Google Chrome has been the leading browser of choice for online consumers for quite a few years now. According to www.statcounter.com, in May 2022, Chrome was dominating the web browser market share by 64.34%. As Google previously announced, Chrome internet browser will stop supporting the user-tracking technology called third-party cookies by late 2024. Cookies are small files sent to your browser from websites you visit. These files track and monitor the sites you visit and the items you click on these pages. This announcement sent shockwaves through the AdTech industry. Third-party cookies have been a foundational part of the technology that enables personalized digital ads, allowing advertisers to personalize ads to users based on their behaviors and site-visit history for decades.
So, what does a cookieless world mean for Dstillery and our audience model building? We have it all covered! Besides offering our premium ID-based methodology for building audiences with cookies, we also have ID-free audiences that do not rely on any user-tracking, cookies, or IDs.
ID-based Custom AI Audiences
Dstillery builds predictive models for audiences by analyzing the data on users who visited certain domains. This data is completely anonymous. It comes from sources like the programmatic ad environment (the bid-stream) and 3rd party data partners, and it is cookie-based.
Once we receive this cookie-based data, we store it under anonymous user profiles in our system. When we talk about examining data on users, we mean understanding their online behavior. After analyzing common patterns among the users who visited a certain domain, Dstillery builds a model consisting of the online behaviors most likely to lead to conversions or actions among all these users. This predictive model is then used for targeting consumers on the internet who would be most likely to exhibit the same behaviors that the model has already identified.
The model is then used daily to score devices that match the model features and build a probabilistic ID-based audience for targeting. All ID-based audiences are refreshed every day by scoring new qualifying devices in and scoring out the ones that do not fit the model criteria anymore. This guarantees that Dstillery’s audiences are always relevant, up-to-date, and ready for immediate use.
Dstillery offers a wide range of ID-based audiences. They fall into different categories based on the way they are built and their composition.
How are Custom AI audiences built by applying the ID-based methodology?
For Custom AI audiences, Dstillery requires first-party data that comes directly from the client, not just any domain or source, in order for the target user behavior for that brand’s signal to be identified. This first-party data can be collected by placing a Dstillery Pixel on the client’s website or by ingesting CRM data offline. The collected first-party data is then analyzed by finding common browsing behaviors among the client’s consumers and used to create Custom AI models that are 100% tailored to the client’s brand and needs.
ID-based Custom AI audiences deliver some of the best results and precision. Why?
First-party data, whether it comes from pixel-based devices or from other data sources such as CRM-based lists, provide the strongest signal for interest in your brand. It is also important to note that the models for Custom AI audiences are rebuilt every 24 hours due to the constant flow of new devices to pixels that sets off the model refresh. Additionally, all Custom AI audiences are refreshed daily by having devices scored in and out of them, which always guarantees strong relevance and excellent performance. Internal studies have shown a 15x lift in conversion rate for qualified customers identified in our ID-based Custom AI Audiences.
ID-free® Custom AI Audiences
Dstillery already has a plan and a solution for the ‘Post Cookie Era.’ Enter ID-free® targeting, a methodology for model and audience building without using third-party cookies. Dstillery has adapted its proven ID-based model-building technology so that cookie IDs are not necessary to create high-quality predictive audiences. In our ID-free environment, we do not rely on tracking the devices we are trying to target. This makes ID-free the ideal choice for advertisers who want to get ahead of consumer privacy concerns without compromising campaign performance and scale.
Dstillery’s ID-free methodology for building audiences is based on patented neural network technology that analyzes hundreds of millions of anonymous digital journey patterns to learn the behavioral signals underlying any web visit. By using this influx of hundreds of millions of anonymous digital journeys our algorithm learns how, when, and from where people browse the web and derive a multidimensional map of the web where the closer the points, the more similar the actions of visitors to those sites. Dstillery builds these statistical predictive models from opt-in first-party data (such as pixels, surveys, and panels) that are then applied to the ID-free universe of prospects. This enables brands to identify the right impression opportunities among all these untracked browsers and devices.
Our Chief Data Scientist, Melinda Han Williams, explains how our cookieless targeting solution is not the same as contextual targeting. “ID-free doesn’t read the words on the webpage or scrape content from websites, it is entirely a behavioral solution. So, it is picking up on signals that reflect interest in the brand, regardless of the content on the webpage. ID-free can take a high-performing 1st party signal, amplify it and activate it at internet scale.”
Who are ID-free audiences for?
The core of ID-free technology, and its inherent cookieless nature, delivers performance and scale for all brands. ID-free audiences are particularly great for industries that have strict privacy standards, like healthcare and finance. We can confidently conclude that ID-free is an excellent solution for everybody, for two reasons:
- Cookies will be going away, so brands can get ahead of the curve now.
- It opens up a whole universe of devices that brands have not been engaged with, so it’s good for performance and incremental scale today!
How is Custom AI implemented through Dstillery’s ID-free methodology?
Just like ID-based, ID-free Custom AI audiences deliver some of the best results and precision because these models are built with first-party data or keywords. Dstillery’s multidimensional Map of the Internet (MOTI) is compounded with the client’s first-party signal. This allows our AI to learn the browsing patterns of anonymous opt-in devices identified in the client’s seed and to create a base model. Then, the identified patterns are extended to all websites by utilizing our neural network technology, allowing the client to select opportunities that are predictive of desired behaviors. When a Custom AI model is generated and applied to the ID-free universe of devices, this creates tailored ‘just-for-your-brand’ predictions to identify the most valuable impressions for the client’s brand.
ID-free Custom AI is a proven post-cookie solution, available now, it works on every browser and across all types of devices including desktop, mobile, and tablet. It performs on par with the best ID-based Custom AI solutions, sometimes even better based on current campaign results. ID-free Custom AI could be used strategically as a way of adding incremental scale to any client’s campaign by unlocking a whole new universe of devices that aren’t being tracked.
How is ID-free Custom AI tailored for healthcare brands?
Custom Patient Targeting is a version of Dstillery’s ID-free Custom AI product made specifically for healthcare. Because the core ID-free methodology doesn’t create profiles, doesn’t “know” the names of the sites that any user visited to make predictions, and doesn’t target any IDs or cookies, ID-free targeting offers optimal precision, customization, and compliance for healthcare brands in programmatic advertising. For advertisers in healthcare looking to engage with consumers in a completely privacy-friendly manner, Custom Patient Targeting is Dstillery’s privacy-safe behavioral targeting solution that provides performance and precise reach without user tracking.