life into contextual targeting, boosting efficiency and scale. Our AI-powered Natural Language Processing (NLP) algorithms analyze specific signals - such as a website’s URL, content category, text, and images - to precisely match ads with relevant audiences based on the content they are engaging with. The level of autonomy and self-learning possible with Deep Learning means the algorithm will continue to improve its targeting capabilities over time, making it a primary, long-term solution [6]. Discover more about our suite of ID-less targeting solutions here: Adlook Cookieless Solutions
Summary
As we transition into the post-cookie era, the digital advertising landscape demands a fundamental shift in targeting strategies. Contextual targeting emerges as a promising solution, enabling brands to reach their target audiences by focusing on the environments where users are actively engaged, rather than tracking their every move. This shift not only respects user privacy but also enhances ad relevance by aligning messages with the surrounding content.
While contextual targeting offers significant advantages, it also presents certain challenges. However, advancements in AI and signal processing are elevating its effectiveness to unprecedented levels, enabling scalable ad placements even among more sophisticated audiences. In many instances, the integration of advanced technologies enables contextual targeting to outperform non-contextual approaches and alternative ID-less solutions like Google’s Topics API, especially within high-involvement categories. This balance effectively marries performance with privacy.
There is no time to wait for the complete phase-out of third-party cookies. Brands must proactively adopt and refine alternative solutions, embracing advanced technologies that facilitate effective audience reach. By implementing these strategies now, brands can develop robust and impactful advertising campaigns that connect meaningfully with their audiences while maintaining the highest standards of privacy and trust.
Sources:
[1] https://www.warc.com/content/article/bestprac/what-we-know-about-moment-marketing/111416
[2] https://www.warc.com/content/article/bestprac/what-we-know-about-targeting-vs-reach/117240
[3] https://dl.xandr.com/2021/06/Shaping-the-Future-of-Identity-Guide.pdf
[7] https://www.warc.com/content/article/warc-exclusive/the-future-of-programmatic-2024/156796
