Averages, the curse of data led decision making
I can’t count the amount of times I’ve heard someone try to disqualify the validity of data because they used it in the past with no success. The most common cause of those misled data “based” decisions are averages. This problem is not unique to website analytics or businesses, it is a problem we face day to day when the media reports on a study, or when you read a headline on Facebook.
We take averages as an unquestionable and holistic metric that applies to everything and everyone. However, your customers are not average, they are a niche, they are people with specific behaviours that make them different from others. If you base your decisions on an average, especially they way you reach and communicate with your target market, you will end up targeting everyone. And you know what happens when you want market to everyone? You end up marketing to no one.
The only way to fix this problem is by always questioning data, especially when presented in averages. When it comes to website analytics there are some metrics you want to be more aware of. Here are some of the most common ones you will want to be more skeptical when seeing them presented in averages.
Bounce rate is a metric that has become a popular measure of success for websites. It measures the percentage of people who leave your website without visiting another page or performing an action that can be tracked. It sounds like a metric that can be directly tied with the performance of your website – if someone bounces it means they aren’t interested, and therefore not purchasing, right?
Not completely true. It is key to understand that not everyone is going to purchase from your website on the first visit. Every industry and website is different, so I am not going to give you an average of how many times people visit a website before making a purchase, but your own analytics can tell you that.
Your website has multiple pages, each of them has a different intent. You may have a page where people find all the information they came to look for and leave. If your website did the job right, they are going to leave very happy because they didn’t have to dig through multiple pages to find that information. On the other end of the spectrum, there are pages where you expect people to perform an action or go to another page in website (like a thank you page). In this specific case, a user who bounced is one who you weren’t able to convince or guide to the next step.
In both of these cases bounce rate would give you valuable information of the performance of that specific page. In one of the cases high bounce rate meant the page worked as intended, in the second it meant the complete opposite. If you were to combine them and create an average it would be telling you a completely different story – the wrong story.
E-commerce and conversions rates
It is common to report conversion rates and E-commerce metrics as averages because they can give you a general idea of the website’s performance. This isn’t a bad thing until you decide to use that information to make strategic decisions.
Ecommerce and goal conversions are probably the most important things someone can accomplish on your website. They are likely the ones that drive revenue or leads. That is why segmenting this data is so important. It will allow you to see specifically what type of people or what behaviours are leading to higher conversion rates.
Imagine you have an ecommerce website that had 4 visitors, one of them spends $100 while the other ones spend $1 each. Looking at the average order value they all spent around $26. With that information you might make the wrong assumption and group the characteristics of those 4 people and design a marketing strategy tailored to all of them. While it would’ve been more profitable and efficient to find more people that look like the one who spent $100 and not $1. (You obviously shouldn’t base your marketing plan on one person, but you get the point of how averages can misrepresent data)
Off course these are not the only metrics you should beware of, but they are the ones that I see being misused most often. Whenever you plan on using data to make a strategic decision, make sure you are not basing it on an average. Instead segment and analyse data further, to find relevant information that will help you guide your strategy.