How AI Targets Audiences
If you’ve ever received an ad and thought, “wow, I was just talking about that” or even gone to stay with a relative and then started receiving ads for the toothpaste that they use, you’ve been targeted.
Ad and search engine algorithms use machine learning to observe what you search, when you search it and any signals that they can glean from those actions. These algorithms gather information about our habits including how many touches or related searches you have to make in order to be primed to buy as well as what you’re generally interested in.
This is called predictive analytics. By analyzing billions of users across the internet, algorithms learn to predict actions based on your “personality” online, or those with similar habits. The AI then matches these personalities to interest-based targeting segments that an advertiser can use to target you manually, or some AI can automatically test, predict and deliver to your audiences.
For example, if your brand is a high-end baked goods company, you can go to a DSP and target those who have an interest in high-end liquors, baked goods generally and have a household income of $60K+. This is what we call manual targeting.
How does manual audience targeting in advertising work?
Manual targeting allows marketers to pinpoint their desired customers by specific demographics, behaviors, interests, or any other data points they have available. This can be done through a variety of mediums including email campaigns, social media ads, and even text messages!
The main difference between manual and optimized advertising is that with manual audience targeting in advertising, you can directly target specific individuals on your platform of choice, while automated audiences rely solely on data collected about your current or prospective customers. This could include age, location, interests, and much more.
However, manual audience targeting in advertising does come with a few drawbacks. For example, you may not always get the results you want on your advertisement when using manual audience targeting in advertising. Also, because different audiences may be more valuable or have more competition, they may have higher CPMs (cost per million impressions).
How has AI improved Targeting with Cookies?
Targeting with cookies became possible with the introduction of AI, which has made it easier than ever before for marketers to create targeted content that is more accurate and effective at connecting audiences with brands they love.
One of Google's technologies that use AI for predictive analytics is called "Targeting with Cookies". It focuses on predicting users' next interactions with a website by tracking their behavior across different websites. It does this because it can learn where the user is headed next and display relevant ads to increase click-through rates (CTR). Cookies are little bits of data stored on your computer. Most of them are useful and help to do things like remember your preferences on a website, or keep you logged in to your favorite shop even after you close out. Other forms are used for tracking, and eventually targeting.
The idea of tracking visitors to a website isn't new, but with the birth of "Big Data," it has become an increasingly important part of online marketing. Websites can collect vast amounts of information about their users and utilize them to target specific ads. These same techniques are also being used by criminals for phishing scams or stealing personal information.
This is one area where AI and machine learning can help improve the ability of companies to track visitors and target ads. This helps to filter out bad actors while targeting more relevant ads at legitimate users. However, it also allows these same companies to build a profile of their users and sell it to others looking for this type of data.
Third party cookies (cookies that are owned and operated by an entity other than the website owner or brand) have recently come under scrutiny, because of the shady business practices of sharing/selling user data across websites or making it available for purchase. Because of the sophistication of these machine learning algorithms, companies have been able to collect, segment and sell audiences based on your perceived political alignment and potential medical ailments which consumers should feel some sort of way about.
How AI Targets Audiences in Advertising & What will happen when cookies go away
Imagine if you could target your message to the exact person who would be interested in buying your product. That's what AI is doing with advertising by calculating each person's browsing history and interests so that only they see the ad. This means more targeted ads for everyone. Artificial intelligence can predict what you want to buy, how much money you have available to spend on a product, and even where you are going next based on your web browsing habits.
Typically this is done by either ingesting your current customer list and creating a pool of users who are “look-alikes” to your current customers; or it can be done through machine learning and pixel placement. The pixels tell the ad server and AI what actions a user/device took on the website which allows the AI to learn and find similar users.
Companies have been using this type of technology for years because it gives them an upper hand over other companies who don't use it as well or at all. It's not just big corporations who are realizing the benefits of artificial intelligence either-- smaller businesses are also starting to see how this type of technology can help them grow their company by targeting potential customers with personalized ads. Companies like Google and Facebook and Microsoft LinkedIn invest heavily in artificial intelligence-driven technologies and DSPs utilize a ton of third-party cookie-built audiences. In 2019, third party audience data use and purchase in the United States amounted to 19.7 billion U.S. dollars, of which 11.9 billion was spent on data itself.
How does targeting work with contextual ads?
Contextual advertising uses tailored advertisements on relevant websites. A page that focuses on books, for example, may use the algorithms behind a contextual advertising solution to include relevant ads specific to the context of the page (such as sales ads for reading glasses or bookmarks)
Where contextual ads showing up can say something about your interests. For example, if you're reading an article about a new exhibit at a museum, you might see an ad for that same museum. Or if you're shopping online and click on a page with women's clothing, your browser might show you an ad for shoes or handbags - since both of those products could be cross-shopped with the clothes you just viewed!
With contextual AI, brands can associate their advertisements with the consumer's interests. This means that ads will only show when they believe it is relevant based on what has been learned about them so far through machine learning methods like artificial intelligence (AI). It's an incredibly powerful tool for marketers because this type of advertising does not require any third-party data or cookies which can result in increased safety when compared to traditional marketing tactics.
CatapultX’s Contextual AI
We are entering an era where AI will be able to target customers in ways that were never before possible. With the loss of cookies, marketers need a new way to track customer behavior and preferences without relying on cookie-based data.
CatapultX’s contextual AI platform allows you to use contextual targeting and signals to serve ads during specific, related moments within video. For the first time in history, an advertiser can upload ad creative, use their previous keywords or targeting settings to allow the AI to read videos across the web to find the perfect spot to serve your ad. Instead of serving reading glasses on Barnes and Noble, reading glasses could be advertised during a library scene on a video or when one of the main characters is wearing glasses.