Analytics products can be complex. There’s a lot that goes on in the background that you, as a user, don’t see. Yet some of these intricacies can affect the results you end up with and ultimately base your decisions upon.
Here are three often overlooked – but very important – things you probably don’t know about your analytics.
1. Tracking Varies By Tool
Every analytics product tracks the activity on your site differently and it’s important to understand how. Each tool will recognize visitors as online, idle, or not online in its own way. It’s especially complex when you get into real-time tracking.
For example, some vendors report based on which visitors have committed a new pageview on your website within the last 5 minutes. That means if a visitor is reading, watching a video on your website, or viewing another tab for more than 5 minutes, it won’t report that visitor as currently on your website. It also means that if a visitor left your website within the last 5 minutes, the tool will actually show that they are still active on your site.
Similarly, idle times are handled differently by each vendor. For example, in Woopra, if a visitor is inactive for more than 5 minutes, they are considered idle. This is, however, only the default setting and is completely configurable. So you could change it to 3 minutes, 10 minutes, 1 hour, etc.
2. You’re Not Getting All Your Data
Well, you *might not be getting all your data, depending on which tool you’re using.
Some of the common digital analytics tools, like Google Analytics, use sampling, especially for high traffic websites. Sampling is when the analytics vendor uses only a subset of your data, rather than all of it.
Sampling can happen in two ways:
- Collecting Data – This is when the vendor only collects a sample of your data. That is, the tool won’t track and store all of your data.
- Generating Reports – This is when a vendor uses only a sample of your data to generate a report. For example, say you have 100 visitors. When you run a report, the tool will only use data from 50 of those visitors.
Many vendors use sampling as a way to cut costs. This can work for you if you’re not too concerned with the accuracy of your measurements and if you do not require individual-level information about your customers.
Still, sampling can be especially problematic if you segment your reports. Once you start segmenting a subset of data and try to run deep analysis on it, your results may be quite skewed.
3. What Are You Measuring?
Different digital analytics tools measure different things, such as people, pageviews, events etc. Traditionally, most web analytics vendors place the pageview as the primary unit of measurement.
On the other hand, at Woopra, we place the customer as the central unit of measurement and events (e.g. actions committed on your site, pageviews, etc) as the secondary unit of measurement.
You will see this difference reflected in a number of ways. For example, whenever you generate a report, you will get your results in people, pageviews, events, etc.
Being aware of these various differences is the first step to making an educated decision about which analytics is right for you. These assumptions, while they may seem like subtleties, will ultimately affect the results you get and thereby the decisions you make. Not every analytics tool is right for every digital property.