The Role of Automation in Digital Analytics Governance & Implementation

by Ed Finn

You are an informed marketing leader. You know how to make your website and marketing technology work for you. You understand how good data informs the needs of your business and helps you take effective action. You know that marketing measurement is an iterative process, and have selected your next goals for data enrichment. You’re already taking on Customer Data Platforms and driving real-time personalized experiences with data.

You are head and shoulders beyond the median performer in your role. You get booked to talk in major cities. People call you The Future.

But as you continue to invest in evolving your business’ capabilities, how can you be confident your data is evolving with you? How can you maintain consistency and quality in how you collect data over time? How do you keep up with more and more teams requiring digital data to perform, and their increasingly complex requirements?  

The Scalability Problem of Too Much Data

The dirty truth nobody is talking about as we build new means of measuring and taking action on digital analytics is that we tend to throw out the old as we produce the new. Marketing campaign launch dates, code freezes, lack of documentation, limited resources … All these things lead people like me, the people architecting your online measurement, to say, “If we can’t tell the client it is valuable and they can’t tell us they can use it, then we can’t justify wasted effort on maintaining or retaining it.”

Diseconomies of scale threaten to undo all the good work that we’ve done to produce actionable measurement in this industry. There is simply not enough manpower (or rationale for the enormous investment of human time) to guarantee that the datasets will remain accurate and consistent. We have too much data, too many websites, and too many conversion points with an increasing demand to continue to be more efficient with less.  

Web and Marketing Analytics – A Tale of Old + New

In the early days of digital analytics, we ran relatively simple websites with equally simple analytics requirements. Each page needed the same javascript to trigger the analytics tool and only a small number of vendor tags (because only a small number of vendors existed). If I wanted to check all the pages for the web analytics tool snippet and then check all the pages for the AdWords snippet, I could do that manually or use a free online tool.

Old Website Measurement and Tag Management Structure

 

Today, we have enterprise websites that run an abundance of unique tagging and measurement requirements that are crucial to informing and empowering the work of multiple distinct business units. Those working in the industry (or those who look at the chart below with a magnifying glass) will know how much marketing and customer retention automation depend on tagging and measurement requirements. These are the babies we tend to throw out with the bathwater as we evolve new features and perform site relaunches.

New Website Measurement and Tag Management Structure

 

… You get the point.

An Example from Digital Analytics

Ok, now let’s run through a short hypothetical situation with both the old and new models in mind.

Suppose that an AdWords tag is missing from one of our product pages. It’s no longer noting add-to-carts on one of our most popular products, and as a result the sales of this item are going to decrease, because we won’t be able to retarget the users that abandon this product.

Old Model: There is no doubt that the AdWords tag is missing. This tag is supposed to be on every page. We may have noticed this by proactively reviewing the pages, using a simple automatic service, or we may have reacted to declines in the volume of retargeting campaign affiliated sales in Google Analytics. We have a number of ways to notice this error without asking a bunch of follow up questions. As soon as we are on that page with a debugger running we can clearly conclude that the marketing tag is missing.

New Model: Well … good luck. Unless you have a robust and dedicated digital governance team working on tag architecture and quality assurance full time, you probably have a vague-at-best sense of whether or not a tag is missing on purpose or by accident because there is no central framework for page types, measurement requirements, and tagging requirements.

The Solution: Automation of Digital Governance (and Implementation)

The solution to the complexity is actually quite simple … and our partners over at ObservePoint already built a solution. We’re so fond of it, we were recommending it to clients long before our partnership began. (No, we don’t take referral fees and no, they didn’t ask us to write this.) I’m an analyst. A big part of my job is to review the latest technology the industry can offer so I can help establish new best practices for our clients.

ObservePoint automatically scans and validates your digital implementation based on rules that you set. Setup is not overly time consuming (comparatively speaking to a manual digital governance framework), especially if you work with someone who knows the product already. Once your scans are set up, automated governance can begin. Care and feeding required is as little as updating it (recommended on a regular schedule) with new tags that need to be brought under the governance umbrella.

There are two ways ObservePoint can be used to help improve the efficiency and accuracy of your governance practice:

  1. During your analytics or martech implementation quality assurance passes in order to validate what you just did is working as expected
  2. After a robust build is complete (and as it is updated from there) for ongoing scans of your full infrastructure

We can say from having tested it internally that the time savings of using a tool like ObservePoint are massive, and the margin of error is dramatically reduced (if not entirely eliminated) for QA.

Digital Governance in 2019

We’re noticing that more and more marketing leaders are bringing up digital governance on calls with us, and we expect to have increasingly more conversations about this over the year ahead … especially as data regulation requirements and privacy concerns become paramount to a data-driven enterprises’ long term success. Not to mention the new marketing tech and measurement requirements of 2019 that we haven’t seen yet.

If you have questions about digital data governance or ObservePoint give us a shout. A wealth of great information can also be found on the Analytics Summit website (with a session on governing digital marketing spend featuring my Napkyn colleague Colin Temple).

Ed Finn

Senior Practitioner, Analysis

Ed’s passion is turning data into knowledge. As a Senior Practitioner on Napkyn's Analyst Team, Ed is responsible for driving the strategic direction behind some of our largest and most complex programs.