Marketing Analytics with Fexingo · 2026-06-25 · 10 min
Key moments - from our scoring
Substance score
50 / 100
Five dimensions, 20 points each
Server-side tagging has emerged as the foundational infrastructure fix that matters more than switching attribution models, because most attribution failures stem from corrupted client-side data rather than flawed modeling approaches. Lucas and Luna discuss how standard JavaScript tags are increasingly unreliable - ad blockers, Apple's Intelligent Tracking Prevention, and slow page loads cause event loss that flows into attribution models as false zeros. A mid-sized e-commerce retailer spending $400,000 monthly on Facebook and Google Ads was losing roughly 30 percent of Facebook conversion events; after migrating to Google Tag Manager's server-side container in late 2025, attributed conversions jumped 22 percent with identical spend, revealing that Facebook was actually their strongest mid-funnel channel. Beyond data recovery, server-side tagging provides first-party data control - allowing teams to hash emails, remove IPs, and enforce consent before forwarding data to ad platforms. The setup barrier is primarily engineering time rather than cost; hosting runs $50-$100 monthly, but requires developer expertise or agency support. Key trade-offs include latency (negligible in most cases) and server reliability risk. The discussion clarifies that cleaner data still requires validation through incrementality testing and geo-lift studies, and can create new biases in view-through attribution if lookback windows are not properly configured. As Apple's privacy restrictions persist and Chrome phases out third-party cookies, server-side tagging is becoming table stakes for marketers managing incomplete data.
The mid-sized e-commerce retailer discussed lost roughly 30 percent of Facebook conversion events to ad blockers and Apple's Intelligent Tracking Prevention - roughly $120,000 in monthly ad spend that appeared unconverted when it actually converted. Ad blocker usage now reaches about 40 percent of global web users.
Hosting costs range from $50-$100 monthly using Google Cloud Run or AWS, and the main barrier is engineering time - typically a few days initially for setup and ongoing monitoring. Setup cost was estimated at a few thousand dollars for the case study retailer, with a payoff of $88,000 in additional attributed revenue monthly.
No. Server-side tagging provides cleaner data and removes technology-based undercounting, but it does not establish causality. You still need experiments to determine whether conversions were actually caused by ads or would have occurred anyway.
The primary trade-off is latency - data traveling from browser to your server to the ad platform adds a few hundred milliseconds, which is usually negligible but can affect time-sensitive events like click-to-call. Server downtime also means complete tracking loss until recovery.
Server-side tagging routes data to your server first, allowing you to transform it - hashing email addresses, removing IPs, and adding consent flags - before forwarding to ad platforms. This contrasts client-side tags, where the Facebook pixel sends data directly from the browser to Facebook with little marketer control.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode packs a reasonable number of non-obvious points into 10 minutes - particularly the warning that cleaner data creates new biases (over-adjustment, view-through inflation) - but some coverage of basics (ad blockers block client-side tags, GTM server-side exists) is familiar territory for the target audience.
Server-side tagging can also reduce the delay in attribution, which might make campaigns look more effective than they actually are
if your tag is dropping 30 percent of events on the floor, no model can fix that
The reframe of 'attribution model failure as a data pipeline problem, not a modeling problem' is a decent organizing argument, but the overall content follows a well-worn industry discussion; there is no genuinely contrarian or first-principles claim that challenges consensus thinking.
The attribution model was not broken - the data pipeline was
Server-side tagging is essentially plumbing. It is not glamorous. But it is the single highest-leverage change you can make to improve measurement accuracy right now
This is a two-host explainer format with no external guests; neither host demonstrates clear senior-practitioner credentials - the key case study is described as 'a retailer I've been following' rather than direct implementation experience, and the buy-me-a-coffee pitch signals a small independent operation.
A mid-sized e-commerce retailer I have been following switched to server-side tagging
Buy me a coffee dot com slash fexingo, if you have gotten something out of them
There are concrete numbers throughout - $400k monthly spend, 30% event loss, 22% conversion lift, $50 - 100/month hosting, 40% global ad-block rate - but the retailer is unnamed, the $88k revenue figure rests on an unstated average order value assumption, and the 40% stat is cited without a source.
spending about $400,000 a month. Before the switch, they estimated they were losing roughly 30 percent of their Facebook conversion events
their attributed conversions from Facebook jumped 22 percent
Luna contributes one genuinely substantive observation (view-through inflation risk from cleaner data), and the pacing is efficient, but the dialogue reads as scripted - Lucas drives all the analysis, Luna mostly provides prompts, and no claim is ever challenged or pushed back on.
with server-side, you might see a jump in view-through conversions, but that could be misleading if you do not set proper lookback windows
Is there any downside? I imagine latency is a concern
Computed from the transcript - who did the talking, and the words that came up most.
