{"id":656,"date":"2022-03-03T10:00:24","date_gmt":"2022-03-03T10:00:24","guid":{"rendered":"https:\/\/graphiclanguage.net\/blog\/?p=656"},"modified":"2022-06-22T20:47:04","modified_gmt":"2022-06-22T20:47:04","slug":"google-topics-explained-a-guide-for-home-builder-marketers","status":"publish","type":"post","link":"https:\/\/graphiclanguage.net\/blog\/google-topics-explained-a-guide-for-home-builder-marketers\/","title":{"rendered":"Google Topics Explained: A Guide for Home Builder Marketers"},"content":{"rendered":"

After announcing it would end support of third-party cookies<\/a>, Google introduced an alternative known as Federated Learning of Cohorts (FLoC). Instead of tracking people on an individual level, FLoC would categorize users in groups, or cohorts, based on shared interests.<\/p>\n

However, Google delayed the launch of FLoC following pushback from regulators<\/a>, industry partners<\/a>, and low adoption rates. This caused Google to shelve FLoC entirely and create a new tracking alternative. Enter: Google Topics.<\/p>\n

Here\u2019s what we know about Google Topics \u2013 and what FLoC was supposed to achieve.<\/p>\n

What the FLoC?<\/h3>\n

Before we dive into what Topics is, let\u2019s discuss what FLoC was meant to do. FLoC was touted to be a new way for advertisers and websites to show relevant ads to online visitors without tracking users across the internet.<\/p>\n

FLoC\u2019s methodology is based on the idea of \u201csafety in numbers.\u201d The API<\/a> assigns the user to a cohort-based on their recent browsing history. Cohorts are dynamic and update every seven days, so as a user\u2019s interests change, so will their FLoC cohort. Unlike traditional third-party cookies, that browsing data would be stored within the individual\u2019s web browser, keeping that information private.<\/p>\n

FLoC privacy issues explained<\/h3>\n

In theory, because these cohorts consist of thousands of people it becomes difficult to identify an individual user. However, privacy experts didn\u2019t think FLoC offered enough protection. FLoC\u2019s use of cohort IDs left users vulnerable to tracking.<\/p>\n

A critical flaw was the lack of protection<\/a> against organizations combining cohort IDs with other tracking mechanisms to identify users. This information can be used for browser fingerprinting since user-specific variation, like browser type, computer, or language, can be used to distinguish users. When combined with cohort IDs, this data can be used to narrow a large cohort to a much smaller group of people. Another issue was the possibility of identifying individual users from multiple visits to the same website. While cohorts are updated weekly, it could be possible to link users to visits over time.<\/p>\n

This criticism led to Google abandoning further development of FLoC for an alternative that addressed these issues. Enter: Topics.<\/p>\n

How Topics will work<\/h3>\n

Like FLoC, the Topics API works within the browser. Each website is assigned to broad categories like \u201cFitness\u201d or \u201cTravel\u201d\u2013 at the moment Google has identified 350 topics<\/a> with more coming as the process matures.<\/p>\n

The browser determines a user\u2019s topics based on their browsing history for that week. Up to five topics are chosen by the browser \u2013 including a random topic that\u2019s meant to complicate identification \u2013 and will be stored for three weeks. When the user visits a website using the Topics API, the browser will pick three topics (one from each of the previous three weeks) to share with the website and ad networks so they can show a relevant ad. However, the website and its associated ad networks can only access topics related to the website\u2019s topic. So, if a user\u2019s topics are \u201cwomen\u2019s clothing,\u201d \u201cautos & vehicles,\u201d and \u201cfitness,\u201d a news website may not be able to access that user\u2019s topics for that week.<\/p>\n

The difference between Topics and FLoC<\/h3>\n

Google acknowledge FLoC\u2019s shortcomings and used the negative feedback to develop Topics. Here are some key differences between the two systems:<\/p>\n