top of page
Writer's pictureJulien De Visscher

The difference between server-side tracking and server-side tagging

For those who have been around in the analytics & MarTech scene, server-side tracking is nothing new. For years if not decades various technologies have been able to send and receive data through their servers. The list of pros & cons is pretty long. From a Human37 perspective, Glenn Vanderlinden, published the following article on the topic in 2021: “What is server-side (or server-to-server) tracking ?


In this article we aim to look at server-side tracking from a marketing perspective. Indeed we’ve observed, over the last months, in most marketing discussions, a clear uptick in interest for everything server-side. Our objective is to share why we believe it is such a hot topic, demystify what it helps marketers achieve, highlight the different paths to get there and share our opinion on the topic.  


Why is server-side tracking so popular?  


First let’s look at what most marketers chase when considering server-side tracking. The graph below is a good summary of what it is all about: 

In a nutshell, marketers are trying to workaround some of the limitations that appeared in the last years and are leading to data collection limitations. Namely: 


  • Technical browsers limitations (ITP as an example)

  • Adblockers (37% of global internet users block tracking technologies) 

  • The announced death of 3rd party cookies (by Google)


Despite what some marketers believe, server-side tracking should at no point be a way to collect data without consent. Consent is a universal concept that applies whatever the technique used to collect and transfer data from point A to B. This means if a user refuses to give his/her (cookie) consent, data cannot be tracked. Even through server-side solutions (unless any of the other principles for lawful processing in article 6 can be applied).


Now that we understand what marketers are chasing when talking about server-side, let’s look into why it has become so popular. In our opinion it comes down to two main reasons:


1.Market maturity


  1. Most companies have now experienced data loss (up to 30% - 40%). As a result, they are looking for solution(s) to get back some of that data.

  2. The technical landscape has matured and more people are at ease with the concept of server-side. This is leading to a democratisation of the implementation. As an example, we’ve witnessed the rise of tools specialised deploying server-side tracking (ie. MetaRouter, Addingwell, Stape)


2.Market standards


  1. Certain use cases that can only be solved using server-side tracking have now become market standards (if not obligations): Meta CAPI, Google Enhanced Conversions and other media/acquisition platforms that are looking to increase the amount of data points they receive to optimise campaigns’ performance.

  2. The importance of 1st party data collection to create audiences and communicate with identified users as a (partial) replacement of traditional remarketing techniques. 


As a marketer how can I deploy server-side tracking? 


As a data agency we are helping clients use data & technology to solve their business needs. As result we often face clients asking us to support them in deploying server side tracking. Our answer is very simple: WHY? What are you trying to achieve with server-side tracking? Depending on their answers, three main roads can be taken: 


1. Pure server-side approach


This approach is at the same time the simplest and most. Simplest because it consists in ensuring data is sent directly from the clients’ servers into the third party receiving server (Meta, Google Ads, etc.). No intermediate layer required. On the other hand, it is the most complex as it relies 100% on technical (read IT) resources. Since IT is the owner of operational systems that gathers the data we are looking to send to third party receivers, they are the one and only stakeholder that make it happen. What makes it so complex is probably the question you are asking yourself? The answer is simple: time & resources. By experience we know IT is so busy ensuring the operational systems are properly working and/or improved. Deploying technical solutions for marketing purposes is usually lower on the priority list. As a result this rarely happens. 


What are the pros & cons?

Pros

Cons

It gives access to the operational data (as opposed to client side data only) and therefore increase granularity and data completeness

Requires IT resources


Data enrichment is harder to deploy

2. The data warehouse (DWH) server-side tracking approach


The data warehouse approach consists in deploying a Data Warehouse that acts as an intermediate layer between the operational systems and the marketing solution and/or analytics in general solutions. By receiving a copy of the operational systems’ data, the data warehouse becomes the single source of truth for all data activations (audience creation, analysis, visualisation, automation, etc.). It sits between sources (transaction databases, customer support, ERP, CRM, etc.) and destinations (data visualisation, advertising and/or media platforms, engagement platforms, etc.). 


What are the pros & cons?

Pros

Cons

One single infrastructure to deploy as many (types of) use cases as possible (conversions tracking, audiences dispatching, BI & Reporting, etc.)


It requires a heavier initial investment (which is normal considering the increased expected return-on-investment)

It gives access to the operational data (as opposed to client side data only) and therefore increase granularity

It doesn’t allow sending behavioural data points* for conversions purposes in acquisition/media platforms

It gives the opportunity to enrich data

It requires data engineering technical skills

 * This is debatable as we could envision a scenario where behavioural data is available in the DWH (for instance: tracked by an analytics solution such as Mixpanel, Amplitude or GA4 and exported into the DWH).  


3. The server-side tag management system (TMS) aka server-side tagging


The server-side TMS (from which Google Tag Manager and/or MetaRouter are likely the most known) approach consists in sending client-side data in a server-side way. It is often referred as “server-side tagging with a first-party collector” and means that instead of sending data straight for the clients’ browsers, it is first sent into an instance stored in the cloud and hosted on a sub-domain of a brand (to ensure it stays 1st party data) before being dispatched to the servers of the 3rd party destinations. 


What are the pros & cons?

Pros

Cons

It requires less technical skills (IT or Engineering)

It only allows to send conversions data (micro or macro) into 3rd party solutions

It allows sending behavioural data points as conversions in acquisition/advertising platforms

It doesn’t allow sending audiences for remarketing and/or exclusion purposes for instance


Below is a summary of the differences of both solutions: 

 

Pure Server-Side 

Tracking

Server-Side 

Tracking (DWH)

Server-Side

 Tagging (TMS)

Use cases covered

 

 

 

Sending Transactions Server-side

Yes

Yes

Yes*

Sending Behavioral events Server-side (ie. click on a button)

No**

No**

Yes

Sending Audiences for other use cases (exclusion, signals, etc.)

Yes

Yes

No

Future-Proofness

Medium

High

Low

ROI

Medium

High

Low

Implementation

High

High

Medium

Maintenance

Medium 

Low 

Low


Our recommendation


As explained above, each solution has pros & cons. The reason why going for one or the other would mainly depend on what you are trying to achieve and your internal context (read resources, culture, etc.). If you are looking for a quick way to activate Facebook CAPI and all other equivalents, TMS server-side is probably the fastest way to get there. If you are looking into building a more robust infrastructure that can help you deploy Facebook CAPI alongside many other use cases, the Data Warehouse is for sure the way to go as the expected ROI will be higher. 


To add more colours to this story be aware that many companies are actually using several of those solutions in parallel. We often have clients using pure server-side tracking to send critical data to analytics, customer data and/or customer engagement platform(s) while having a data warehouse to serve more advanced use cases: audience creations, BI & reporting, etc. 


Interested? Feel free to reach out!

Comentários


bottom of page