Hey Reader!

How do you listen for listenable events on a website you can't see, even with a VPN? That's the challenge I was working through last month, and I came up with a solution that might help some of you facing the same issue.

First, I created a trigger in Google Tag Manager that listens for any event at all. It's a custom event trigger with an event name of .* (and make sure to turn on regex matching).

Then, for your tag, it will look like this:

This way you can see all the possible events coming through in your reports. Anything that starts with gtm_gtm. (for example, gtm_gtm.click) is a built-in GTM event. If you want to ignore these, change your trigger regex to: ^(?!gtm\.).*

Publish your container, wait a couple of days, and see what pops up in GA4. We used this method to find undocumented dataLayer pushes from a third-party software product that one of our clients was using. Now we can use those dataLayer pushes as triggers to create key events in GA4.

If you want to see what's in the payload of the dataLayer push itself, I have some JavaScript to make that happen. Just reply and I'll send it off to you, because if I include JavaScript in this email, it'll probably go to your spam! Let's avoid that.

Any fun measurement hacks you've discovered lately and want to share? I'd love to hear what you're working on.

Dana DiTomaso Founder ​dana@kpplaybook.com​

New Guide: Using Engagement Rate to Diagnose Traffic Quality

Engagement rate in GA4 is one of those metrics that looks basic on the surface but becomes extremely useful once you know how to read what it's telling you about traffic quality.

I just published a guide that walks through what engagement rate actually measures, how it differs from the bounce rate you might remember from Universal Analytics, and how to use it as a diagnostic tool. When I see low engagement rate from a traffic source, it tells me one of two things is happening: you're attracting the wrong audience (an intent problem), or your page experience isn't meeting visitor expectations (an experience problem). Either way, you've got something fixable.

​Read the full guide →​

Understanding Query Fan-Out (And Why It Matters for Your Content)

Salt Agency published a solid explainer on query fan-out, a technique that AI search systems use to expand a single user query into multiple related sub-questions. When someone searches for something like "best running shoes for flat feet," AI generates dozens of related queries to gather a more complete answer.

This changes how we think about content strategy. AI systems now evaluate whether your content covers the full topic, not just whether you've mentioned the right phrase. If you're responsible for content that needs to show up in AI-generated answers, this piece gives you a framework for thinking about topical coverage rather than keyword matching.

The Google Analytics MCP, Explained

The latest episode of The Analytics Power Hour tackles Model Context Protocol (MCP) with guest Sam Redfern, and it's worth your time if you've been hearing about the Google Analytics MCP server and wondering what it actually means for your work.

What I appreciate about this discussion is the honest acknowledgment that the Google Analytics MCP is basically an API layer. That means that it lets an LLM write queries for you, but it's not doing anything you couldn't already do with the API. The more interesting conversation is about where MCP is heading and what it means for how we'll interact with analytics data in the future. If you want to understand MCPs without the “this’ll change everything!!!” hype, this is a good starting point.

Why Dashboards Fail Before They're Built

Michelle Franklin at The Data Drop published a piece on why dashboards fail, and the core insight resonates with something I see often: dashboards fail because no one agreed on what the dashboard is supposed to support.

For example, when someone says "We just want visibility." That sounds harmless, but Michelle points out it really means "We don't know what we want, so give us everything." Then you get the kitchen sink problem where metrics are added "just in case" until there are 20 metrics in a single view and the only person who can interpret it is the person who built it.

Her fix is straightforward: design dashboards around decisions, not data. Most teams start from "What data do we have?" but high-performing teams start from "What decision do we need to make?" If you can't articulate the decision, you're not ready to build. She even put together a free decision-first dashboard worksheet if you want to apply this to your own work.

Speaking at SMX Munich 2026

I'm heading to Munich again for SMX Munich 2026 from March 9-11! I'll be presenting a Looker Studio workshop and giving a talk. If you're going to be there, come say hi — I'd love to meet some of you in person!

That's it for this edition of The Huddle. As always, if you have questions or want to share what you're working on, just hit reply!

Want more? You got it!