What happened?

On May 20, 2026, Google presented Google Marketing Live 2026 on the last day of Google I/O. One day later, Google presented major updates to their Marketing Platform stack to the EMEA region. The event was streamed live and our engineers have analyzed its announcements. In this article, we will go beyond the marketing hype and share our notes on the most important takeaways for Google Analytics, Google Tag Manager, and your Data Stack.

Google Analytics 4 x Meridian

The third wave of marketing attribution in GA4 is coming right at us (who remembers Google Attribution 360?). In 2024, Google announced Meridian as an open source Python project. Now they are integrating this feature as a core GA4 capability. Over the last couple of years, Google has quietly removed first-click, linear, time decay, and other deterministic attribution ‘models’ and is now going all out on their probabilistic approach.

We’ve seen huge investments from organizations for the adoption of expensive attribution platforms, but most of them feel the same pain: your deterministic model may be state of the art and extremely optimized, but if it relies on data from Google Analytics and other major platforms, the data quality is too low to make trustworthy predictions. As always: garbage in = garbage out.

Going probabilistic with marketing mix modeling can tackle some of these challenges, but there are some caveats:

• It requires a completely different mindset: one in which you dare to experiment. If you’re not willing to turn the dials on ad spend, MMM is not going to help you.
• It assumes that organizations will go for Google as their main marketing analytics warehouse, even though we have seen more and more organizations moving to true first-party alternatives.

Google Tag Manager x Google Tags

If you’ve ever dealt with Google Tag Manager, you were probably puzzled by the jargon jungle. Google Tag Manager (gtm.js), Google Tags (gtag.js), Google Tags (in Google Ads and Google Analytics), Google Tags (in client-side GTM), Google Tags (in server-side GTM), and Google Tag Gateway; they all sound the same but serve completely different purposes. Google has tried numerous times to make tagging and measurement easier for non-technical users while sticking to extremely confusing terminology. The Google tracking and measurement stack has a long history with a huge amount of users depending on it to keep working like it always has. This has introduced a lot of legacy and redundancy.

In 2022, Google introduced the feature in the interface of Google Analytics 4 and Google Ads to set up the behavior of your own ‘Google Tag’ (gtag.js). This allowed users to fine-tune the collection behavior of Google’s main script without the need for a developer. We think that Google hoped for a widespread adoption of this feature, but it seems that the majority of users still prefer to use Google Tag Manager or ask someone to directly change the script themselves.

Now, Google is pushing for more unification: the GTM container (gtm.js) will become the Google Tag itself, and inside Google Tag Manager, you will configure which destinations data should be routed to (GA4, Google Ads, Floodlight, etc.). Technically speaking, this will all be orchestrated by a single script with fewer dependencies, which will improve performance and reduce complexity.

This sounds great, but remember that 90%+ of the organizations we work with also use Google Tag Manager to orchestrate other scripts like the Facebook Pixel, LinkedIn Insight Tag, Hotjar, Klaviyo, etc. So, while this is a step in the right direction, as we adopt this new architecture, we will develop new best practices to juggle the unified Google scripts with other tracking scripts that need to run.

Universal Commerce Protocol (UCP)

This is a brand-new standard, in development since 2025. It is not exclusively developed by Google, as they are part of the UCP Tech Council. As of yesterday, this council is represented by Google, Shopify, Etsy, Target, Wayfair, Amazon, Meta, Microsoft, Stripe, and Salesforce. The goal of UCP is to create a standard for e-commerce data that can be used across all platforms. This will allow for a more unified and consistent way of tracking e-commerce data, which will be a huge benefit for marketers and analysts.

Sounds familiar? Because it is. Feed management has been around for what seems like ages, and anyone dealing with e-commerce data knows about XML files and JSON feeds. After 20 years, it’s finally time to start pulling data at scale, instead of pushing data. Google would love to fetch and display e-commerce data in near real-time so that they can populate search results (including “AI Overview”) with relevant offers. Consumers tell Google what they are in the market for, and companies tell Google what they have to offer. A tale as old as time, but now with a new twist.

Integrating with this new protocol is a technical feat, and our team is here to help you. Just keep in mind that this is still a closed program, currently only available to companies within the U.S., Canada, and Australia. We have applied for access and will keep you posted on what we find.

To be continued…

That is it for today, but we will keep an eye on other announcements from Google and do our best to de-fluff and demystify them for you. If you have any questions in the meantime, please don’t hesitate to reach out.