With the “death of the cookie” approaching, the entire digital advertising industry is looking for ways to retain and grow advertising spend in this new paradigm.
Today, the majority of digital advertising is centered around behavioral advertising: showing precisely-targeted ads to users based on data that indicates they are likely to be interested in them. For example, showing an ad for sneakers to someone who recently visited an ecommerce store browsing for shoes.
New privacy-preserving forms of advertising will focus less on tracking users and their interests, and more on other signals to show the best ads for different scenarios.
One of the methods for serving ads that is getting a lot of attention right now also happens to be one of the oldest forms of advertising: contextual advertising.
Contextual advertising is fairly straightforward and easy to understand, and it can still be very powerful when used correctly.
What is contextual advertising?
When you open a magazine, you expect to see ads that are related to the topic of the magazine. Home decor magazines tend to feature ads for a furniture store or paint brand. Video gaming magazines are more likely to have ads for video games or computer hardware.
At its simplest level, that is contextual advertising — ads with content that matches the environment the ads are in.
Contextual advertising can get even more in-depth by combining information about the environment with other data and assumptions.
Back to our magazine example — in a luxury travel magazine you’d expect to see ads from a cruise line, but it also makes intuitive sense to see ads from high-end car companies. People booking expensive trips are good candidates to buy a fancy car.
You’ll also see products that are associated with the content in other ways, like a sports magazine with ads for sneakers that are endorsed by athletes covered in the magazine.
On the internet, contextual targeting can get even more precise. If a reader is looking at a chicken parmesan recipe, they might see an ad for a brand of mozzarella cheese. If another reader is checking out a movie review, a streaming service might show an ad for a recently-released TV series of a similar genre.
Context can even be expanded into the ad itself, where it can influence the actual ad creative. Imagine looking at a recipe for a smoothie, where an appliance company shows an ad for a blender. The visual content for the ad itself can be customized to show blueberries in the blender if it’s a blueberry smoothie recipe, or strawberries if it’s a strawberry smoothie!
The benefits of contextual advertising
Privacy regulations like Europe’s General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA) restrict behavioral advertising because it uses information about the user to serve targeted ads. Since it uses information about the page the ad is appearing on instead of information about the person who is viewing the ad, contextual advertising is an attractive option as interest in online privacy increases.
Contextual advertising can also help advertisers maintain brand safety, since it’s all about where the ads will be displayed and what content will surround the ads. Behavioral advertising focuses more on reaching the desired user, which runs the risk that advertisers may be showing their ads next to content that doesn’t reflect well on their brand.
Finally, contextual advertising can help brands reach people when they’re in the right mindset and most receptive. Take the earlier example of shopping for shoes and then seeing shoe-related advertising, or making a smoothie and seeing ads related to that.
Just because someone was shopping for shoes at an earlier point doesn’t mean they’re ready to buy shoes or even interested in shoes at all when they pull up a smoothie recipe in the kitchen. But they might be in the right mindset to pay attention to ads about blenders, or blender accessories, or smoothie straws, or fruit juices, or healthy supplements.
The limitations of contextual advertising
Contextual advertising can be a powerful tool, but it does have limitations.
One limitation relates to the way contextual advertising is bought and sold. Today, publisher-side contextual targeting systems can only be used for more limited direct deals, not plugged into broader programmatic advertising pipelines.
Another issue with contextual advertising is that not all content has context that’s valuable to advertisers. News content is a great example of this issue. Some news content may have valuable context, like local news being valuable to local advertisers who want to reach that audience. But some content that’s broadly-interesting, like content related to a national election, has little context associated with it to help advertisers figure out what to show the audience.
AdThrive’s approach to contextual advertising
Many advertisers rely on third-party technology to determine the contextual content of a page. While functional, these systems are quite simplistic. They work by classifying content into very large buckets like “technology”, “family”, or “pets”, and displaying ads based on these assumed buckets of interest.
For some advertisers, that high-level classification may be enough, but it misses out on the full power of contextual advertising.
AdThrive publishers are fortunate as most of their content is highly contextually relevant. But unlocking the value of that content goes far beyond the generic technology solutions relied on by many advertisers.
Unlike these basic systems, we deeply analyze the types of content that are common across all AdThrive publishers to understand the rich context available in many content categories.
We even built our own contextual targeting system called Marmalade that uses machine learning to determine deeper contextual information about the content!
Marmalade has been running since 2017 and we’ve used this data on thousands of advertiser campaigns across every type of content that AdThrive publishers create. We’re continuously upgrading it in preparation for the death of the cookie in 2022.
How does Marmalade work?
For the technically-inclined, we have built a huge database (running on Amazon Athena) where we can run SQL (Structured Query Language) queries against the content and metadata of any AdThrive page.
Then, we create targeting for ad campaigns based on that data and push it into Dynamo, a high-speed, in-memory database. From there, our ad code, Nucleus, accesses that targeting via a Lambda function to send contextual information to our ad exchange partners so they can deliver the highest-paying ad based on that info!
TLDR: Every ad impression on every page across the AdThrive network can get super-detailed contextual targeting data sent to advertisers to maximize the value of your ads!
The future of contextual advertising
Contextual advertising has a bright future as the entire industry places a new emphasis on high-quality content.
Through our ad code, AdThrive publishers are already able to send advertisers lots of data about their page content, and this head start will prove even more valuable when advertisers need alternative ways to reach their target audiences in the future.
With the clock running out on third-party cookies, we’re excited about developments that are expanding the capabilities of contextual advertising and helping to connect your incredible content with even more sources for ad revenue!