Server-Side Tagging is geïnstalleerd, wat nu?

Once your server-side tagging setup has gone live, you'll naturally want to know if everything is working correctly. You've invested in a first-party tracking solution and expect more accurate measurements and better insights. But how do you know if your setup is functioning correctly?

In this article, we'll guide you step-by-step through all the essential checks you need to perform after your Server-Side Tagging implementation. We'll focus on practical checks you can perform immediately within your main platforms: Google Analytics 4, Meta Ads, and Google Ads. Additionally, we'll cover the most common problems and their solutions.

What you'll find in this article:

• What checks you need to perform in GA4, Meta Ads and Google Ads
• How to solve common problems (such as high Unassigned traffic)
• What to expect in terms of measurement accuracy

This article is intended as a reference work.

Server-Side Tagging

The visuals below explain as simply as possible how events in marketing and analytical channels end up based on the site interactions a visitor has. With this, we try to make it as clear as possible what a Server-Side Tagging setup entails compared to Third-Party Pixels or Client-Side Tagging.

Third-party pixels can be placed on your site, giving third parties access to create their own events based on site interactions:

With Client-Side Tagging, you place a Google Tag Manager script on your site so that you can specify within a Google Tag Manager web container which events may be created based on site interactions. These events are sent directly from the web container, from the user's browser, to the platforms.

With Server-Side Tagging, events are still created in a Google Tag Manager web container, but they are sent to a server linked to the main domain. A Google Tag Manager server container can then check which events are arriving at the server, and thanks to the Google Tag Manager server container, the events are forwarded to the correct platforms.

The difference compared to the old situation? With third-party pixels, you gave external parties direct access to your visitors. With server-side tagging, your browser sent data directly to those platforms. Now, as a first-party, you are in control and can precisely determine what happens with the data.

Parts to check after initial implementation

Google Analytics 4

Google Analytics 4 is the central hub for data analysis for most websites. After your server-side tagging implementation, there are three crucial checks you need to perform.

Please bear in mind that GA4 operates with a data processing period of 24 to 48 hours. Therefore, yesterday's data will never be fully processed. Always check with certainty for periods up to and including the day before yesterday.

1. Check if your conversion events are being received

Go to GA4 Reports > Engagement > Events and find your most important conversion event (for example, ‘purchase’ for webshops or ‘generate_lead’ for leadgen sites). Check if these events are being received since the implementation of Server-Side Tagging.

Please note: A small margin of error between measured and actual conversions is normal. Without webhooks, 100% accuracy is not achievable due to ad blockers, cookie refusal and technical limitations. The important thing is that the events are recorded at all, not that there is a perfect 1-to-1 match with your backend.

2. Check the percentage of ‘Unassigned’ conversions

Open in GA4 het Events report, select your conversion event, and add the dimension ‘Standard Channel Group for Session’ to the blue plus sign.

Ideal Fewer than 151 TP3T Unassigned conversions. A high percentage indicates tracking issues that require further investigation (see the “Troubleshooting” section below).

3. Analyse the percentage of ‘Unassigned’ traffic

Go to Report > Lifecycle > Acquisition > Traffic Acquisition. Ensure your date range starts from the day after your Server-Side Tagging implementation.

Aim for: less than 15% Unassigned traffic.

Meta Ads

Meta Ads has been receiving events via your server container since your Server-Side Tagging implementation. Here are the key checks:

1. Check alerts in Meta Events Manager

Open business.facebook.com and go to Event management through the menu.

Please note the warning: “Custom events cannot be used in conjunction with ad units.” You need to resolve this by confirming the unknown events (if they are from your tracking) or rejecting them (if they are from elsewhere).

2. Check if the Meta Events Tool is still active

Check in your Meta Events Manager if there are any events still set up in Event Tool Configuration. If these exist, remove them. You are now sending events via server-side tagging and do not want duplicate measurements. Follow this action plan to perform this check.

