Your web analytics solution is only as good as the data being passed to it, and sometimes the data being passed is simply “bad.”
Tags break, pages are coded incorrectly, tools conflict, implementations fail—you name it, it happens and it impacts your data, your analytics team’s credibility, and ultimately your organization’s ability to be agile and truly data-driven.
This Halloween, take your analytics data from bone-chilling to bona fide with these seven steps:
1. Promote a Pumpkin King
The first step towards data quality nirvana is to assign ownership to data quality. The first step towards data quality nirvana is to assign ownership to data quality.
A data quality owner should have the right background and authority to influence organizational practices and beliefs. A data quality owner should have the right background and authority to influence organizational practices and beliefs.
4 Strengths of a Data-Quality Leader:
- Strong Technical Background – Competency with the technology and familiarity with the platforms each digital property is built upon is a must.
- Seniority – A person with context, influence and trust improves the odds you’ll have a better marketing technology ROI.
- Understand Business Value of Data – Understand and create dialogue about the requirements, purposes and methods of data collection.
- Data Orientation – A data-oriented steward understands the effects of breakage, knows where to look for problems and will choose effective corrective actions.
2. Gather Your Goblins
Define and deploy a data governance strategy
Data collection and governance best practices revolve around three key processes:
- New Tags – Create clear, consistent business processes for data collection, integration with development and content deployment cycles.
- Verify Tags – Confirm variables are properly populating and code functions as expected under all conditions. Ensure data collection is as consistent as possible across browsers, devices, geographies and does not hurt the user experience.
- Conduct Audits – Deploy a process to confirm that tag deployments are maintained over time. Websites have enough moving parts and enough people touching the code that tagging problems are common and frequent.
3. Skeletons in Your Data?
Watch for duplicate tags on your site
One of the lesser known problems in the web analytics world—data inflation—is exactly what it sounds like: digital data that is reported inaccurately high.
When more than one instance of a tag fires on a page, data inflation occurs. Each time a tag fires, an action is counted, and when a tag fires more than once, the action is being counted more than once.
For example, if two Adobe Analytics tags exist on the same page, a single visitor can be inaccurately recorded as two visitors.
The causes of data inflation are common and occur often. Tags can be coded to multiple places on the skeleton of a page, transposed from other pages or duplicated in some other way. Data inflation also happens when new versions of tags are deployed, but the old tags are not removed.
Inflation can affect web analytics data such as traffic counts, sales attribution data advertising, data such as clicks and impressions and a variety of other important metrics. And as this inflated data gets passed into other systems throughout your MarTech stack, the effects of the inflation are magnified.
4. Things That Go Bump in the Night
Be aware that website tags are bound to break
One of the most dreaded conversations a digital analytics professional has to face begins when an executive with a printed report in his or her hand asks, “Why don’t these numbers look right?”
Have you ever found yourself in this web analytics death spiral? In a split second, you must decide if you are going to defend your data and risk your reputation (possibly even your job) or decide to backpedal saying that you will have to recheck your data.
It’s a lose-lose situation.
When it comes to web tags and implementations, things break and go bump in your reports.
Take the appropriate measures to continually review and verify your web analytics efforts, and follow best data governance practices to help work towards greater analytics data quality.
5. Got Ghosts?
Prevent data leakage
Ghosts are so scary because you might not even realize they are present, haunting your space, you just know something isn’t quite right…
You won’t see flickering lights or hear a poltergeist plodding down the hallway in your analytics architecture, but that doesn’t mean data leakage ghosts aren’t lurking around your site.
In the white paper Data Quality and the Digital World, Eric Peterson warns that companies have started to leak data through the multitude of tag-based data collectors deployed across their digital properties.
“Given the relative ease with which they can be added to a website, combined with the fragmented approach companies take to digital measurement, analysis and optimization, it should be no surprise companies have started to leak data.”
Deploying data collection systems without a clear plan for maintaining accuracy, validity and security of data collected is undoubtedly a poor practice.
Consider These Five Scenarios:
- An employee leaves the company to work for a direct competitor but maintains access to traffic and revenue data through unknown deployment of analytical tools.
- An agency deploys tools with questionable Personally Identifiable Information (PII) collection and handling practices.
- A Tag Management Solution (TMS) is deployed in an effort to consolidate the chain of authority for controlling site tags, but that doesn’t prevent other groups from circumventing this process and deploying tags outside of the TMS to meet their own needs.
- A third-party vendor deploys tracking that sells data to other third-parties, which potentially exposes your data to direct competitors.
- A policy bans particular technologies on all sites, yet their tags continue to appear.
6. Count Your Candy
Audit your data collection once, twice, three times
Website content is constantly added, changed and retired—and this creates tracking challenges. You must manage for these changes while still being productive with your other duties.
A THREEFOLD AUDITING STRATEGY IS PRESCRIBED FOR ALL NEW CONTENT:
- Test frequently during the development process—Audit multiple times during any large-scale deployment effort. This instills resources throughout your company with data collection best practices. Share the results with developers and project managers when appropriate.
- Test often during quality assurance and testing phases—Closely monitor the data being passed to your analytical systems after your development project has passed into your QA environment.
- Test immediately following deployment to confirm data is being collected as planned—Confirm data collection has successfully migrated from the development and staging environments to the production environment. When problems are identified in this phase, act quickly to correct them. Assign resources to post-deployment efforts tasked with ensuring data collection is correct.
The deployment process is not considered complete until all aspects of the code are deemed functional in a production environment including data collection code. Commitment to analytics data accuracy and utility within the broader enterprise means asterisks next to numbers are seldom seen.
7. Beware! Keep Out! Danger! Caution!
Deploy a Data Quality Assurance Platform & maximize the ROI on your marketing technology stack and be alerted to problems before they become data disasters
Of course, all of the above steps—crucial as they may be to fixing your scary data—are also labor intensive, expensive, time-consuming and still prone to human error when your web analytics team tries to monitor it all without automated technology.
“We have 40 or 50 developers dedicated to the website and we roll code twice a week. [Before ObservePoint’s Data Quality Assurance Platform], we didn’t have an easy process to validate page tagging on such a frequent basis. I needed to hand check new or altered pages with each roll-out, which could easily lead to missed errors in the tags.” – Tim Munsell, Lead Web Analyst at DaveRamsey.com
DATA QUALITY MATTERS – Web Analytics, Advertising, Testing tools, DMPs and TMSes are only as good as the data they process. Data Quality Assurance confirms that the source of this data—the tags on your website, in your video and in your apps—are deployed correctly and completely.
MEASURE AND IMPROVE DATA QUALITY – Compared to the cost of traditional Quality Assurance processes and the risk associated with bad digital data, you can’t afford to not deploy Data Quality Assurance.
IMPROVE ROI IN ALL DIGITAL – Consistent, correct and compliant analytics data improves the ROI of your entire digital marketing technology stack.
Want to see if you have any ghosts or goblins lurking in your analytics data?
Using your company URL, and without deploying a tag, we’ll initiate a custom audit on your website and guide you through our findings, alerting you to:
About the AuthorLinkedIn More Content by Chelsi Linderman