Three popular GTM server-side tagging misconceptions
With the recent release of server-side tagging for Google Tag Manager, a whole new woooooorld of possibilities has opened up. But on that, some people have got the wrong idea, so with that in mind, here’s three popular (but wrong!) things people believe about server-side GTM.
1) You still need client-side GTM to make it work
This is slightly unfair, as I can see why this misconception arose. You do need something to send your hits to the GTM server, and in many cases the easiest and best thing to use will be the already-existing client-side GTM. A lot of the examples you see will do this, and in a lot of real cases, it will be the best and most sensible route to take. However, there is no necessity for it. If you wanted to remove GTM entirely, you could have your developers insert calls directly to GTM rather than data layer pushes, for instance. It can be treated more like Google Analytics’ Measurement Protocol—or, for that matter, the prelapsarian state before man ate the fruit of the knowledge of GTM and event calls had to be delicately hardcoded onto the page individually.
As I say: I suspect in a lot of cases you will want to keep client-side GTM around to send the hits to the server, but it’s not strictly speaking necessary.
2) It’s a 1:1 replacement for everything
While some tag providers will be largely unaffected by having their tracking run through SST, there are others whose tracking ability depends on their code actually being on the page. Most heatmapping—think Hotjar, SessionCam, CrazyEgg—or ‘track first, classify later’ solutions like Heap, FullStory or Snowplow—require their code to be on the page. They can’t hoover up and send off every little thing the user’s doing from the server like that—and if you were to send everything in that way, it would obviate some of the benefit of moving to server-side tagging in the first place. The bottom line, I suppose is—if you’re heavily dependent on those tools, maybe SST isn’t for you; or at least, maybe you want to keep client-side GTM around too.
3) You don’t need to worry about user tracking preferences
There’s been some beef lately, though thankfully not between Brian Clifton and Simo Ahava (very much the Biggie and Tupac of the digital analytics world). The beef has been between people who see the introduction of server-side GTM as an excuse to bypass people’s express preferences with regards to tracking and those of us on the side of the law and God. Server-side GTM is a brilliant new tool that will allow for lighter page weight, fewer calls to external sources and better security around things like content security policies. It is not an excuse to ignore people’s choice to not be tracked. Any legal concerns aside, it’s just not a good thing to be doing. In the words of our Lord and Saviour Simo: “I hope people who think like this will take a moment to consider the ethical ramifications of what they’re doing”.