You can read hundreds of think pieces on new approaches to targeting in a cookieless world. I’m here to tell you they’re all wrong.
The most viable, sustainable and effective alternative isn’t new: Contextual targeting is the only proven approach to targeting that fundamentally embraces the privacy-centric paradigm.
I know your objections. But it’s time advertisers rethink outdated conceptions about contextual’s limitations. AI and machine learning have transformed what’s possible with modern contextual targeting, allowing brands to precisely target their unique audiences.
Allow me to explain.
Advertisers need to look beyond unsustainable work-arounds
By now it’s apparent to most advertisers that putting more money into already-saturated walled gardens isn’t a smart option.
Brands need to be on the open web. But they’re right to be nervous about unproven unified IDs. The fragmented identity landscape delivers frustratingly limited reach.
I also understand why organizations might question the buzz around first-party data: Most companies would need to make massive investments in the infrastructure and tools needed to collect and activate first-party data – and the entire approach falls apart the second users log out of systems and applications.
The bigger problem, though, is that these approaches are unsustainable work-arounds. If you’re still tracking users around the web, I think we all know that, at some point, regulation (or public opinion) will catch up with you.
It’s no surprise, then, that we’re seeing a major resurgence in contextual targeting. But here’s the thing: Contextual should not be viewed as a last-ditch alternative. Advertisers need to understand how the new generation of contextual ad tech is erasing traditional questions around the effectiveness of contextual targeting.
Contextual 2.0 is here: AI + ML redefine capabilities
Over the last decade, innovation moved contextual targeting from basic keyword inclusion/exclusion to a much deeper and more accurate understanding of digital content.
Natural language processing (NLP) models provide incredibly accurate sentiment analysis, going beyond keywords to understand the true meaning of the content. For example, NLP tools know that “attack” has a negative connotation in an article about terrorism or crime – but “attack” can be used in a positive way in a sports article.
Computer vision technology accurately analyzes image and video content that dominates digital content today. And speech-to-text engines do contextual analysis of audio.
Together, these tools dramatically improve the effectiveness of contextual targeting. They’re also a major boon to brand safety, not only ensuring ads don’t appear alongside inappropriate content but creating the new concept of brand suitability – aligning creative with the tone and sentiment of content.
Get ready for Contextual 3.0: adding scale and improving precision
As advertisers turn their attention and budgets back to contextual, they’re riding a new wave of innovation.
Rather than just analyzing specific URLs within certain predefined categories, leading contextual technologies, such as Network Level Analysis (NLA), look at the entire universe of URLs. NLA develops an understanding of the network as a whole to understand content clusters and topics that audiences are engaging with at that moment.
Advertisers can impact their unique audiences by targeting the context that best fits their campaign brief, without the loss of accuracy caused by the standardization process. For example, they can go deep into the concept of sustainability by aligning their creative with the most timely, salient topics for an audience.
This full-scale breadth leads to a major disruption: Using advanced machine learning models to do custom targeting based on a specific campaign briefing. Let’s say a brand wants to align itself with “happy moments for the holiday season.” The traditional contextual approach limits advertisers to predefined content categories. You might target the broad “lifestyle” category and then refine using keywords – but you’re invariably missing relevant audiences and targeting others that aren’t relevant.
These accuracy flaws grow larger with scale: When you want to increase reach, you expand the size of the predefined category. This nets you a few relevant targets but even more irrelevant ones – and those “misses” get further off-target.
The new approach takes a brand’s brief and scans the whole network of URLs to find the articles that are semantically closest to the brief: “happy moments for the holiday season.” Your campaign is focused on the true center of your relevant audience. And as you scale to increase reach, you’re expanding the size of the customized category – but still staying centered. It’s a unique exception to the rule: You’re scaling while keeping a very high level of precision.
Three tips for choosing a contextual ad tech partner
Over the next 18 months, I predict contextual will move from niche tactic to essential strategy. For brands to separate the true contextual innovators from the me-too players, I recommend they assess players in these three ways.
- Test your options head-to-head
The best way to cut through the marketing BS is to test head-to-head. The differences between a company that’s built around contextual and one that’s jumping on the bandwagon will be clear.
- Look for a full-stack contextual solution
The real magic of modern, AI-powered contextual targeting comes when you integrate control of all three levers: media, creative and targeting. Brands should seek out full-stack solutions that provide the inventory, deliver consumer-first creatives and the most advanced contextual targeting capabilities that can create campaign-specific machine learning models tailored to the specific opportunities.
- Don’t limit contextual to branding
Contextual hasn’t traditionally been used in performance advertising. But I’m seeing companies investing big budgets to figure out how they can be competitive using contextual for mid- and low-funnel KPIs.
Advertisers should look for contextual partners with the vision and confidence to apply their tool set for performance advertising, helping brands create an integrated and consistent customer journey.
Contextual will reshape paradigms in the digital ad world
I’ve always found demo-based targeting to be problematic. Defining a target audience beforehand is inherently limiting. Traditional assumptions and biases creep in. The process ignores customers that don’t fit the profile.
I see more brands who recognize that cookie deprecation gives them an opportunity to change how they target: Instead of finding people, brands are thinking about finding moments when an individual has signaled they are interested in becoming a customer based on what they’re reading, watching or doing right now.
Who you are is much less important than what you care about.
Armed with modern AI and machine learning, brands can finally bring this idea to life while genuinely respecting the concept of user privacy.