16 Jul Analyze This! Lesson Four: Data Misinterpretation – Yep, Happens A Lot
On an old episode of Penn & Teller, they addressed numbers, stating at one point, “Numbers don’t lie; the people who use them do.” This couldn’t be more true when interpreting web analytic data. My last post talked about errors matching PPC data with your analytics. This lesson concerns common misinterpretations of data within Google Analytics.
When I take on a new contract, I typically sit down with my client and talk with them about what useful data they need to see. Out of this discussion comes a custom report for analytics data. Typically, I have found that sending over a scheduled Google Analytics report is confusing to executives, since many do not understand it. Often it’s easier to simplify it based on what they’ve identified as their “useful data.”
I assume other people in the industry do this from time to time. When you are reporting daily and weekly numbers, it is easy to misinterpret data. Below are two examples of how this can happen.
New & repeat visitors
If you came to my site for the first time today, left, and then came back again in the same day, you would count as a new visit and a repeat visit. However, the total number of visits would only be one. In this case, if you added new plus repeat visitors together, this is not going to be an accurate total.
Unique visitors summed daily
When you work in an office environment, sometimes executives will look for daily reporting and then at the end of the week ask for a summary of performance. If you are using Excel to track unique visitors per day, you may wind up reporting inaccurate data. A great example of this is if you post a page that gets 2,000 unique visitors on Monday and the SAME visitors return each consecutive weekday — the total of unique visitors for the week would still be 2,000, NOT 10,000.
Be sure you understand the fundamental differences here to ensure you don’t fall into the trap of critical data misinterpretation on your job. In my next post, I will explore the steps you need to take in order to limit data misinterpretation and improve your web analytics setup to show you the most accurate numbers you can get.