3. Understand low quality scores

Meta automatically assigns quality scores to events. These will almost never be close to 10/10. These quality scores are often lower because parameters such as email address, country, first name, etc. are not sent with every event.

Meta Ads will also indicate that you need to send an fbc parameter more often. The fbc parameter is the Facebook Click ID parameter. The only way to send this more often is to increase the percentage of visitors to your site who have clicked on a Meta Ads advertisement. You only increase this by advertising more via their platform, and thus spending more budget on Meta.

Important: These scores have no impact on your campaigns. These are automatic alerts with no practical impact. You do not need to do anything with them.

Google Ads

Google Ads has created new conversion actions alongside your existing ones after your Server-Side Tagging implementation. Here's what you need to pay attention to:

1. Check the status of your conversion actions

Go to Google Ads Goals > Conversions and look at the status column. There are 4 possible statuses:

• ActiveConversions are being received and assigned to campaigns ✅
• InactiveA conversion has not yet been received; the setup may be incorrect.
• No Recent ConversionsEvents are received, but they cannot be assigned to campaigns.
• Needs AttentionThere is a problem with consent mode, enhanced conversions, or data formatting

Only ‘Active’ is what you want to see. Other statuses require action.

⚠️ If your conversion action is showing the status Needs Attention due to Cart Data issues, it means the product IDs in your Google Merchant Centre are not the same as on your webshop's frontend. Would you like us to help with this? Please submit a support request via: https://help.adpage.io/nl/contact-us

2. Determine which conversion actions are driving your bidding strategy

Following the implementation of Server-Side Tagging, new conversion actions have been created which are set to "Secondary" by default. Secondary conversions are reported, but they do not inform your bidding strategy. Your responsibility is to analyse which conversion action (Server-Side, Client-Side, or GA4 import) provides the most accurate measurements and set this as the "Primary" for your bidding strategy. The Server-Side conversion is not automatically the best; this depends on Google's modelling and how well different conversions can apply attribution. For more information, please refer to this blog article: Which Google Ads conversion action do you set as primary?

Other platforms

Do you also have tracking set up for TikTok Ads, LinkedIn Ads or Pinterest Ads? Then check there too to see if events are coming in:

TikTok Ads: Open je Pixel via Tools > Events Check the Event Health for Complete Payment and the number of events received per domain. Ensure there are no inactive events.

LinkedIn Ads Go to Track Conversions Check that your conversions have the status ‘Active’ (not ‘Inactive’ or ‘Not verified’).

Pinterest Ads: Open it Event Summary and verify that your Conversion API events are being received and the numbers reasonably match what you're seeing in GA4.

Common problems (and how to solve them)

Problem 1: High percentage of ‘Unassigned’ traffic in GA4

A high percentage of ‘Unassigned’ traffic (above 15%) in GA4 is a common issue following a server-side tagging implementation. Here are the steps to resolve this:

Step 1: Determine if Unassigned is truly unassignable

Open het Traffic acquisition report, add the dimension ‘Session source/medium’ And filter on ‘unassigned’.

What do you see?

• Unrecognised sources/mediums (such as ‘email/unknown’ or rare combinations) → This is a UTM parameter issue. You have used the wrong or missing UTM parameters in your campaigns.
• Only ‘(not set)’ This is truly non-attributable traffic, often due to cookie refusal or privacy browsers.

Step 2: Check your cookie banner and consent processing

A misconfigured cookie banner or incorrect consent processing in Google Tag Manager can lead to high unassigned traffic. Visitors must give consent before tracking begins, and if this is set up incorrectly, events cannot be linked to their original source.

Read us Article on setting up cookie banners via GTM for the correct configuration with Server-Side Tagging.

Optimise your Consent Management Platform

To minimize the impact of cookie refusal:

• Make your cookie banner unignorableVisitors must explicitly opt in before they can use the site
• Show on the first screen Reject all“ knop – dit verhoogt het aantal cookieweigeringen significant
• Accept and customise preferences“

This approach is GDPR-proof and minimises the number of visitors who refuse all cookies. How to do this with Cookiebot or Consent Studio can be found in this helpdesk article: The ideal cookie banner setup.

