Recently Google has brought about a huge leap in the way digital marketers approach site tracking in 2021. Google Tag Manager Server-side offers a powerful alternative to regular tag management, but what the hell is it, and why should Digital Marketers even care? If you were to ask Google, they’ll probably hand you a complex flowchart with some technical jargon about payloads, endpoints, and servers. So I’m going to explain this to people who don’t want that.

Sounds Complicated.. Why Should I Care?

It’s easy to jump to the conclusion that this is all too technical and new to affect me, my clients, or my business, well you’d be wrong. Server-side does bring a few hurdles, but the positive impact far outweighs the drawbacks. It not only offers ultimate control over the data you feed hungry marketing platforms like Facebook and Bing, but can significantly increase site speed too, and in turn, SEO performance. A complete list of the benefits of server-side would include:

• Enhancing your data accuracy
• Full ownership of your data as you decide what third parties receive
• Speeding up your website loading times
• Reducing potential privacy and security risks

How It Currently Works

Let’s imagine your website that sells bikes is a shop on the high street. To sell as many bikes as possible, you hire some sales executives, these represent the usual marketing channels like Facebook Ads, Google Ads, Bing Ads etc. To measure their effectiveness each wants to know how many bikes they sold.

To record their sales, you hire a handy cashier, they’ll act as our regular Google Tag Manager set up.

Each sales exec tells the cashier they want full access to the till, so they can write their own receipt for every bike sold. This way, they can be as accurate as possible in measuring their sales. Now for every customer who buys a bike, the cashier must call over each manager to write their own receipt. He then writes one for his own till record and finally one for the customer to leave the store bike in hand. You can imagine how this becomes a slow, frustrating process for the customer.

This is the process happening on most eCommerce sites using a standard Google Tag Manager implementation. For every page you load in your browser, Google Tag Manager is loaded in too, acting as access for marketing platforms like Facebook to add their tracking scripts onto the page to write and send network requests back to their own database, each request requiring the time and energy from the customer’s device. This happens on every page and continues to occur for each action you take, like clicking buttons, completing forms, and making a transaction. Here’s a diagram I whipped up.

If you’ve ever run a page speed report, you’ll know that these marketing platforms scripts will be at the top of the list for long waiting times. Especially on something like a mobile device with not much computing power. So what’s the solution?

How Serverside Changes Things

Let’s imagine the same scenario, but instead of just having a cashier, we also hire them an assistant, they will represent a server, which in reality is just a computer in a cold dark room somewhere. This assistant is given the sole task of taking the cashier’s receipt to write and distribute custom copies for each sales executive. Now the cashier doesn’t have to give everyone full access to their till, and the executives get the information they need from the assistant. They aren’t having to be called over each time and write their own, they are simply hand-delivered their custom copy of the first receipt.

This is where the efficiencies of adding an assistant for the cashier become apparent; it lightens the load on the cashier, making the buying process much faster. It also gives them complete control over the data these sale executives are given, and in turn more privacy for their customers.

There are some downsides to this; customers may want to know what their receipts are used for and might ask to opt-out of contributing to the sales executive performance. With the assistant handling the distribution of receipts, the customer can’t see if sales are taking a copy, or stop them from getting it. Finally, let’s not forget about the additional cost of hiring an assistant in the first place.

This approach on an actual website is essentially taking a lot of the heavy lifting away from the page. Using a server that the website owner maintains to assist with distributing tracking information to these marketing platforms.

So instead of loading every tracking script on the page for a transaction, and then wasting resources sending the same bits of information like revenue and quantity to each platform individually. The page sends the information once to the server which is running its own special version of Google Tag Manager. When this info reaches the server, it can chop up and reuse it for each marketing platform, with its own energy and time, instead of the user’s device. This means the original page can focus on loading functional and visual elements, instead of getting held up with scripts, increasing the speed it loads, and our SEO performance. Here’s another fun diagram.

The downsides in reality are interesting, once this information is sent to our server, the customer on the page has no view of where it’s distributed. A process like this can easily avoid cookie blockers and ad blockers, the information sent to a server that doesn’t have something like ‘facebook.com’ in its address (this could now be anything), wouldn’t be easily recognisable as a road to Facebook. It puts a lot of responsibility in the hands of the site owner to not do anything that might be regarded as intrusive, or unethical in regards to their user’s privacy. The other put off for some is that maintaining a server does come with costs, although for a small site with low traffic this would be nominal, it’s when you get into high volume traffic that this could quickly become expensive.

So What Now?

If you desire to implement serverside on your website today, I would seek proper guidance from an agency or marketer with experience in this. You would also do well first to run this in parallel to your current solution. It’s still very much a product in its infancy with many unknowns, and as it stands, it is still missing many features you get through standard GTM.

Hopefully, you now have a better idea of how this whole tracking thing works, both in and out of the server. Appreciate it if you’ve made it this far. If you have any feedback or burning questions on server-side tracking, you can reach me on my Linkedin or Twitter, or shoot me an email at richard@bedrock42.com. If you want to see more articles like this, you can sign up to our Newsletter below!