Linear Web Analytics Can Be Bad
I was working on a data project the other day and a group of people were looking at some web analytics data sent over by an advertising agency.
The conversion rate and other key metrics were based on assuming viewing behavior went something like this:
Visitor A is exposed to an ad or link on social > they click the ad and go to page X > A certain percentage then download the white paper
Great. But that’s not the full story.
This linear view on analytics can make conversion rates easy to calculate but it can leave half the information on the table. It takes the viewer behavior and assumes it follows a linear, or straight, pattern.
Problem is, the viewing experience could easily be something like this:
Visitor is exposed to an Ad > They keep growing through pages or apps > Visitor is exposed again and clicks ad > They land on the page but leaves again > Visitor is exposed to a different ad and clicks > Visitor downloads a white paper on a different page
The point here is that basic web analytics can be providing false information because it doesn’t take into account the complexity of the viewing behavior.
And data can be misleading for multiple reasons, for example:
- The user denies a cookie so can’t be tracked
- The analytics program doesn’t map the full behavior
- Weighting isn’t given to the first impressions of the ad but only to the last touch
- The ad could be converting on other company pages that aren’t being tracked
These are just a few examples. The point is that a linear-based conversion rate figure or other digital marketing calculation could be prone to errors.
As digital analytics become a cornerstone of the information development of organizations, it’s worth considering the linearity of your web analytics and whether a lack of understanding is impacting your business. Be careful not to rely on a single analytics tool to map all of the different variations.
The key here is to be aware of the short-term limitations of the analytics work and continually look for ways to measure the natural viewing behavior of your audience. That might mean using different tools or investing time and experienced resources to continually improving your tracking.
