The Future of Tag Management

Tag management is still relatively new. What was the first TMS? TagMan? It was probably another one… my memory is a bit fuzzy. Though still young, the industry has evolved through several iterations. The early tag management systems relied on templates. The language everyone spoke was in the context of these templates. “We have a template for over 300 media tools!” That actually sold products more effectively than crushing competitors in proofs of concept. Companies were looking for ways to manage their media pixels – they had dozens of them on their site concurrently! I worked on a few clients that had over 20 pixels on a single page. What question do you think those clients wanted to answer most? “How can I most effectively manage my media tags?” Templates were specifically intended to answer that question.

With the release of Google Tag Manager and the acquisition of Satellite (to become Dynamic Tag Manager), the narrative changed. The concept of templates more-or-less fell to the wayside. The TMS was mainstream. The increased exposure to these tools showed that these were more than just pixel managers. Instead, a tag management system could be an incredible powerful tool to quickly implement detailed tracking – and that became its primary function. It put more power in the analyst’s hand to change the narrative from discussing pixel placement to leveraging tags to describe user behavior.

Now where are we in the TMS space? Well, we’re developing process and standards around tag management. The clear leaders (Adobe and Google) are getting comfortable in their chairs – and rightfully so. These two are phenomenal tools – not to say some other tools aren’t (Tealium is a wonderful tool and Signal takes a really unique approach to tag management). So what’s next? … Or rather, what do I think is the next big evolution in the tag management space?

Governance will be HOT

They always say when you make lists to lead with something attention-grabbing and I open with governance (don’t you see I capitalized “HOT”?). OKAY, WAIT DON’T LEAVE! I promise I won’t make this one boring. Let me just say this… there haven’t been enough horror stories yet. I’m not talking about your paranoid developer who went through the proof of concept and was all “Yeah we’re going to lock this bitch down.” I’m talking about the moment the developer got bored of QA’ing every single release and now the analyst (or your consultant) runs the roost… and proceeds to burn it down. Do I know from personal experience? Absolutely. Was I the one who burned it down? More than once.

There’s an entire workflow in Google Tag Manager and Dynamic Tag Manager. It takes one highly-paid person to say “I need this thing now!” enough times for the analyst to eventually get admin access. The problem is the workflow is a pain in the ass. It’s usually a string of 4 or 5 emails asking which rules need to be approved – or what changed in the version. “Wait, what code did you add?” Yes, it’s right there in the interface but we’re dealing with humans here. It takes an army of developers to maintain a site but 1 document.write(); in your TMS to take the whole thing down.

Prediction: TMS’s will explore integration with ticketing systems like JIRA, Trello, and others (via API). More realistically, there will just be loads of redundant conference sessions talking about how you should totally focus extra extra hard on testing your tags before releasing them and it’s really worth going through QA.

We realize we went overboard with our tagging

You have how many tags? How on Earth can you manage that many? I get that you were kind of just doing some tagging here and there – we had some stakeholder requests but like… holy crap! We’re not even using half of these tags and about 20 of them are broken. When was the last time you even looked at this tag? Is it supposed to say “TEST TEST TEST” or is it just broken? Let’s consolidate.

We’ve abandoned our tools of yesteryear… our SDR’s, tagging documents, tag repository. We really don’t have any easy way to keep a concise inventory of our tags. This is a problem. There was a reason we did this in the past. Remember? Not just because these tags were in-line. They helped us tie back the tags to business requirements. They were a form of communication and enforced accountability for your implementation. Most importantly, they made you THINK about what you were actually implementing. Tag management systems encourage a “set it and forget it” mentality – not intentionally. They’re so easy to use that you don’t remember what the hell you implemented during holiday season a few years ago. This means our libraries are also growing in size at an alarming rate. How are we holding ourselves accountable? Simple. We’re not.

Prediction: Tagging documentation will slowly return. It will mostly serve as a tagging inventory for periodic audits. Tagging consolidation will become a “thing”. While we will strive to use more dynamic tagging methods to consolidate rules, we will also be trimming more excess fat this year.

We’ll focus on extensibility

Once we’ve done everything with our TMS we’ll crave more. Don’t limit yourself to the roadmap of your tag management system. There’s so much you can do with your TMS – like query API’s, measure scroll distances on pages, create unique combinations of conditions to trigger events or surveys. Test your boundaries (and make my first prediction come true). There is so much untapped potential and focusing on extending the utility of your tag management system outside the bounds of its default framework provides unlimited opportunity. Utilities will become tool-agnostic and our implementations will become more powerful than ever.

Prediction: We’ll get tired of working within the bounds of our TMS and find ways to boost the capabilities of our tools in ways we never considered.

5 thoughts on “The Future of Tag Management”

  1. This is funny: “More realistically, there will just be loads of redundant conference sessions talking about how you should totally focus extra extra hard on testing your tags before releasing them and it’s really worth going through QA.”

    I’m currently building a talk with Craig Scribner where we try to talk about data quality at two levels:
    – Craig focusses on end-to-end data quality, the ObservePoint approach that allows an Analyst or Marketer to see where they stand
    – I will ramble on about automated testing of data layer and tag management structure

    Our belief is that those two will make a huge difference. My hope is that DTM will at some point include some integration with Selenium or other test frameworks.

    • Hey Jan – That’s great to hear! It’s funny you mention that – Selenium and ObservePoint absolutely make a huge difference. These tools are key to diagnosing the gun-slinging! At the end of the day we’re still prone to error via the workflow. Selenium/OP can only preemptively catch errors when it is being run before a library is published. I can’t count how many times I’ve said something was ready to be published and it was just published outright. I think cases like that will become more and more prevalent and we’ll see many folks come out and say “Let’s slow things down!”

      • Not sure about slowing things down. I think it is more a question of marketing demanding a solid foundation for their frolicking about.
        I think testing (continuous, integration, whatever) at the level of the data layer is acceptable for dev and gives analytics/marketing what they need.
        The two huge advantages would be a) automated tests, no more surprises with missing data or non-firing rules b) the tests would become an explicit contract between dev and marketing/analytics, which is needed badly.

  2. I have seen pages with over 120 pixels. However I use TM routinely in my work and I think the future will have to be better support for integrating with project management and development tools such as version control and Continuous integration. On another front, I see that tags are changing as well and as web components become mainstream the whole approach to tag management is going to have to change.

    • Over 120 pixels? Holy. Cow.

      I didn’t even consider web components either. That will make things very interesting. I’m not an expert on this by any means – but is there visibility into the “shadow DOM” elements via jQuery/Sizzle? If not, things could get very interesting.


Leave a Comment