Hey Reader!
I've had a post rolling around in my head for a while, and instead of waiting until I have time to write the full thing, I wanted to share this idea with you now.
Here’s the core idea: GA4 is really built with ecommerce in mind, which kind of sucks for the rest of us who don't have ecommerce websites. But that shouldn't stop us from using some of GA4's most useful features.
What if you recorded form fills (or really any non-purchase conversion) as a purchase? I know it sounds weird, but hear me out. GA4 has some really useful metrics that only work when you supply it with a purchase event — things like First-time purchaser rate, First-time purchasers per new user, and 31-90 day repeat purchasers. These metrics can be incredibly valuable for understanding your audience behavior, whether you're selling products or collecting leads.
To record a purchase event, you do need to send along a few required parameters: currency, value, transaction_id, and items (which is an array of what was purchased). But you get to decide how you want that to be populated. Google isn't going to get mad at you for using the purchase event when it's just a form fill. The data is yours, and you can structure it in ways that make sense for your business.
If you do try this out, let me know! I'd be happy to include your experience in my post when I do get to writing it. Too many things to cover, not enough time.
New Resource: AI Traffic Analysis Beyond Basic Reports
I just published a detailed guide on moving from counting AI visitors to understanding their behavior patterns. This came directly from a recent Analytics for Agencies office hours session where Jason Dodge of Black Truck Media asked about AI traffic analysis strategies.
AI traffic audiences reveal patterns that change how you evaluate ROI. I've consistently found that AI-influenced customers show higher return visit rates, faster journey completion from first visit to conversion, and increased page engagement. But you won't see any of this if you're only looking at traffic numbers.
The guide walks you through setting up four key audiences in GA4 (First Touch, Any Touch, Return Touch, and Zero Touch AI), then shows you how to use cohort explorations in GA4 to see which visitors actually come back and convert. These audiences don't backfill, so even if your AI traffic is small right now, future you will be thankful you set these up today.
Read the complete guide →
Articles Worth Your Time
———•
|
The Attribution Mirage: When Your "Truth" Becomes a Liability
Connor Phillips published a thought-provoking piece on how attribution models can become more misleading than helpful. The core insight? Your "single source of truth" might actually be a liability when it creates false confidence in flawed data.
My favourite line: “Digital marketing attribution creates a mirage of certainty.” Yes, yes it does.
This matters because many organizations treat their attribution model as gospel without questioning whether it's actually capturing reality. Attribution fraud, channel conflicts, and technical limitations all contribute to a picture that looks precise but might be fundamentally wrong. Before you make budget decisions based on your attribution reports, it's worth asking: what aren't we seeing?
Advice I Wish I'd Gotten About Data Quality
June Dershewitz wrote about the data quality lessons she learned while engaged in "active combat" — which is exactly how it feels sometimes, isn't it? Her advice is practical and hard-won: keep one shared list of issues, agree on what matters most, and be patient because this takes time.
What I appreciate about June's approach is the honest acknowledgment that leaders often want quick fixes when they see patterns of failure, but you can only move forward through observation and learning. There are no shortcuts, and pretending there are just sets everyone up for frustration.
How To Build A Content Audit Assistant With ScreamingFrog and AI
Ian Lurie walks through his process for using AI to improve (not perform) content audits. Like all of Ian’s posts, he lays everything out in detail and doesn’t gloss over any of the challenges.
His workflow uses tools you might already have at hand: ScreamingFrog for crawling and data extraction, APIs for performance data, and AI for specific tasks like assessing user intent or identifying applicable personas. What I love about this approach is that it’s an improvement, not a replacement — you're not throwing your entire existing process in the trash.
Where You Can Find Me
———•
|
Speaking at Tech SEO Connect
I'll be speaking at Tech SEO Connect on December 4-5 in Durham, North Carolina. This is the only conference in the world completely dedicated to technical SEO — no "SEO 101" talks or link building basics, just deep technical discussions from some of the best technical SEOs around. And me. (Before you come at me, I’m not downplaying my technical SEO experience, I definitely know my stuff but it isn’t my full-time focus anymore.)
If you're working on crawling and rendering challenges, data analysis, or machine learning applications in SEO, this is where you want to be. Hope to see you there!
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!