Google announced today that it will postpone plans to phase out third-party cookies in the Chrome browser until 2023, a year or so later than previously planned. Other browsers, such as Safari and Firefox, have already implemented some blocking against third-party tracking cookies, but because Chrome is the most widely used desktop browser, its shift will have a greater impact on the ad industry. That is why the term “cookiepocalypse” has gained currency.

Google says in a blog post announcing the delay that the decision to phase out cookies over a “three-month period” in mid-2023 is “subject to our engagement with the United Kingdom’s Competition and Markets Authority (CMA).” In other words, it is attributing some of the delay to the need to collaborate more closely with regulators in order to develop new technologies to replace third-party cookies used in advertising.

Few will cry for Google, but the company has found itself in a difficult position as the sole company that dominates multiple industries: search, advertising, and browsers. The more Google restricts third-party tracking, the more it harms other advertising companies while potentially increasing its own ad dominance. The less Google restricts tracking, the more likely it will be chastised for failing to protect user privacy. And, no matter what it does, it will be heavily criticized by regulators, privacy advocates, advertisers, publishers, and anyone else with a stake in the web.

To say the least, finding a way to balance those competing incentives has proven difficult. One reason for this is that, as a steward of the open web, Google is attempting to develop its new privacy technologies in the open through the normal process of developing web standards. It has bundled several efforts under the banner of a “Privacy Sandbox,” a catch-all term for a slew of new Chrome and web proposals.

The “Federated Learning of Cohorts” technology, or FLoC, has been the most contentious of these proposals. It is a complex attempt to create groups of demographically similar users in a semi-anonymous decentralized system that advertisers could use to target advertisements. However, no other browser vendor has indicated support for FLoC, and several have explicitly stated that they will block it. The best response Google has received is this analysis from Mozilla, which identifies some issues while not completely dismissing future Firefox adoption.

Google is referring to a “rigorous, multi-phased public development process, including extensive discussion and testing periods” for FLoC and other proposals, which is a fairly clear indication that FLoC will be changed or replaced. “We intend to complete this origin trial in the coming weeks and incorporate feedback before moving forward with further ecosystem testing,” Google says.

The company promises that a “more detailed schedule” will be posted on its Privacy Sandbox website, but in the meantime, here is its current timeline — complete with a disclaimer that this revised timeline will need to be approved by regulators.

Following this open development process, and subject to our engagement with the CMA, our plan for Chrome is to phase out third-party cookie support in two stages:

Stage 1 (beginning in late-2022): We will announce the start of stage 1 once testing is completed and APIs are launched in Chrome. Publishers and the advertising industry will have time to migrate their services during stage 1. We anticipate that this stage will last nine months, and we will closely monitor adoption and feedback before moving on to stage 2.

Stage 2 (Beginning in mid-2023): Chrome will phase out support for third-party cookies over a three-month period that will end in late 2023.