GA4 vs server-side tracking. What you actually need in 2026.

So what do you actually need?

Both, and a bit more. Let me take them in turn, because the honest answer is not "buy the fancy one".

Is GA4 still good enough?

For most people, yes, as long as you use it for the right thing. Ever since privacy regulation tightened worldwide, and since Apple restricted tracking years before that, what GA4 can capture has genuinely changed. It will never again be a perfect, every-user record.

But it still tells you plenty, if you read it as a compass rather than a courtroom.

"GA4 is a compass, not a courtroom. Read it for the trend. Stop trying to make every single number perfect."

I do not obsess over individual figures. I watch whether organic is growing, whether social is growing, whether search and paid traffic are heading the right way. There is always going to be a big "direct" bucket that is really a mix of things, and that is fine. If the trends are up and to the right on the channels I am pushing, I know it is working, and I go elsewhere for the numbers I need to be precise about. The things I do want accurate are lead generation and where those leads genuinely came from.

What does server-side tracking actually do?

You asked me to expand on this, so here is the plain version.

Normal tracking is client-side: little tags run in the visitor's browser and fire data straight off to GA4, Google Ads, Meta and so on. The trouble is that ad-blockers, Safari's cookie restrictions and browser privacy settings quietly eat a chunk of that, often anywhere from a tenth to half of it, before it arrives.

Server-side tracking puts your own server in the middle. The browser sends its data to a server you control, and that server decides what to forward on to each platform. Because those onward calls happen server to server, far less gets blocked, your first-party cookies last longer and are more durable, your page can load a little lighter, and you control exactly what leaves and where it goes. In short: more complete, more durable, more vivid data on what people did and where they came from.

Two honest caveats. First, it is not a consent workaround, a myth worth killing. You still need proper consent under UK GDPR, and Google's Consent Mode is still mandatory for UK and EU visitors regardless of where your tags run. Second, the "cookie apocalypse" it was sold to survive did not fully arrive: Google reversed course and kept third-party cookies in Chrome. So server-side is now less about survival and more about accuracy and control.

Who actually needs it?

Honestly, it is overkill for most companies. It earns its place when you have got serious volume, a lot of monthly events, meaningful ad spend that depends on accurate conversion data, a regulated industry, or an audience heavy on Safari and ad-blockers. Running it costs relatively little in hosting terms, from around the price of a couple of coffees a month on managed services up to modest cloud costs, but the real cost is the engineering complexity of setting it up and keeping it healthy. For a smaller B2B site with modest ad spend, GA4 read sensibly will do the job.

What tracking actually changes decisions?

Here is the part I care about most, and it is neither GA4 nor server-side. It is influence tracking.

There are tools now that sync with your CRM and tell you every person who has been shown your ads, say on LinkedIn. That changes everything about how you value marketing. Picture the classic case: a lead comes in, books a call with sales directly, and tells the rep "you're the first person I've spoken to". That goes down as a pure sales win, and marketing gets nothing.

Except the influence data shows that same contact has spent three months being shown your thought leadership and a few retargeting ads, and actually engaged with one of your posts eight weeks ago. They knew you long before they "found" you.

"Marketing's job isn't to claim the win. It's to show it helped the win happen."

This is not about scoring points off sales. It is all one commercial pipeline. Marketing is just the earlier portion of the same sale. The reason to track influence is to make investment decisions: if your LinkedIn thought-leadership campaigns keep showing up in the deals that close, they are worth it, keep investing. If you pour money in for months and it never appears in a single closed deal, pull back. You will never trace every pound to a pound of revenue, and you should not try, if you did, brand awareness would never get funded and every business would be poorer for it. But you can see influence clearly enough to decide where the next pound goes. If you want that influence layer wired into your CRM, no lock-in, request a quote.