Posted
March 29, 2024
Anyone who has worked with a Consent Management Platform (CMP) knows this well: you plug it in, fire it up, and it starts analyzing the cookies and tags on your site. You watch the scanner chug, and you’re excited, because this should be easy – with a couple clicks, you should be able to use the output to power your cookie disclosures, as well as tell the CMP which tags and cookies to suppress after a consumer opts out.
The scanner finally spits out the list, and you’re thrilled to see vendor names and classification suggestions. But then as you scroll down, you see them – the dreaded Unknowns.
“Unknown” is a common designation for tags and cookies that a CMP doesn’t recognize. Most CMPs suggest that clients use a search engine to figure out what the unknowns are. If it’s one or two, that’s not a problem, but for some companies, the list of unknowns can be a large number – we have clients whose CMP scans come back with hundreds of unknowns. Manually searching for hundreds of things, one by one, is time-consuming and painful. For many companies, it’s not practical or even possible to complete the search project at initial implementation, or to keep up with it as new tags and cookies get put on the site over time, leaving them with potential compliance gaps.
This isn’t a knock on CMPs – recognizing all of the thousands upon thousands of tags and cookies in use across the digital ecosystem is a tall task. But clients need more help managing tag and cookie inventories that can easily run into the hundreds and thousands.
There is a better way.
As discussed in previous posts on our Boltive AI Engine (patent pending), we utilize state of the art AI technology to solve problems for clients in both security and privacy. Seeing our clients struggle with the manual search process for tags and cookies – what we call “Objects” in Privacy Guard – we recognized an area where we could apply our AI skills to ease significant pain.
Traditional search engines are hard to use for this kind of research because they rely heavily on keyword matching and surface-level indexing. As a result, they don’t understand context well, and can struggle to deliver relevant information. You may have some luck finding some information on common cookies in the first page of search results – such as a discussion on a developer forum or a disclosure in someone else’s privacy notice – but the entire exercise quickly becomes looking for needles in the world’s largest haystack. This is why so many companies abandon the project, and learn to live with the risk.
AI tools, on the other hand, excel at interpreting context, extracting meaning, and synthesizing complex information, resulting in highly relevant, accurate suggestions tailored to the user's query. We have developed a new AI model – which we call the Boltive Object Identification Engine – that leverages AI’s ability to understand key inputs from our clients’ tag and cookie inventories, to produce high-quality results that make the vast majority of their unknowns known, in a matter of minutes.
We have a large ad-supported publisher client who found themselves struggling with hundreds of new cookies appearing in their CMP scans every month, most of which would come back as unknown. This publisher invests a lot of effort in respecting consumer choice, so leaving these cookies as-is and just living with the fact that they wouldn’t be suppressed after consumer opt-out was not an option.
Every month when their CMP scan would run and produce this large list of unknowns, their privacy operations manager would stop what he was doing and spin up his search engine. 40 hours of mind-numbing work later, he would have his cookies identified, classified in the CMP, and be able to get back to more strategic initiatives.
When we released the Object Identification Engine, this entire monthly exercise went away.
Our client has gotten a whole week back. Each month.
He’s now spending this time on things that help move the business forward, such as building cross-team collaboration with stakeholders across the organization and preparing for laws going into effect later this year.
This client is not alone in having to cover a lot of ground, and getting dragged down by manual processes. In addition to eliminating the pain of using search engines to classify your cookies and tags (or the risk of not doing it), Privacy Guard helps automate a number of processes that you may be doing manually today. If you find yourself needing to do more with existing – or perhaps fewer – resources, consider how automation can help you.
If you would like to find out how we may be able to help, please reach out – we’d love to offer you a free assessment of your site.
