Server-side tracking: It’s happening


Hey Reader!

I want to talk about something I've been deliberately not writing about for a while and that’s server-side tracking.

The honest reason I've held off is that for a lot of the people reading this, it hasn't been a practical option. The setup is technical, the cost is real, and most clients, when you describe what's involved, look at you like you've suggested they renovate the kitchen during a dinner party. There's no good way to time that conversation.

But at the same time what’s happening is that the conditions that make server-side tracking necessary aren't waiting for anyone to feel ready. Between massive ad blocker uptake, increasing browser restrictions, and continuing privacy regulations, it’s already tough to capture as much of the picture as possible. Client-side tracking is just getting harder to execute cleanly.

Now I’m not saying that server-side tracking is some sort of magic wand for tracking accuracy because it absolutely is not. And we aren't at a point where you need to switch everything right away. But I've been doing this long enough to recognize when an industry is about two to three years away from a major shift, and that's where we are with a wider adoption of server-side tracking.

So I'm going to start covering it! Not as a "you need to do this right now!!!!!" panic alert, but as the orientation I wish had existed when I first started trying to wrap my head around how it actually works and what a realistic implementation looks like if you aren't a huge enterprise.

Before I dive in, I want to know where to start. What would be most useful to you right now? Tell me your server-side tracking questions (and no question is silly, ask away!) and I'll use those to shape the content I put together.

Dana DiTomaso

Founder
dana@kpplaybook.com


KICK POINT BLOG

How to Work With a PPC Agency (and Get Better Results)

Sammy Cisek wrote a practical guide to the client side of a PPC relationship, and really it’s applicable to all marketers who deliver leads, not just PPC folks.

Sammy’s core insight is that you see everything up to the click, but without full-circle analytics (which you can do, but it’s often difficult to get set up) you’ll have no idea what happens after someone submits a form or calls. But the campaigns that get better over time are the ones where clients share that information with us.

Sammy's breakdown of how to give useful lead quality feedback is especially good. "Leads are bad" gives an agency nothing to act on. "We got three calls this week from people asking about residential service but we only do commercial" tells them exactly which keywords to add to the negative list.

Even if you don't run PPC yourself (or work at an agency), this is a good framework for thinking about any relationship where there's a visibility gap between what they can measure and what you actually know.


Articles Worth Your Time
———•

Google AI Mode Is Testing Overlay Cards

Barry Schwartz at Search Engine Roundtable flagged something worth paying attention to: Google is testing a change where some links in AI Mode open as overlay cards rather than sending users directly to the website. That extra step between the AI response and your site means even fewer clicks making it into your analytics.

This matters because we've already been watching organic click data erode as AI Overviews answer more queries without requiring a site visit. Overlay cards are another layer on top of that. If you're currently benchmarking your organic traffic, now is a good time to document where things stand because whatever "normal" looks like today is going to keep shifting.


Why SEOs Need to Start Looking at Social Data

Celeste Gonzalez at RicketyRoo built a Looker Studio report template that brings social and search data together in one view, and it's worth grabbing even if social isn't your primary focus.

What I appreciate about Celeste's framing is that this isn't a social media report, but rather a search report that includes social signals. The template pulls from Google Search Console, GA4, and Metricool (for TikTok and YouTube data), and the core question it helps you answer is: are months with more social activity also months with higher branded impressions?

The reason why you should try this out is the connection to AI search, which is of course so hot right now. We know that LLMs are pulling from TikTok videos, YouTube content, and other social platforms as source material for the answers they generate. That means that if you're only reporting on what's happening on your website, you're likely looking at the end of a journey that started somewhere else. Having a way to at least see the correlations across channels is a useful starting point, and the template is free! (We love free.)


The Browser Signal That Could Quietly Kill Your Retargeting

I flagged server-side tracking as something I'm starting to cover, but I also want to be clear about what it doesn't solve, and Global Privacy Control (GPC) is one of those things.

I will be writing about this myself but for now I recommend this piece from the team at Seresa as it’s a useful breakdown of GPC and what it means for your tracking. GPC is a browser-level signal that automatically opts users out of data sale and targeted advertising across every site they visit, without them needing to click a single opt-out link. As of January 2026, twelve US states legally require businesses to honor it and Chrome, Safari, and Edge are required to include GPC natively by January 2027, which means a significant increase in opt-out signals is coming.

I recommend sharing this information with your team or clients, especially those who run paid campaigns, because this has serious implications for retargeting audiences. When a user has GPC enabled, you can't add them to behavioral segments or retargeting pools. And this applies whether you're using client-side or server-side tracking as GPC restricts what you're allowed to collect and share, not just what you're technically able to collect and share.

Start educating people around you about this now. It's easier to have that conversation before the data gap is visible than after someone notices their retargeting audiences have shrunk.


Where You Can Find Me
———•

MeasureSummit — May 12-13

I'll be speaking at MeasureSummit this spring — it's free and fully online. Details on my session are coming soon, but if you're looking for a solid analytics-focused conference that you can attend without a travel budget, this is it!


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!

📈

Practical GA4

Sign up →

📘

Free Resources

Get yours today →

🛠️

GA4 Workshops

Level up in GA4 →

Was this email forwarded to you? Sign up here!

PO Box 68171 RPO Bonnie Doon Shopping Centre, Edmonton, Alberta T6C 4N6
Unsubscribe · Preferences

Analytics Playbook by Dana DiTomaso

Analytics Playbook gives you the analytics skills you need to land more clients, level up your career, or make smarter marketing decisions. Get bi-weekly insights curated by analytics expert Dana DiTomaso. Each issue includes expert tips, must-read articles, and free resources, all designed to help you take action and see real results.

Read more from Analytics Playbook by Dana DiTomaso

Hey Reader! Here is a mantra that I want you to carry with you in your professional career: we are not data accountants. Somewhere along the way, marketers got sold the idea that good analytics means precise analytics. That if the numbers don't add up perfectly, something is broken. That your job is to account for every single session, every click, every conversion, and if you can't, you've failed. The problem with that whole idea is that the data was never complete. It just used to be less...

Hey Reader! I've been seriously irked lately. Client sites are getting hammered with spam and bot traffic, primarily from China and Singapore, and I've been running experiments to see what I can filter out in Google Tag Manager before it ever pollutes our client’s production GA4. Here's where I'm at so far: GeoIP — block based on country. Only worked a little, since the IPs that spammers use typically aren't in geoIP databases. Browser version — most spammers are using an older version of...

Hey Reader! I’ve been working on an audit of a client’s dataLayer implementation for a custom-built eCommerce website. There’s a lot going on, but one of the major issues is that their developers got bad dataLayer advice in the past and now we’re working on cleaning it all up. But what's been on my mind more than the technical issues is the documentation itself. When you're working with people who only encounter analytics occasionally, your audit document has to do two jobs at once. It needs...