Giving a cookie-free future some context

As signal loss continues to make targeting more challenging, and marketers look beyond the cookie, one area of focus for the industry is contextual targeting, which places ads close to highly relevant content. The goal is to find a placement that resonates with the brand and engages the user.

Since contextual targeting focuses on content or keywords instead of users, marketers often assume that this is an effective, privacy-compliant, cookie-free solution. But it’s important to understand the full picture before you implement new strategies based on these assumptions. There are a few questions you can ask to guide you toward flexible, extensible, and high-performing programmatic media buying.

• Targeting: Can you find and serve messages to people you believe are in-market for your goods or services?
• Measurement: Can you receive insights on users who are exposed to, engage with, or ignore your ads? How do you attribute this to sales?
• Optimization: Can you adjust targeting, media mix, placement, context, or messaging to help increase the likelihood of positive outcomes based on campaign activity?

Contextual targeting does not require cookies. But without some kind of identifier, measurement and optimization become near impossible, making it hard to answer some of advertising’s more challenging questions:

• Cross-site tracking: Did a click lead to a purchase on the advertiser’s site?
• Customer journey tracking: How many ads does a user need to see/click for a purchase to happen?
• Frequency capping: Without identifiers, advertisers are only targeting content. How do you know if the user is seeing your ad?
• Attribution/optimization: What keywords are driving outcomes for which segments of your audience and what is wasted spend?

The good news is there are many solutions that can be combined with contextual targeting to help fill in these gaps and answer these questions. And they don’t require a lot of extra work to implement.

• Durable IDs: IDs like Unified ID 2.0 (UID2) are tied to persistent identifiers — like phone number and email — and are a good tool to test in cookie-free browser environments like Safari and Firefox. This will help you understand the scale of your first-party data, what you’re able to measure, and whether your optimizations work.
• Contextual benchmarking with cookies: We recommend using all available identifiers for your campaigns, including where cookies are still available. Ahead of cookie deprecation, this gives you an opportunity to develop benchmarks. Cookies can give you a view into performance of contextual campaigns, helping you answer questions like:
  
  
  • What do my audience’s online behaviors looks like?
  • How big are the gaps between my cookie-free and cookie audiences?

• Probabilistic IDs: These tools are common in Europe but are increasingly becoming table stakes in North America and Asia-Pacific as well. These IDs help close the gap between authenticated and ID-less traffic. Here are a couple of questions to consider:
  
  
  • How big is your gap? And what do you need to close the gap?
  • How can you balance the confidence that you’re reaching your target audience with scaling your campaign?

Here’s how you can take advantage of contextual targeting on our platform today:

• Remember that you still want to be able to frequency cap and measure outcomes with a cookie-free solution like contextual.
  
  
  • Target all trackable users and enable Identity Alliance. This will ensure the most accurate frequency capping and conversion measurement. Without these, you will bid on nontrackable users, which means ads served above and beyond optimal frequency spend is wasted and user experience degrades.
  • If desired, you could layer on a UID2-only audience. This will test what scale might look like in a cookie-free world.

• When cookies go away in 2024, emerging probabilistic solutions will help get you more scale.
  
  
  • Be specific with targeting. Select your contextual targeting keywords intentionally.
  • Use negative targeting. Try targeting keywords you do not want to show up against for a powerful and scalable application.
  • Layer on brand safety. When layering brand safety onto your campaign, you can ensure you’re not exposed near content that is not favorable to your brand.

• Generate additional insights.
  
  
  • Look at broader KPIs like incrementality. For example, enable iSpot if the client is running linear television, so you can measure the incrementality of your linear TV buy.
  • Add Inferred Brand Intent pixels to measure those conversions.
  • Look to emerging data sources and measurement providers. For example, you can use attention metrics and TV Quality Index to understand how content and context work together to enhance buys.

While some marketers are turning to contextual targeting to solve for signal loss, it can result in some gaps in your targeting and your measurement. It's important to consider contextual targeting alongside strategies that also use identifiers like UID2 to ensure your targeting is as well rounded as possible.

Want to make sure you're deploying robust targeting strategies on The Trade Desk, including contextual targeting? Reach out to your account manager today.