In a recent blog post, Google announced it’s “moving on from Universal Analytics”. This means it’s sunsetting the old Universal Analytics (UA) tool and asking users to migrate to the new Google Analytics 4 (launched in October 2020). Our friends at Merkle | Cardinal Path have already provided a POV on how companies should prepare for this decommission.
What I found most interesting is an update to GA4 that was mentioned in the post. Google says, “Google Analytics 4 will also no longer store IP addresses.” This is almost certainly in response to increased legal pressure related to consumer privacy from around the globe. What does this update mean for marketers?
How To Determine Customers’ Locations in GA4
One of the main use cases for IP address is the determination of a customer’s location. Fortunately, GA4 offers several “modeled attributes” that can be used for that purpose. For example, if you are using the native BigQuery export function from GA4, you can use:
- geo.continent
- geo.sub_continent
- geo.country
- geo.region
- geo.city
- geo.metro
Google creates these modeled attributes through a variety of processes. Some of the attributes are created via configuration settings on the browser. Settings like language preference and time zone can be leveraged to support modeled location attributes. In addition, Google deploys artificial intelligence (AI) to infer location from context and past behavior.
But what about organizations who don’t want to consolidate to an all-Google tech stack? How will the new GA4 support location without the native BigQuery extracts? Google recommends configuring the dimensions for City ID, Country ISO code, Region ID, and Sub Continent Code. These dimensions are based on widely used standards, so sharing the data with other users and applications is simpler.
Support for Customer Identity Use Cases in GA4
Another main use case for IP address is identity of unknown visitors to your channels. How will removal of the IP address from GA4 affect identity use cases? It’s certainly a concern for marketers – however, given the existing privacy restictions and policies, organizations should already be looking to other identity signals outside of IP address to remain compliant. IP address is considered personal identifiable information (PII) under prominent privacy regulations GDPR and CCPA, so it’s subject to the same rules as name and address.
Additionally, IP address has a number of built-in issues that make it unreliable. ISPs typically issue “rotating IP addresses”. They’re also typically tied to entire households rather than individuals.
Google has created a number of other ways to support identity use cases within GA4. User ID is a function within GA4 that allows companies to use their own individual keys and load them into GA when an individual is authenticated. GA4 also has functionality called Google signals that allows users to leverage other Google applications, such as Gmail, to improve GA functionality, including:
1. Remarketing with Google Analytics
2. Advertising reporting features
3. Demographics and interests reports
4. Cross-device reports (this feature is currently in beta)
Additional Reasons to Transition to GA4
Google promises several other benefits of consolidating to a Google-centric tech stack, anchored by GA4. When companies switch from UA, they will have the ability to track YouTube engaged views. This engaged views metric will only be available to GA4. Google has promised to enable premium connectors and metric capture for all users using both standard and 360. These include Search Ads 360 (SA360) and Display and Video 360 (DV360) integrations to publish audiences and improve campaign performance.
Google hopes that by enabling native connectors to Google Ads and BigQuery, companies will find efficiencies and pursue this consolidation.
The deprecation of IP addresses within GA4 will impact marketers. Fortunately, Google is providing a set of tools that will help mitigate those impacts. This is especially true for companies that invest in the full Google tech stack. At the end of the day, impending privacy changes would have necessitated this change eventually. Marketers should always be thinking in terms of privacy-safe ways to advertise and create curated customer experiences.
This post was originally shared on Merkle’s blog.
