Skip to main content

Menu

Part of: Data and analytics

Server-side tracking and GTM

Server-side tracking sends some of your tracking data through a server you control, so less is lost to browsers and blockers. It can improve your data, but it adds running costs and upkeep, and it does not remove the need for visitor consent.

Is this right for you?

Who it is for

Stores with meaningful ad spend whose tracking has gaps that make decisions harder, and whose basic GA4 and conversion tracking are already sound.

When it is not

Stores with low ad spend, stores whose basic tracking is still wrong, and anyone hoping to collect data without visitor consent. It will not do that, and we would not set it up to try.

What is included

  • A check that your basic tracking is sound before anything moves to a server
  • A plan for the server container, its hosting and who looks after it
  • Setup of the server container on a domain of your own
  • Your Shopify store connected through its customer events and pixels
  • Consent choices passed through, so visitor decisions are respected
  • Testing against your real orders, with the remaining gap written down
  • A handover note on monthly costs and what to check if something breaks

What it does, and what it does not

Normally, tags run in the visitor’s browser and send data straight to each platform. With a server container, the browser sends data to your server first, and the server passes it on. That can make tracking more resilient, and it gives you control over which details leave your site.

It does not make consent optional. If a visitor declines tracking, that choice still applies and the server must respect it. It also does not give you perfect data. Some visitors will still be missed, and some difference between your store, GA4 and your ad platforms will remain.

The tags on your site send data to a server container that runs on a subdomain of your own, such as data.yourstore.co.uk. It then passes the data on to Google and other platforms. Because it runs on your domain, browsers treat it as part of your site, though they still limit how long some cookies last.

Complexity and running costs

We would rather say this plainly than leave it out:

  • It is more to look after. There is a server to host, a domain to configure and another set of tags to maintain. When something changes on your store, the server side may need updating too.
  • Hosting is an ongoing cost. It is paid to the hosting provider, separate from our fees, and it usually rises with traffic.
  • It can fail quietly. If the server stops, data stops, and nobody gets an error. Monitoring needs to be agreed from the start.

The container can run on Google Cloud or with a specialist host, and the cost depends on your traffic. We give you the likely cost before you commit.

On Shopify, storefront and checkout tracking runs through the platform’s customer events and pixels, rather than code added to checkout pages. That shapes how events reach a server container, and what is available at checkout can depend on your plan and setup.

Consent has to be passed through too, so the server only sends what the visitor has agreed to. Google’s consent mode is one way to do that.

On Shopify, tracking runs through customer events: app pixels and custom pixels that run in a sandbox, on the storefront and through checkout. That includes the checkout events, and the pixels respect the visitor’s consent settings.

For visitors in the UK and EEA, Google needs consent mode, with its signals for analytics, ads, ad data and personalisation. The server container receives those signals with each event and passes them on.

What we measure

We measure how closely tracked orders match your real orders, before and after, and write down the gap that remains. We do not promise a particular improvement, because it depends on your visitors, your consent rate and your setup.

Where this connects to other work

Better conversion data helps Google Ads learn from the sales that really happened. The same data is what attribution and ROI dashboards rely on. If you are not sure whether you need this at all, get in touch and we will say.

How we do it

Every project is different, so timings depend on its size and complexity. You get a realistic timeline in writing before work starts, and we never promise results by a date.

  1. Step 1: Check the basics

    We compare tracked orders with real orders. If the basics are wrong we fix them first, because moving bad data to a server only gives you bad data.

  2. Step 2: Decide whether it is worth it

    We look at your ad spend, your gaps and the running costs, and tell you plainly if it is not worth doing yet.

  3. Step 3: Plan and set up

    We agree the hosting, the domain and who owns each account, then build the server container and connect your store.

  4. Step 4: Test with consent in mind

    We test with consent granted and declined, and compare the results with your orders.

  5. Step 5: Hand over

    You get a written note on what was built, what it costs each month and how to tell if it has stopped working.

How every project runs

What it costs

Retainers start from £500 per month and cover our services only. They do not include your ad spend, or any software or other subscription costs. One-off optimisation projects are priced by current need or an agreed hourly rate. Our pricing page has the detail.

See how we price our work

Questions people ask

Is server-side tracking worth it for my store?

It depends on your ad spend and how big your tracking gaps are. With meaningful spend and clear gaps it can be worth it. For a small store, or one whose basic tracking is still wrong, the running costs and upkeep usually outweigh the benefit. We will tell you which applies before any setup work begins.

Does server-side tracking get around consent?

No. Visitors still need to be given a choice where the law requires it, and their choice still applies. Moving tags to a server changes where data is processed, not whether you need permission to collect it. We set it up to respect consent. We are not lawyers, so take legal advice if you are unsure. The visitor's consent choice is passed from the website to the server container, which acts on it.

What does it cost to run?

There is a setup cost, then an ongoing hosting cost for the server container, which usually rises with traffic. Someone also has to look after it.

Does it replace GA4?

No. It changes how data reaches GA4 and your ad platforms. GA4 is still where you analyse it. If GA4 itself is set up badly a server container will not fix that, which is why we often start with a GA4 audit.

What happens if the server stops working?

Data stops arriving until it is fixed, and you may not notice straight away. That is why the handover note covers how to check it is running, and why we agree who is responsible for monitoring before we build it.

Will it close the gap between Shopify and GA4?

It may narrow it, but it will not close it. Some difference is normal and comes from consent, attribution rules and how each system counts. We write down the gap that remains so you know what to expect.

Ready to talk it through?

Get a free, top-level audit of your website for SEO, your ad account for PPC or your analytics account. A subject matter expert at Advent reviews it, and we aim to send it within 5 working days of receiving access.

Request an audit