What Are FLEDGE and TURTLEDOVE?
Everything you need to know about two of Google’s proposed solutions to target online ads without third-party cookies and without exposing user-level data.
TURTLEDOVE is an acronym that stands for “Two Uncorrelated Requests, Then Locally-Executed Decision On Victory.” TURTLEDOVE is an advertising API proposed by the Google Chrome group as a potential way to target ad campaigns without exposing user-level data. With TURTLEDOVE, an advertiser might target people who visited an online shoe store or who viewed a specific product page but without identifying any specific individuals. The key to preserving user privacy with TURTLEDOVE is that the browser itself tracks users and assigns them to interest groups for targeting.
The first tests of the TURTLEDOVE framework are called FLEDGE, which stands for “First Locally-Executed Decision over Groups Experiment.” Note that the acronym for LED in both are the same, a reference to the fact that FLEDGE only tests components of the full TURTLEDOVE proposal. FLEDGE will enable advertisers to replicate the benefits of retargeted advertising in a more privacy-centric way where site actions cannot be tied back to individual users.
TURTLEDOVE was proposed in 2020 by the Google Chrome group and grew out of a previous effort to enable behavioral targeting without user-level identifiers. That effort was called “Private Interest Groups, Including Noise,” or PIGIN.
The FLEDGE trials planned for Q2 2021 have been pushed back. The FLEDGE trials will test the TURTLEDOVE tech, as well as potential changes proposed by outside adtech companies. For example, Criteo proposed that an independent ad server or industry group could control campaign bids and budgets instead of the Chrome browser.
In the meantime, Google Chrome, Google Ads and other digital media and advertising stakeholders are debating TURTLEDOVE in the W3C, along with other ideas in the Chrome Privacy Sandbox. The FLEDGE trials are much anticipated — because until then, the framework is theoretical, and companies don’t yet know how campaigns will perform.
The Google Chrome and Google Ads groups are working on ways to target online ads without third-party cookies and without exposing user-level data. TURTLEDOVE and products like FLoC are among those changes. Google, other browser operators and privacy advocates say these changes are the way online advertising will work in a privacy-centric environment.
TURTLEDOVE, FLEDGE and all the variants of the concept are promising ways to help advertisers maintain a core capability of reaching (or retargeting) interested consumers who have already expressed interest in a product or brand through the actions they take on-site. But in a way that also increases user privacy. A capability like this will be essential to maintaining the effectiveness of digital advertising.
Independent adtech and the publishing industry have pushed back on TURTLEDOVE. For one thing, reporting would operate on a lag and only be available in aggregate, which means programmatic companies would only receive campaign reports every few hours. DSPs might place thousands of bids without knowing which impressions they won and at what rates. This would be a wrench in the gears of real-time bidding (RTB), since bids couldn’t be optimized in real time. Budgeting for campaigns could be a mess, too, since DSPs might discover hours later that they won far more or fewer impressions than expected.
Some of those problems are addressed with the FLEDGE trials, which will test the original TURTLEDOVE ideas and industry amendments. But depending on how long the trials are delayed, it might be difficult to incorporate new ideas before Chrome deprecates third-party cookies.
Others in the adtech ecosystem have voiced concern that Google would control so much of this proposed architecture. Criteo, for example, has proposed that an independent ad server or industry group could control campaign bids and budgets instead of the Chrome browser. FLEDGE actually takes this proposal into account as some evidence that Google is listening to the industry.
What it means:
Advertisers and publishers have been less involved in nitty-gritty infrastructures such as TURTLEDOVE. The debates are happening in the W3C, not in a familiar venue like the IAB, and are heavily technical. But the development of TURTLEDOVE in the Chrome Privacy Sandbox and the prospects of the FLEDGE trials will impact everyone with a stake in programmatic.
It’s impossible to say exactly how online advertising will change because the ideas haven’t been tested yet. As it stands, the TURTLEDOVE framework is an entirely new way to buy online ads, and impression rates would likely change dramatically when put into effect. If independent DSPs and SSPs are hampered by TURTLEDOVE, it would be another incentive for advertisers to work directly with publishers and Walled Gardens instead of using programmatic pipes to span the open web.
For users, TURTLEDOVE aims to protect their privacy from advertisers. Though without any updates to the current framework, adtech companies worry it will lead to a sub-par user experience. People might be bombarded by certain ads because DSPs, ad servers, and SSPs couldn’t effectively frequency cap campaigns.