I'm from a marketing background. I work quite a lot with online marketing teams.
And the question "do we need to go server-side" comes quite often.
There's a hype to it combined with a lack of knowledge what it is useful for which leads to confusion and mistakes.
My goal is not to explain how technically server-side works. Many have dont that very well LINK LINK
My goal is to bring clarity what you get, the impact for your marketing team and a bit broader for your company, so that you can make the right move and understand better if it's worth the investment.
3 reasons to implement server-side tracking
1. Reduce impact of ad blockers
I start with this one as you - performance marketing - will directly be impacted by this one. But even if the uplift might not be negligible, the 2 other reasons described below have a bigger range.
You certainly saw Facebook pushing for its Facebook Conversion API. And maybe yourself want to push for the integration of Facebook CAPI in order to get more conversions.
Expect an uplift from 5% up to 30% (but more generally around 10-15% I'd say), which is not nothing. Facebook CAPI does overall many things but the one thing that will impact the conversions is that it will reduce the impact of ad blockers.
Ad blockers prevent the Facebook pixel (and any pixels) to run. So your marketing platform won't record any conversions.
A server-side solution such as Facebook Conversion API can solve this. And your marketing platform start tracking users and their conversions on your website.
Is Facebook CAPI the only thing you need
As Meta is the biggest advertising platform out there (outside Google which works a bit differently), there are many solutions out there that implement only Facebook CAPI.
If you rely solely on Meta Ads, this could look like a good and cheap solution. For quick and short-term effects, it could indeed suit your needs if you're only interested in a conversion uplift.
However I would give you 3 reasons to think about it carefully:
- What about the other platforms? You have here a solution just for Meta
- Tracking technologies keep evolving and using a server-side service to solely work around ad blockers might become useless in the future when the tech will adapt. (Because don't forget that these users have ad blockers for a reason....)
- You might be interested short-term on a conversion uplift. But server-side could bring more on the table than that. And this kind of offers don't provide this to you and your company.
So I would always recommend setting up a complete server-side structure, also because the following 2 points are quite exciting.
2. Increase website load speed
If you advertise on many marketing platforms, you probably integrated them and added their tracking technology to your website. So your website communicates with their servers, sends information, receives some, etc.
This needs to happen as quickly as possible, ideally before the document (the page) loads as we're talking about tracking here.
You can test it yourself on your website with the developer tools of your browser, it takes time. And the more platforms you use, the longer it is.
One big advantage with server-side is that you stop communicating with all these different platforms but you focus on one: your own server.
The impact is that your page will load faster. Which is a great factor for the success of your campaigns: A report from Deloitte showing that websites with a 0.1s improvement in site speed record 10% more conversions.
3. Control the data you send
This one might not have a direct impact for your performance marketing campaigns but this is a very important topic: by sending information first to your server, you own the data and you can control what you want to send to the vendors.
I'm not going to harass you with Data Governance, but I'm sure your Data team understands the potential of server-side tracking in this regard.
Main advantage: You can make sure that you don't send any PII (Personal Identifiable Information) to your marketing platforms, which is prohibited without explicit consent. And this is not the cookie banner.
NOTE: IP Adress is considered PII. Hashed pseudonymous email is considered PII.
How does server-side tracking with Google Tag Manager work?
Concretely if you implement a server-side setup via Google Tag Manager, you will create a new "server" container in addition to your (probably already existing) "web" container.
The web container will read what's going on on the website and instead of sending a tag to a vendor, it will send to your server.
In the "server" GTM container preview, you see therefore the data you receive on your server.
From there you can decide which data you send to your different platforms, whether it is Google Analytics 4, Google Ads, Meta, etc.
You can delete information (such as PII) overall, just for one platform, just for one tag.
You can also enrich it with information from other sources:
- Send your CRM data to Meta for example: if you have offsite conversions (eg for lead gen), send this info from your CRM to your server and send an event to Meta.
- Enrich/enhance the information of a new lead with information from external sources, calculate a lead value and pass it as a conversion value to your marketing platforms.
So yeah, this one might be for the nerdier of yours. But this is the BIG advantage of a server-side implementation for you Marketer but also for your company.
What is NOT server-side tracking:
There is maybe a term that I haven't written yet and that you were expecting: third-party cookies
This is exactly what I meant with confusion earlier in my introduction
"We need server-side tracking because third-party cookies are dying"
Third-party cookies are being blocked more and more, this is true. Even Chrome starts doing it.
But marketing platforms haven't waited to adapt and most of their cookies are first-party already. An example is the Google Conversion Linker that you (hopefully) have in your Google Tag Manager "Web" container.
So server-side tracking won't change much in this regard.
You might see out there that you can extend the cookie expiration. It is unfortunately less true since last year and the update from WebKit.
In a nutshell: if your server-side tracking server runs on a third-party cloud service (such as Google Cloud Platform), the first half of your server IP won't match the first half of your website. WebKit will therefore limit the cookie lifetime to a maximum of 7 days.
"We'll track more users because we don't need to get their consent"
Nah this is completely wrong.
Users' consent has nothing to do with the technology you use to track them.
You need their consent for cookies you set regardless if in a third-party or a first-party context.
And, of course, if you want to respect GDPR and your users' consent, you will need to configure your tags in your "server" GTM container according to their choices.
