We still fail at basic analytics tagging and it’s hurting us

It’s all (still) about Javascript

If you have a website, you very likely installed a web analytics solution such as Google Analytics to track your site’s performance, which marketing channels drive the most traffic or even measure basic conversion. That can be done by just placing a Javascript tag on every single page of your website. That’s all you need to do to get basic traffic data flowing into your reporting account. And even something so technically trivial can be botched, because different site section are owned by different business units or different managers, who in turn use different IT departments or integration partners. Sometimes it’s a lack of education on the part of the marketing team about how to set up an analytics solution. And before you ask, using a tag management system does not necessarily help.

Adding a bit of tracking to a website can look like an afterthought, something that “can be done by an intern”. The truth is that too many websites end up with missing tagging.

“Hey guys, we forgot to put Google Analytics tags on the site, do you think the marketing guys will notice?”

The end result is that data scientists or your analysts get to work with inaccurate or incomplete data. Which means they cannot do what they’re paid for: finding insights to optimize your business conversion and generate more page views, more leads, more sales, more registrations – you get my drift. In the case of websites monetized with ads, if your tag coverage is off by a couple points, that’s as much of a dent that can be felt in your ad revenue.

This problem is amplified by the activity in a web property, with new content being created every day and new microsites and/or landing pages being added for different campaigns. How many of those sites and pages do you think are audited for analytics tags? Ask your marketeer in charge of the campaign why (s)he has no website engagement data 30 days after the campaign is launched.

See what I meant?

One analytics tagging plan to rule them all

And we’re only talking about basic tagging here. Bigger, more mature companies use advanced tagging that can capture dozens of visitor and product criteria but also rich user interactions. Advanced tagging is crucial to capturing the data you need for all that “Big Data stuff” you’ve been hearing about for the last 10 years: link your clickstream data with your CRM database, send targeted e-mail campaigns that actually reach the right audience, optimize campaign ROI and acquisition costs – and a lot more.

Advanced tagging requires a tagging plan that serves as your Holy Javascript Bible, the master blueprint for your data collection strategy. This document (or set of Excel sheets) describes which pages and clicks should be equipped with tagging to collect the desired data elements, sometimes just for a 10-page website, sometimes for entire brand sites covering dozens of local markets and hundreds of microsites.

Many of the companies I work with seem to suffer from the “my dog ate my tagging plan” syndrome where the tagging plan is either missing, outdated or just not adapted to the company’s data collection requirements. These companies have a serious analytics governance issue.

Implementing and maintaining this tagging plan can be time-consuming. Even more time-consuming is the quality assurance work required to ensure that the data you’re collecting is compliant with your tagging plan. In other words, ensure that your reports about visitors viewing the Refrigerators product category on your ecommerce website actually contain data relevant to fridges. You also need to explain why it is that you are not capturing the right product SKU or that the product you’re viewing is incorrectly flagged as out of stock.

How about that free sample?

Running data quality assurance for web analytics tags is either an afterthought or done poorly using the sampling method: take a look at a handful of pages on the site and surely all other tagging elements are guaranteed to be squeaky clean. Except they’re not. While this method requires very little time, resources and budget, it cannot produce accurate results. This is the iceberg metaphor all over again.

Like I hinted at earlier, quality assurance can be very time-consuming if done manually using sampling and even more so when done manually trying to be exhaustive. Of course, there are browser plugins or companion tools that analysts and developers can use, such as WASP, Fiddler, or Charles. Even Google recently announced an update to their Chrome extension Tag Assistant that helps with analytics QA work. These tools are great for spot checks but they cannot be used for quality assurance on a large scale, over multiple websites and in a short time. They also include no features for tagging plan validation.

This is when you start to either invest heavily in quality assurance with lots of valuable dedicated resources or start using an automated quality assurance solution such as Hub’Scan. This type of solution scans your website and checks which tags are installed and validates the data they collect. This allows you to save time and money in quality assurance so that your analysts and developers don’t waste time on menial QA tasks when they could be doing something more productive. Hub’Scan is not the only solution in the field, and others such as ObservePoint and Tag Inspector also offer analytics QA services.

Tools for analytics tagging are great but what about people?

Even if solutions such as Hub’Scan can provide recommendations on how to fix and improve tag-based data collection, real value in analytics comes from a corporate culture of data governance that is part of your company’s digital transformation. Your analytics team should be focusing on data, not tags; on value and not compliance.

If not done already, start defining processes for:

– data collection requirement gathering
– tagging plan definition and maintenance
– communicating tagging guidelines to development/integration teams
– running tagging quality assurance processes
– build quality scorecards that show areas for improvements
– apply recommendations for optimization
– profit!

If you caught my subtle plug earlier, Hub’Scan can assist with a lot of these steps so give us a try.

tl;dr / cliff notes

– Your website is a communication platform with visitors/customers. Use web analytics tracking to measure user engagement and conversion
– Make sure your tags (Javascript code used for tracking) are deployed on every single page and interaction of your site(s)
– Build a tagging plan to list all tag values per site/page/click
– Verify that tagging is 100% present. Either check manually on a handful of pages (slow & inaccurate) or use a solution such as Hub’Scan (automated, fast and exhaustive)
– Get your company to adopt data and data quality as part of their company culture and daily tasks.

Questions? Comments let us know in the comments below!


I mentioned Google Analytics in this post because it is by far the most widely used web analytics platform in the world but of course there are other, including (but not limited to): Adobe, AT Internet, Comscore, IBM, Webtrends or Webtrekk.