Episode 74 of Marketing Analytics with Fexingo dives into server-side tagging - a technical shift that is quietly reshaping how marketers measure campaign performance. Lucas and Luna unpack why client-side tags miss conversions from ad blockers, browser privacy updates, and slow page loads. They walk through a concrete example: a retailer that lost 30 percent of its Facebook conversion data after Apple's Intelligent Tracking Prevention kicked in. By moving tags to a Google Cloud server, the retailer recovered a 22 percent lift in attributed revenue from the same ad spend. The hosts explain how server-side tagging preserves first-party data integrity, reduces data loss from tracking blockers, and gives marketers cleaner signals for multi-touch attribution models. They also discuss the trade-offs - higher engineering cost, latency risks, and the temptation to overcorrect with inflated view-through windows. No platitudes, no vendor pitches. Just a focused look at why this infrastructure change matters more than any attribution model tweak.
Transcribed and scored by The B2B Podcast Index.
Lucas: We have talked for weeks about why attribution models break - incrementality tests, geo lift studies, view-through inflation. But there is a more fundamental reason many of these models fail. They are working with bad data. Luna: Bad data coming from where exactly?
Lucas: From the tags themselves. Standard client-side tags - the little snippets of JavaScript that fire when someone loads a page or clicks a button - are increasingly unreliable. Ad blockers block them. Browser privacy updates delay or kill them.
Slow page loads drop them entirely. And that missing data flows into your attribution model as a zero, which is a lie. Luna: So you are saying the problem starts before the attribution model even touches the data. Lucas: Exactly.
And that is why server-side tagging has become the hidden infrastructure fix that actually matters more than switching from last-touch to multi-touch. Server-side tagging moves the tracking logic from the user's browser to your own server, or to a cloud server you control. The tag fires on the server, not on the client. That means ad blockers cannot see it.
Intelligent Tracking Prevention cannot delay it. It just runs. Luna: I have heard server-side tagging described as just moving the tracking from the browser to the cloud. But does that really fix the data quality issue?
Lucas: It does, and here is a concrete example. A mid-sized e-commerce retailer I have been following switched to server-side tagging via Google Tag Manager's server-side container in late 2025. They were running Facebook and Google Ads, spending about $400,000 a month. Before the switch, they estimated they were losing roughly 30 percent of their Facebook conversion events due to ad blockers and Apple's Intelligent Tracking Prevention.
Luna: Thirty percent is huge. That is $120,000 in ad spend per month that looked like it was not converting, when it actually was. Lucas: Right. After moving to server-side tagging, their attributed conversions from Facebook jumped 22 percent - same ad spend, same campaigns, just more complete data.
Their cost per acquisition dropped accordingly. And they were able to feed that cleaner data into their multi-touch attribution model, which suddenly started showing that Facebook was actually their best-performing channel for mid-funnel conversions, not just top of funnel awareness. Luna: So the attribution model was not broken - the data pipeline was. Lucas: That is the key insight.
Marketers spend all this time debating attribution windows and weighting models. But if your tag is dropping 30 percent of events on the floor, no model can fix that. Server-side tagging is essentially plumbing. It is not glamorous.
But it is the single highest-leverage change you can make to improve measurement accuracy right now. Luna: And it also protects first-party data, right? Because you control the server, you can strip out personally identifiable information before sending it to ad platforms. Lucas: Yes, that is a second major benefit.
With client-side tags, when a user lands on your site, the Facebook pixel sends data directly from the browser to Facebook's servers. You have little control over what gets sent. With server-side tagging, the data goes to your server first, you can transform it - hash email addresses, remove IPs, add consent flags - and then forward what you choose to Facebook. This is increasingly important as regulators tighten rules on data sharing.
Luna: I can see why larger companies with engineering resources would jump on this. But what about smaller teams? Is server-side tagging accessible to a brand spending, say, $50,000 a month? Lucas: It is becoming more accessible.
Google Tag Manager's server-side container is free, but you need to host it somewhere - Google Cloud Run, Amazon Web Services, or even a virtual private server. The hosting cost can be as low as $50 to $100 a month for a small to mid-size advertiser. The bigger cost is the engineering time to set it up properly and maintain it. That could be a few days of a developer's time initially, plus ongoing monitoring.
Luna: So the barrier is less about budget and more about technical capability. Lucas: Exactly. If you do not have a developer who understands JavaScript and cloud infrastructure, you might need to hire an agency or a freelancer. But the return on investment can be substantial.
That retailer I mentioned saw a 22 percent lift in attributed conversions. For a $400,000 monthly spend, that is an extra $88,000 in attributed revenue per month, assuming a conservative average order value. The setup cost was a few thousand dollars. Luna: Is there any downside?
I imagine latency is a concern. Lucas: Latency is the main trade-off. With client-side tags, the tracking fires on the user's browser immediately. With server-side tagging, the data has to go from the browser to your server, then to the ad platform.
