Table of Contents
Behavioral targeting in marketing and advertising has been hit by some big changes in the past few years, namely third-party signal loss and cookie deprecation. These structural shifts have especially affected health advertisers who target specific older audiences.
At Digital Brothers, we help clients reach durable, over-50, health-motivated audiences in the wake of these changes. Our approach: use declared and contextual signals instead of cookies. This has pushed our focus back to signals that are actually reliable, which, ironically, has made our targeting better in this vertical.
Here’s what we’ve learned, and how you can apply these tactics to build a strong program that can withstand changes in a post-cookie landscape.
Strategy 1: Target a Declared, Durable Audience
We recommended that one particular client, a direct-to-consumer hydration and wellness brand, build on data that doesn’t decay when cookies disappear. For older health audiences, this makes a lot of sense: These audiences were never perfectly captured by behavioral cookie trails anyway, and a lot of the inferred data about them was noisy.
Why Inferred Demographic Data Fails for a 50+ Health Buyer
Inferred or guessed demographic data is often wrong for a precise audience like this. The budget you spend on that audience is optimizing against a blurry picture of who you’re actually targeting. Instead, we recommend leaning on two aspects that actually hold up: the context of the page someone is reading, and declared or first-party audience data signals.
For an older buyer searching for solutions, the article they’re reading right now tells you more about their intent than a fragmented cookie history ever did. It’s a more durable intent signal that will also withstand any new privacy changes. Context and declared data is a more honest signal in the first place for this audience.
Declared Demographics Plus Publisher First-Party Interest Data
Declared vs. inferred demographics information can make a big difference in marketing to older health audiences. Declared data is someone telling you who they are, while inferred data is a model guessing who the audience is. Inferred behavioral data is probabilistic, looking at which pages a device visited to guess at what kind of person the user is. That data degrades fast and gets stale, and from the start it’s fuzzier than dashboards might imply.
Declared data is closer to the truth: it’s stated, it’s durable, and it doesn’t evaporate with the next browser update. We still use modeled audiences at Digital Brothers, such as for conversion-seeded prospecting, but it’s essential to seed it with real, declared conversion events to ensure that the predictions will be spot-on. If you build your whole strategy on inferred third-party data trails, you’re building on sand that’s actively washing away.
For this hydration and wellness brand, we recommended using publisher-sourced declared demographic data and interested segments gathered from Taboola’s more than 9,000 publisher integrations and Connexity e-commerce site. That allowed the brand to reach the right readers with more confidence.
Strategy 2: Place Ads in the Right Content Context
Where the ad appears can qualify the audience when cookies are no longer available. It’s possible to do this safely and effectively with modern advertising platforms.
While behavioral targeting tries to reconstruct audience interest from a trail of past activity, it becomes less accurate and fragile over time. Contextual targeting doesn’t replace everything (we still seed conversion-based audiences for prospecting), but it can be ideal for reaching a problem-aware health buyer at the right moment. Contextual targeting also sidesteps privacy exposure issues that make behavioral targeting more and more difficult. It’s a durable piece of information you can trust.
Contextual/NLP Placement in Health Editorial
We also suggested that this health brand use natural language processing (NLP) contextual intelligence across code-on-page publishers. This reaches readers who are already engaged with relevant health content. If someone is reading in-depth content about a health problem, it’s intuitive that the context is the intent — you don’t need a behavioral profile to know they care about the topic. Aligning your ads with the content they’re actively reading is a strategy that punches above its weight in terms of return.
Staying Cookieless and Brand-Safe
Contextual advertising is also privacy-durable and brand-suitable. It reaches people, including private or incognito browsers, without requiring PII or third-party cookies. So, contextual advertising can be a powerful tactic for health verticals in particular.
Key Takeaways
The advertisers who prepared for these changes are going through the transition smoothly, and even clarifying their audience targeting. Those who relied entirely on third-party cookies are scrambling. For an older, health-motivated audience, it makes more sense to rely on durable first-party and contextual signals than decaying, cookie-based data. It also future-proofs your strategy as signal loss accelerates; you can start owning your signal instead of renting your audience data. Build your first-party data now by making sure pixel and conversion events are clean, capturing declared data wherever you legitimately can, then treating that information as the asset it is.
Frequently Asked Questions (FAQs)
What are Taboola 1P Audiences in Realize, and where does the data come from?
Taboola First-Party (1P) Audiences are more than 500 first-party segments across over 200 categories, such as declared demographics, interests, and intent. They’re built from Taboola’s exclusive integrations with more than 9,000 publisher sites, plus Connexity, Taboola’s e-commerce data business. You’ll find them labeled “Taboola 1P Audiences” in the Marketplace Audiences section of Realize, and you can mix them with third-party segments. Because they’re publisher-sourced and declared, rather than inferred from third-party cookies, they stay durable as signal loss accelerates.
How do declared demographics in Realize differ from inferred demographics?
Inferred demographics are guessed from behavioral signals and will continually decay as cookies disappear. Declared demographics (such as age and gender) are self-reported and sourced through Taboola’s publisher network. In Realize, the guidance is to use the declared segments when targeting age and gender for precision. If you layer segments with AND targeting, keep it to about five maximum, so you don’t go too narrow and lose scale.
How does Contextual Targeting work in Realize, and why is it cookieless?
Realize’s Contextual Targeting capability uses AI and NLP to read the content and metadata of pages across more than 9,000 directly integrated publishers. It then places your ad alongside relevant content, rather than tracking the individual. Because it targets the page context, it needs no PII or third-party cookies and can reach users even in incognito/private browsing mode, which also makes it brand-suitable for sensitive verticals like health. A best practice is to combine contextual targeting with demographic/behavioral signals and align creative to the content context. For niche or trending needs, Topics Targeting extends this to more than 70,000 granular topics.