Problem 2: Advertising platforms claim different conversions than GA4

This is normal and unavoidable. Each platform (GA4, Meta Ads, Google Ads) uses its own attribution models and attribution windows. A conversion can be claimed by multiple platforms because they use different rules for assignment.

Example: A visitor arrives on your site on Monday via Meta Ads, on Tuesday they click on a Google Ads campaign after Googling your brand, and on Wednesday they return directly and convert.

• Meta Ads claims the conversion (visitor first came via their campaign)
• Google Ads claims the conversion (last click before conversion was via their campaign)
• GA4 registers the conversion as ‘Direct’ (the session in which conversion occurred was direct traffic)

All three are right in their own way. Discrepancies between platforms are inevitable – focus on trends and relative performance, not exact numbers.

Do check your attribution model in Google Ads

Go to Google Ads Goals, open a conversion and check at Settings the Attribution model:

• Last ClickConversion goes to last click before purchase (traditional, simple)
• Data-drivenMachine learning analyses your historical data to determine the value of each touchpoint (recommended, but requires sufficient conversions)

The data-driven model generally provides a more realistic view of how different channels contribute to conversions.

Understanding Consent Modelling

Google vult ontbrekende conversiedata aan van gebruikers die geen tracking-toestemming gaven via modellering. Dit helpt je prestaties nauwkeuriger in te schatten, maar kan wel kleine afwijkingen veroorzaken tussen conversiedoelen binnen Google Ads zelf.

This is a feature, not a bug. It compensates for the loss of data from cookie refusal.

More information Google Consent Modelling documentation.

Problem 3: Margin of error between measured and actual conversions

Even with server-side tagging, you will never see 100% of your conversions tracked in your analytics and marketing platforms. This is because:

• Ad blockers and privacy browsers which also block first-party scripts
• Cookie refusal via your consent banner
• Script errors and connectivity issues
• Abandoned during the loading process (visitor closes tab before scripts have loaded)

Accept that a small margin of error is normal. Without a backendwebhooks (where your backend reports conversions directly) perfect accuracy is technically impossible.

What can you do?

1. Optimise your CMP as described above to minimise cookie rejection
2. Set alerts on the trytagging environment to automatically receive an email when a status change is detected with your server or tagging pixel
3. Monitor trends rather than absolute numbers – if your margin of error is consistently 5%, you can take that into account

Focus on improvements in conversion rate and trends over time, not on a perfect 1-to-1 match between your backend and your tracking platforms.

Maintenance and monitoring

Your server-side tagging setup is not “set and forget”. Regular maintenance is essential to maintain data quality:

Set up alerts for automatic monitoring

On the trytagging environment, you can set up alerts to automatically receive an email when a status change is detected with your server or tagging pixel. This helps you to react quickly if something goes wrong.

Check your key metrics regularly

Make it a habit to check monthly:

• Percentage of unassigned traffic in GA4 (must remain below 15%)
• Status of conversion actions on all platforms (must remain ‘Active’)
• Warnings in Meta Event Manager and Google Ads

Be alert to changes by others

Many different people may have access to your website, Google Tag Manager or platform settings. A change by someone else could unintentionally affect your tracking. Regular checks and alerts will prevent problems from going unnoticed for too long.

Conclusion: Focus on what matters

Server-Side Tagging gives you more control and accurate measurements, but it also requires more technical attention than old-fashioned third-party tracking. The checks in this article will help you to:

1. Direct after implementation to verify that everything works
2. Identifying problems quickly before they have an impact
3. Realistic expectations to talk about measurement accuracy

Remember that small margins of error are normal and acceptable when you are not using backend webhooks and are dealing with consent processing via a cookie banner, that discrepancies between your marketing platforms' reports are inevitable, and that you should primarily focus on trends and relative performance rather than absolute perfection.

Do you have problems that aren't covered in this article? See our helpdesk for further troubleshooting Please get in touch for personal support.