Google's Justin Cutroni: There's no such thing as perfect data
We asked one of our analysts to tell us the top two things he would always ask any prospective Google Analytics expert. That is, if he was about to hire a new analyst, what questions would he ask to make sure he was choosing the right candidate?
Here’s what he said...
Q What, in your opinion, are the biggest challenges in accurate data collection and how do you plan to overcome them?
A As Google's Analytics Evangelist, Justin Cutroni, astutely asked at the 2013 Google Analytics User Conference: How many of us had perfect data last year? What about in 2011? 2010? No hands went up for these questions because all analysts are constantly made aware of the limitations in our data and reports. That said, it’s also crucial that we continue to seek out other ways of extracting what's needed. A common method is to look at other related metrics (eg, analysing the titles of landing pages for searches where the keyword was not provided). A great analyst, however, will try to go beyond this and find out as much as possible about where the data might have errors – in order to prevent the organisation from being systematically misled by their reports.
Q What commonly reported data should our teams NOT be focusing on?
A Most company analytics have evolved gradually over time and have a few quirks whereby they find themselves getting obsessed with pinning down metrics that might have been important once upon a time, but now aren’t necessarily relevant for the business. A fresh pair of eyes is a great way to start applying the "less is more" principle and seeing what aspects of your reporting you should be farewelling. This is particularly relevant for non-actionable data, which looks pretty in reports but doesn't go anywhere – beyond cluttering the landscape and taking our attention away from insights that can actually lead to action. It’s important to make sure team members aren't going to be gung-ho on adding more "stuff" to reports. An effective analyst is aware of the context of what they're presenting and will continue striving to remove the clutter.
The takeaway here is that if you’re looking to employ a new analyst, or considering signing an agency to analyse your website, the two things you should have top of mind when interviewing candidates for the job are accuracy, and actionability.
As long as you’ve found an analyst who understands the limitations of the data sets they’re working with, it’s then your responsibility to ensure you’ve equipped your Google analyst with the business information they need to contextualise the data.
That way, they’ll know what they should be looking for and which insights are irrelevant to your needs – and you’ll end up with actionable insights that can make a solid difference to your business.