Google Floodlight to Recover Cross-Device Conversions Using Signed-In Data

Which conversions Floodlight did not count

Floodlight serves as the conversion tracking system of Google Marketing Platform and sends conversion data to Campaign Manager 360, Display & Video 360, and Search Ads 360 through a single set of tags. When a user lands on a tagged page, for example an order confirmation page, the system checks whether they saw or clicked the advertiser's ad within the attribution window. If a connection is found, Floodlight counts the conversion.

Such a connection is hardest to establish when the user's path runs across several devices or environments. A person can see an ad in a mobile app and then buy on a laptop in a browser. If the system cannot link the impression or click to the subsequent action, the conversion remains unattributed and does not appear in the campaign report. These are exactly the cases Google names as the target of the November change.

How the new mechanism will work

Google describes the change briefly: Floodlight will use Google signed-in signals to recover unattributed cross-device and cross-environment web conversions. A cross-device path means a move between devices, and a cross-environment path means a move between an app and a browser. The wording applies only to web conversions, so the target action must take place on a website, even if the user saw the ad in an app.

This refers to Google signals about users who are signed in to their account. This is how the new mechanism differs from Enhanced Conversions in Floodlight, where the advertiser provides hashed data, such as email, for matching with Google accounts. Google does not name an exact launch day; only the month is given.

The change continues a series of Floodlight attribution improvements in 2026. In its third-quarter plan, Google announced expanded app-to-web and web-to-app conversion measurement across all three products, support for CHIPS partitioned cookies, and the use of browser local storage alongside cookies for post-click conversion attribution. All these changes are aimed at making Floodlight less dependent on third-party cookies.

What Google has not disclosed yet

The help article does not include the details that working with reports depends on. Google has not said:

• how many conversions may be recovered and for which campaign types the effect will be more noticeable
• whether the recovered conversions will be observed or modeled
• whether they will appear in separate columns or be included in the overall conversion metrics
• whether the change will affect automated bidding strategies in Display & Video 360 and Search Ads 360
• whether it is possible to opt out of the new mechanism and how user consent and EEA requirements are handled
• whether historical data will be recalculated.

If recovery starts only in November and does not affect past periods, the increase in conversions after launch may reflect a change in measurement methodology rather than an improvement in campaign results. Google does not comment on this question.

Measurement Implications for Advertisers

Advertisers who measure conversions through Floodlight will start to see in their reports some of the conversions that were previously lost when a user moved between devices or from an app to a browser. This may be especially noticeable for e-commerce with mobile apps and for campaigns where the impression and the purchase often happen on different devices.

Agencies and analytics teams should mark November 2026 as a point of methodology change. When comparing periods, evaluating CPA and ROAS, and analyzing media mix performance, they need to take into account that part of the conversion growth may be related to the new attribution method, until Google publishes details on how recovered conversions are calculated and displayed.