Continuing the discussion on the 10 essential tips to achieving data quality nirvana from Web Analytics Demystified’s white paper, what’s the next step? Define governance and controls for your organization’s data quality.
Designing for Quality
Quality data does not just organically appear, especially in a complex enterprise. Data quality happens because the organization agrees that having accurate, trustworthy data is important and because business stakeholders are willing to design the necessary internal processes and behaviors to facilitate data quality. Data collection best practices revolve around three key processes:
Adding new data collection devices
Data collection is a key process in any business, but one especially important in complex, distributed, or multi-brand businesses where an array of resources are developing and deploying measurable content. All too often web analytics is considered an afterthought during the design and deployment process and, typically at the eleventh hour, code is slapped on the page without regard for proper deployment and variable setting.
Creating a clear and consistent business process for adding data collection that is integrated into development and content deployment processes is critical to your success in creating actionable and trustworthy data.
Verify proper functioning of data collection devices
Just deploying data collector code to your pages and applications isn’t enough. You will need a process to confirm that variables are set properly and that the code functions as expected under a variety of conditions. Given the still changing browser landscape, coupled with the emergence of mobile devices (with additional limitations), verifying proper coding is more important than ever. While there is likely some temptation to assign this responsibility to your normal developmental quality assurance (QA) staff, resist that temptation and keep the data confirmation processes within your digital measurement group!
Keep in mind that it’s not enough for data to simply be populated—the data has to be correctly populated. A combination of knowledge about data collectors, analytics systems, and the systems generating the data is critical. While proper function can be verified using simple technology found in a number of QA tools, such as ObservePoint, there is no substitute for knowing why the data is being collected and how the data collected will be used in a business context.
Conduct a data collection device audit
Once you believe that data collectors have been correctly deployed, you then need a process to confirm the deployments are maintained over time. While it may seem odd—the need to confirm that deployed code has stayed deployed—in our experience most modern web sites have enough moving parts and enough people touching code that failing data collectors are surprisingly common. Fortunately this process is easily facilitated thanks to tag auditing and management platforms.
Bonus Tip: Tag Management
A new product category has emerged in the past few years that, like tag auditing solutions, is designed to help the enterprise improve the overall accuracy and utility of tag-based data collection. These systems similarly help mitigate common challenges associated with digital data collection.
While TMS is very much a step in the right direction, these platforms also require care and consideration in their deployment. According to Patrick Foster, senior director of analytics and messaging systems at Turner Broadcasting Systems, “If there was some way to guarantee that our tag management system was deployed correctly every single time we wouldn’t need something like ObservePoint but, at least today, we don’t have a tool to confirm the accuracy of our TMS deployed tags and, even if we did, we believe the human element will always necessitate validation.”
The third step to achieving top-notch data quality? You need to find a robust tag-auditing platform for your organization.
This post is based on the white paper Data Quality and the Digital World by Eric T. Peterson, principal consultant at Web Analytics Demystified.
About the AuthorLinkedIn