Monday, 4 January 2016

Why Tag Management Is The Way Forward In 2016

group-of-tags-marketing-old-ssIn our recent white paper, “State of Digital Marketing Analytics in the Top 500 Online Retailers” (registration required), my colleague Amardeep Singh and I found that a full 42 percent of Internet Retailer’s top 500 retailers have yet to adopt a tag management system (TMS).

While this may not have been surprising if we had been looking at the top 5,000 retailers, we were shocked to find that so many leading e-commerce organizations had yet to deploy a TMS.

Even more surprising was that, when digging into those 42%, a full third of them were using multiple analytics tools. Managing one analytics implementation with a TMS is tough enough, so trying to manage more than one is quite the undertaking.

ml blog post

Anecdotally, I would say that within the digital analytics industry, there’s broad agreement that implementing a TMS is now a best practice.

I’d go so far as to say that a TMS is practically a prerequisite to managing analytics efficiently, and ultimately to deriving serious value from analytics at all.

This got me thinking — if a TMS is so important, why aren’t more of the leading retailers using it?

Clearly, many organizations are still concluding that the costs (or perceived costs) of a TMS exceed the benefits. So I thought I’d do a quick recap of those costs and benefits and — if you’re still on the fence about adopting a TMS — try to convince you that it’s the right way forward in 2016.

Of course, this table doesn’t capture all the benefits or costs of a TMS, but these are the points that I’ve found come up for discussion most often.

table

How Do I Get The Benefits Of A TMS Without The Risk Of Employees Making Mistakes?

Of course, there will never be a 100-percent guarantee that one of your team members won’t make a mistake when deploying a tag, causing an issue on the front end. In reality, for all the talk about convenience and agility, a TMS was never supposed to remove the need for a QA (quality assurance) process.

Fortunately, a TMS actually helps to standardize that process, as all users — whether business or technical — have to go through the same workflows to ultimately publish a tag.

A great way to mitigate the risk of “user error” when it comes to tag management is to invest in some basic training for your business users. TMSes generally have easy-to-use interfaces that are intended to help business users make changes effectively, so the groundwork is already laid.

Providing some conceptual training on the kinds of rules and logic that can be used to fire tags, as well as some tactical training on QA changes, can go a long way.

In addition to some training, defining a governance plan will help you avoid TMS missteps.

For example, perhaps you allow business users to create and edit tags, but you require a technical user or TMS expert to review and publish tags. Decoupling the initial tag workflows from actually publishing the tags adds a bit of time to the overall publishing cycle, but it may reduce risk and provide a useful “check and balance.”

Doesn’t A TMS Create A Single Point Of Failure On My Website?

This concern comes up in a couple different varieties.

First, what happens if the TMS you’ve chosen goes belly-up? Second, what happens if the TMS you’ve chosen is “having a bad day.” Won’t that cause serious issues for your website? Let’s discuss each of these in turn.

I suppose it’s possible that your TMS vendor could go out of business or go through a merger/acquisition that results in substantial changes to the product.

If this is of particular concern, then tools like Google’s or Adobe’s might seem especially appealing, since both of these companies are large, stable, publicly traded companies unlikely to disappear on short notice.

On the other hand, marketing technology is inherently a fairly unstable world. New tools routinely come along and displace others.

If “stability” is of paramount importance, it’s possible you’ll avoid some bumps in the road, but also that you’ll fall behind competitors who are willing to adopt systems and tools created by startups or smaller companies.

Second, what if your TMS happens to fail on any given day? Given that the TMS is intricately woven into a site’s functionality, it’s common to wonder if a TMS failure would be catastrophic.

Fortunately, most TMSes allow you to load the TMS asynchronously, which means that even if your TMS fails, other page content can load as normal, minimizing the impact on the user experience.

In fact, taking all of your third-party scripts and abstracting them away from your content (and into your TMS) actually helps you reduce the risk of code failures impacting user experience.

Won’t Implementing A TMS Take Too Long, Be Too Expensive Or Difficult?

Depending on where you are with analytics as an organization, implementing a TMS can be a surprisingly easy process. A very basic implementation can be done within days, although not everyone should hope for this.

In my experience, things tend to get tricky when the pre-existing analytics deployment is quite complex. Using Google Analytics as an example, if you’ve already got a robust, custom implementation without GTM (Google Tag Manager), you’ll have to ask yourself: How do all of those events, custom dimensions and so on get “translated” from on-page JavaScript to TMS-based tags?

The good news is that if you’re savvy enough to have developed a complex analytics implementation without the benefit of a TMS, you’re surely well-suited to do so once you have a TMS at your disposal.

The question really becomes: Why make the effort? Well, as mentioned above, the benefits of adopting a TMS are numerous:

  • More nimble marketing: Tags are added or edited more quickly, and with reduced involvement from IT stakeholders, who are freed up to do more strategic work.
  • Better governance: With a single system in place at the center of your tag management efforts, it becomes much easier to both define and apply governance, reducing mistakes and duplicative efforts.
  • A data layer: Though not required when adopting a TMS, a data layer often goes hand-in-hand with a TMS. The data layer helps you standardize the definitions of your digital data, allowing multiple applications to read and react to data much more efficiently.

Conclusion

Ultimately, making the jump to a TMS gives you a flexible foundation upon which you can build as the world of marketing technology grows ever more complex.

It’s true that, especially for organizations already using a complex analytics implementation, migrating to a TMS isn’t a trivial project. But as the majority of the top 500 online retailers are now finding, the long-term benefits outweigh any short-term costs.

No comments:

Post a Comment