Automated data validation tools such as ObservePoint are certainly time savers, but they only identify data collection issues. It’s still up to you to solve them.

ObservePoint

We all know the importance of data integrity. We understand the need for continuous, uninterrupted, accurate data collection. This is why we spend time confirming any new analytics implementation is accurate from the start, all changes are thoroughly verified, and ongoing monitoring occurs.

Historically, data validation has been a time-consuming, manual effort. Build-time unit testing, error monitoring and reporting tools such as Splunk and TMS error logs, and routine data reviews are also key for ensuring consistent data collection.

Automated data validation tools like ObservePoint* provide an alternative to the manual approaches listed above.

What is automated data validation?

Automated data validation tools enable continuous, real-time monitoring of data collection in any environment, including your production websites. They can be configured to monitor what you want, when you want and, if problems are found, ObservePoint will notify you right away.  

How do you fix a data collection issue?

Identifying data collection issues is just the beginning. Whether an issue is reported by an automated validation tool, client, or colleague, the next logical step is to do something about it. The question here isn’t what we’re going to do. We’re going to fix it. The real question is how we go about it.

As an aside, I’m a firm believer that the first step to solving any problem is to verify that the problem truly exists. For example, as a developer, I often have custom settings in my local environment that can appear to interrupt data collection. Confirming the presence of the issue in another environment or with another team member is a critical step in any problem resolution process. We can’t fix something that isn’t broken, and raising false alarms can be both frustrating and tiresome for all involved.

Once we’ve verified an issue is real, the actual steps we take to address it will vary from situation to situation. However, the process we follow as corrective action is taken should be consistent, with plenty of communication along the way.

  1. Prioritize and communicate the issue to necessary teams and individuals
    • What is the issue?
    • What impact is known at this time?
    • What is being done to address the issue?
  2. Identify the root cause
  3. Plan the resolution approach and prioritize the work.
  4. Communicate the plan, LOE, and ETA to the appropriate teams/individuals
  5. Implement the plan, including a thorough validation/QA to confirm that the fix addresses the issue.
  6. Communicate the resolution to the appropriate teams/individuals, making sure to document the issue and resolution for future reference (documentation repository).

Data issues are a fact of life. They happen. Whether a web page changes, a URL changes, a data layer is modified or removed, or a bug is introduced in your analytics code, data issues exist. It’s how we prepare for and respond to them that makes all the difference. 

*Evolytics is an ObservePoint partner. We carefully choose our partners, believing they are best-in-class. 

Written By


Brian Johnson

Brian Johnson is the Evolytics Director of Analytics Implementation & Data Engineering. Specializing in Adobe Analytics, Brian helps brands including QuickBooks, eHealth, and TurboTax get the most out of their web analytics.