Data Driven Business

7 years ago
Data Quality: Simplifying Isn’t Simple

It happened again, and it happens on a daily basis. Someone posts a desperate plea for help in the 30,000-member strong Google Analytics G+ Community I’m co-managing with other industry experts: "Help, the data doesn’t look right."

"But I followed the instructions!"
We’re sure you did, but it isn’t always as cut and dried as the examples in Google’s knowledge base. And now your marketing manager is seeking answers and your whole data set might be wrong!

As Alex Langshur, Cardinal Path co-founder and senior partner puts it, "Don’t make the right decisions on the wrong data." Indeed, there’s nothing worse than going down the wrong path without even knowing it!

Never assume<
Oscar Wilde is quoted as saying "When you assume, you make an arse out of you and me." Urban folklore or not, we’ve all heard it before… yet too many analysts "assume" their data is right – or worse – "assume" that because Google Analytics is simple to implement, they can do it on their own.

I have been working with messy, dirty, inaccurate data since 1985 and it’s still that way today – except there’s more of it, it accrues faster, and it’s exponentially more complex. Since developing the first ever tool specifically built for web analytics quality assurance, (WASP in 2006, which is light years in web analytics terms) I’ve developed a kind of 6th sense for spotting instrumentation and data issues.

We’ve gone from log files in 1993 to vendors claiming that a single line of JavaScript code on every page would be the answer. Today’s instrumentation of web sites, mobile apps and any connected device through technologies such as Google Tag Manager and Universal Analytics makes "page view" tracking sound like an artifact of another century.

Turning the light switch on doesn’t make you an electrical engineer
When well-intentioned people fancy themselves implementation specialists, they not only waste precious time they should be spending doing analysis, they put their businesses at risk! Literally, deploying Google Analytics (or any other web analytics solution) can look as easy as turning the light switch on – but is it?

Here’s a recap of some of the most shocking discoveries of the week:

  • A large government agency fires two pageviews on every page
  • SaaS-based service company recently launched their website with Google Tag Manager deployed at the bottom of every page – because their CMS didn’t allow otherwise (so they said!)
  • An ecommerce company discovered they had totally forgotten to include social tracking…a couple of days after the code freeze preceding the holiday season
  • A financial institution deployed with a single web property across many websites when there should have been multiple
  • A manufacturer relied on their web agency to deploy Google Tag Manager and Universal Analytics… they did a great job of defining everything as a custom-HTML tag relying on jQuery rather than leveraging the incredible power of GTM – they simply didn’t know otherwise
  • More than one company has deployed paid Tag Management Solutions less than a year ago and now wants to throw them out the window… sometimes because the tool is inherently flawed, but most often because it’s a lot easier to blame technology than it is to admit a lack of expertise and inappropriate deployment strategies

Ok, I get it. Now what?
Before you venture into deploying GTM and UA, ask yourself: Am I the right person for the job?

Put aside any ego and the obvious appeal of tinkering with a new toy and answer honestly: How long are you willing to mess around with tags, lack of confidence in your data and excuses? Can you afford to put your job, your reputation and your business at risk?

There are specialists available who have done hundreds of implementations and mastered the art of configuring Google Analytics and other web analytics platforms. At Cardinal Path, we are receiving a tremendous level of inquiry from clients who want to be ready to leverage Universal Analytics. We are deploying best practice implementations for all verticals faster – and to the highest standard – than ever before.

Technologies evolve and things that used to be amazingly tedious and complex are now available in seconds with the click of a mouse – but only if you know what you’re doing.

Attending eMetrics San-Francisco? Don’t miss Stéphane Hamel’s new workshop on Mastering Google Tag Manager (with Universal Analytics), March 20. He will also present "Shocking! Tricks Your Tags are Playing on Your Data

By: Stephane Hamel, Director of Innovation, Cardinal Path

One thought on “Data Quality: Simplifying Isn’t Simple

  1. Pingback: Tag Management Does Not Make IT Redundant - Business 2020

Leave a Reply