Analytics Roundup – Updates from May 2025

Proactive Analytics and Metrics Monitoring

When your job is analytics, it’s easy to get in the habit of letting others set your priorities. There are always tasks to be done, and to the stakeholders who submit them, every broken metric or new conversion is urgent. I do what I do because I like helping people, so their urgency becomes my urgency. This mindset and the business realities that drive it can mean that strategic work never gets done.  

One of the services we offer is managed analytics, which seeks to get ahead of this dynamic by proactively and continuously ensuring that everything is set up correctly and working properly. An aspect of the service is daily monitoring of Google Analytics property metrics with follow-up investigation if a metric spikes or tanks. We’ve tried various approaches for this, and the one we’ve settled into is to use custom Analytics Insights and aggregate alerts in a Google Sheet. It has proven to be very reliable and helpful for discovering issues and sharing insights. I wrote about the process and shared code for doing it yourself in this article: Google Analytics Insights Monitoring Tool.

I’m also working on a Dataform process for more granular event monitoring and alerting. I’ll be writing about that soon, but ping me in the comments if you’d like an early preview (and would be willing to be an alpha tester 🙂 ).

Google AI Mode

The topic that caused the most buzz in my world this month was Google’s AI Mode for search. My five-cent take is that it really does represent a seismic shift. From a pure technology perspective, it may be comparable to what Anthropic and OpenAI are doing, but 90% of all searches happen on Google. As such, Google has the reach to transform how people behave in a way smaller companies do not. If you want to sound smart when your boss asks you about AI mode, read this.

I also had a few people reach out to me about the trackability of AI Mode – specifically the analyst’s nightmare scenario: is it trackable at all? Turns out it wasn’t, but now it is, and it will be supported in Google Search Console to boot. So at least we’ll have a seismograph.

Product Updates

• Looker Studio properties pane overhaul – if you use Looker Studio, you will have noticed that the properties pane looks totally different. Many of the changes are cosmetic, but even so I find the new look a lot easier to navigate and have found myself using settings I previously overlooked. There are also a few feature enhancements that I like:
  • When you click to apply a filter to a chart, it shows you filters that are based on the data source applied to the chart. This is a nice improvement if you tend to use a lot of filters.
  • When you apply a display format to a metric, then switch to custom format, it populates the custom format field with the correct syntax for the previously selected format. I don’t want to say how much time I’ve fussed about trying to get custom formats to look the way I want – it’s really helpful to have a starting point.
• Set up Google tag gateway for advertisers in Google Tag Manager with Cloudflare, Google
  I was in the midst of auditing a GA property several weeks ago, and Tag Diagnostics threw the message, “Get better signals by enabling Google tag gateway in just a few clicks”. My first thought was, “wait, how did I miss this important-sounding new feature?” Well, it turns out it’s really just a rebrand of first-party mode, which I covered in my roundup last October. Google is pretty vague on the benefits of first-party mode, but the main one I hear about is the fact that it will circumvent ad blockers that prevent Google tags from loading. My POV on this is that if a user does not want to be tracked, they shouldn’t be tracked, period.
  But if you are interested in learning more about “Google tag gateway”, Julius Fedorovicius explains it better than Google, as always.

Workflow

• Where Google Analytics (GA4) Audit Tools Fall Short, Jude Nwachukwu Onyejekwe
  I certainly agree with Jude’s perspective that Google Analytics auditing tools fall short, but what I really like about this article is how he details what an audit should include to be impactful for a business.
  Related: Dana DiTomaso just released a new course: Google Analytics Audit: Your Path to Better Data. Way back I cut my analytics teeth by working my way through Annie Cushing’s fabulous (and sadly defunct) GA audit template. These two resources are a solid replacement.
• Hat tip to Kyle Rushton McGregor for pointing me to Steve Lamar’s GA4 Bulk Annotation Wizard add-in for Google Sheets. This is a handy time-saver for annotating events that affect multiple properties, like bots, news events, and Kyle’s example: major Google search updates.
• How Dataform Handles Incrementality in BigQuery, Krisztián Korpa
  Incrementality in Dataform allows you to make updates to a table according to rules you specify. For example, with GA4 export data, you can make daily updates that merge the last three days worth of data with the data you’ve already written. This means you get fresh data in your reports, but you can also correct your already-processed data when Google retroactively changes previous days events. I love this feature, but it can be hard to get right. This is a great guide.
• How to configure the MCP Server for Google Tag Manager, Uliana Lesiv
  Model Context Protocol (MCP) is emerging as the dominant framework for agentic AI. Stape created this open source solution that allows AI tools to interface with Tag Manager – I haven’t had a chance to try it out yet, but I do a lot of analytics audits and I love the idea of being able to ask questions and get things done directly from an LLM.

Attribution

• Do users really show higher intent when they click through from an LLM to a website?, Dan Taylor
  The answer is … nope, not gonna tell you, you’re going to have to read it yourself. Also includes a breakdown by business category.

Ideas

• Data Micrography, RJ Andrews
  I admit it, most of the time I just take what my charting tool produces and show it to a client. That’s kind of like shopping in the deli section at Whole Foods and calling it cooking. This article is a playful reminder to be deliberate with how we present data.
• The Double Three-Layer Framework for Tracking Setups, Timo Dechau
  This reminds me of one of my all-time favorite concepts in analytics, the Impact Matrix, but Timo’s approach is more down-to-earth, and much easier to follow.

Privacy

• How to Verify Consent Signals Are Being Received Correctly in GA4 for Modeling Eligibility and GDPR Compliance, Dana DiTomaso
  This is part of the auditing course I linked-to above, but available for free. Consent mode is really confusing. This helps.