That adds a few hundred milliseconds. In most cases, that is negligible. But for time-sensitive events like a click to call or a real-time bidding response, it could be an issue. Also, if your server goes down, you lose tracking entirely until it recovers.
Luna: So it is not a silver bullet. But it seems like the direction the industry is heading. Lucas: It is. And there is another pitfall to watch for.
Some marketers, once they switch to server-side, start seeing more conversions - because they were undercounting before - and then they over-adjust. They think, 'Wow, my campaigns are performing way better than I thought,' and they start increasing spend without checking whether the attribution window is inflated. Server-side tagging can also reduce the delay in attribution, which might make campaigns look more effective than they actually are. Luna: So you still need to validate with incrementality testing or geo lift studies.
Lucas: Absolutely. Server-side tagging gives you cleaner data, but it does not tell you causality. You still need experiments to understand whether the ad actually caused the sale, or whether the sale would have happened anyway. But at least now your data is not corrupted by ad blockers and browser policies.
Luna: One thing I find interesting is that server-side tagging also changes how you think about view-through attribution. Because with client-side tags, view-through conversions are often undercounted due to tracking blockers. With server-side, you might see a jump in view-through conversions, but that could be misleading if you do not set proper lookback windows. Lucas: Great point.
The cleaner data can actually create a new bias. If you suddenly see more view-through conversions, you might be tempted to give them more credit. But the view-through window itself - 1 day, 7 days, 30 days - is still an arbitrary choice. Server-side tagging does not fix that.
It just removes the technology-based undercounting. Luna: So really, server-side tagging is the foundation. Then you layer proper attribution modeling and experimentation on top. Lucas: Exactly.
And it is becoming table stakes. Apple's privacy changes are not going away. Google is phasing out third-party cookies in Chrome. Ad block usage is growing - it is now around 40 percent of web users globally.
If you are relying on client-side tags, you are making decisions on incomplete data. Server-side tagging is the fix that lets you trust the numbers you are looking at. Luna: It really does tie back to something we have talked about before - that marketing measurement is as much about infrastructure as it is about models. Lucas: It is.
And if these conversations have sparked something you have actually used - maybe a new approach to attribution, or a decision to finally move to server-side - a couple of dollars a month is genuinely what keeps these going. Buy me a coffee dot com slash fexingo, if you have gotten something out of them. Luna: Yeah, it makes a real difference. We keep these ad-free because of that kind of support.
Lucas: And we appreciate it. So back to the practical side - if you are considering server-side tagging, where do you start? The easiest entry point is Google Tag Manager's server-side container. Google has a pretty good step-by-step guide.
You will need a cloud hosting account, but you can start with a small instance. The key is to map out every tag you currently have on your site - Facebook, Google Ads, LinkedIn, Twitter, any ad platform - and decide which ones to move first. Luna: Would you recommend moving all tags at once, or testing one at a time? Lucas: Test one at a time.
Start with the tag that drives the most attribution - usually Facebook or Google Ads. Run it in parallel with the client-side tag for a week, compare the data, and make sure everything looks right. Then move the rest. Also, set up monitoring alerts for your server.
If it goes down, you want to know immediately. Luna: And do not forget to update your consent management platform. Server-side tagging can still respect user consent; you just need to configure it properly. Lucas: Yes, absolutely.
Do not bypass consent. That would be a legal disaster. The beauty of server-side is that you can enforce consent on your own server before any data leaves your control. Luna: It sounds like server-side tagging is not just a technical SEO or IT project.
It is a core marketing analytics initiative. Lucas: It really is. And the companies that adopt it early will have a measurement advantage for the next few years. As the tracking landscape gets more restricted, having a clean, server-side data pipeline will become a competitive differentiator.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
How Kubernetes Topology Spread Constraints Create Scheduling HotspotsDevOps Daily with Fexingo · features Luna95 / 100
Why Pipeline Velocity Trumps Deal Size Every TimeThe Growth Operator with Fexingo · features Luna95 / 100
Why Enterprise Software Deals Now Include a Vendor AI Model Explainability MandateB2B SaaS Talks with Fexingo · features Luna94 / 100
How B2B Brands Wreck Pipeline with Unsyncroned CRM DataThe Marketing Operator Podcast with Fexingo · features Luna92 / 100
Why API Webhook Payloads Should Be Signed Not VerifiedThe Developer Tools Podcast with Fexingo · features Luna90 / 100
How to Sell Against a Competitor Already in the BuildingSales Leadership with Fexingo · features Luna85 / 100
Why Attribution Models Ignore Missed Calls79 / 100How Incrementality Reveals True Marketing Impact90 / 100How Marketing Attribution Fails on Subscription Models82 / 100Why Marketing Attribution Misses the Seasonality Pattern91 / 100
Why Marketing Attribution Breaks on Subscription Models92 / 100