Customer Experience Optimization is the next big thing in digital commerce. It is a set of optimization activities that span across the wide spectrum of demand generation activities, customer journey, and different customer experiences.
CXO is both a big opportunity and a very complex problem whose solution requires systematic approach, knowledge, and use of smart technology. This mini series of posts helps business professionals understand just how big and how interconnected is the problem they are dealing with, and to use the posts as stepping stone towards architecting an effective approach.
This post, is the final installment in this series and is dedicated to helping you understand the time varying nature of the customer experience optimization problem. Your ultimate goal is to achieve sustainable results and if you do not understand or ignore the fact that visitor preferences are always changing, you might end up spending a lot of money and time, only to achieve temporary results, if any.
Customer preferences are always changing
It is quite remarkable how often we hear from digital commerce professionals and business owners that their key website metrics is very stable.
The following is our attempt to change that perception. We have to do it because this is the biggest misconception and one of the biggest threats to your successful customer experience optimization initiative.
Let’s start by saying that it does not matter what we say. Instead let’s use real life data. The following is the example produced through use of Google Analytics (GA) results for one of our real life customers.
As shown below, even those that do not ‘live and breathe’ data, will still see the significant difference in results between different months.
In the example below one can see 2x swing between different months.
If we go deeper into data and increase the resolution of the report to examine daily data, we will begin to see more of the time varying behavior. The less smoothed metrics swings will now be 5x-10x bigger.
Further increase in the resolution will just reinforce the key message: visitor behavior is always changing and the deeper you look, the more changes you will see. Hourly samples will now vary at 100x or higher rate.
Why would you care about time varying metrics?
After seeing the time varying pictures above, one can still legitimately argue that monthly variations are due to seasonality, daily variations due to different buying patterns during different days in the week, while hourly variations could be attributed to time-of-day behavior of visitors.
Yes. Probably most of the variance can be attributed to those explanations but this should not prevent us from considering the question: are these changes telling us something that will empower us to learn or do something to enhance revenue?
Following that thought process, we might compare two performance metrics on the same chart. Suddenly a picture begins to emerge that conversion and revenue lines did not move in lock steps.
How does this become actionable information?
In the ideal world, you would expect two lines to perfectly overlap or always be parallel to each other. Detecting significant discrepancies between two lines is the indicator that you should then dig deeper and try to understand what caused such business outcomes.
In general, you can explore two angles: visitor demographics and customer experience variables.
You should first examine the different visitor types to see if any of them is significant contributor to such trends.
For example, you may determine that your e-mail campaigns are causing a significant jump in conversion rate but they are not producing proportional lift in overall revenue. This might guide you in how to examine and improve the e-mail marketing campaigns.
If you are not running any customer experience optimization initiative, you can still get a glimpse if you have effective user experience for visitors using different type devices like mobile or tablet.
If you compare metrics for desktop and mobile device visitors and if you determine uneven trends for the same type of traffic, then you my conclude that your mobile or desktop form factor should be optimized for better customer experience.
The impact of time varying behavior on traditional testing and targeting solutions
It is unfortunate but testing and targeting vendors, lately called personalization vendors, are contributing to the confusion about the existence of time varying visitor behavior.
Number one, none of them are talking about the fact that visitor preferences are changing. They are promoting a “heads are in the sand” approach where they pretend that this ‘thing’ does not exist at all.
Number two, they are diverting customer attention on statistical confidence. Their recommendation is to start testing and implement results when they reach certain statistical confidence. For those who know that visitor preferences are always changing, this recommendation amounts to driving a car while watching the rear view mirror.
We have a deeper discussion of this topic in our blog post: Lies, damn lies, and statistics.
The fact that visitor behavior is changing is a good thing. Each time you have change, you have an energy that is driving your digital commerce in a certain direction. For many, this is a challenge. Our perspective is that this creates a huge opportunity for businesses to detect and harness this energy. To paraphrase Wayne Gretzky, businesses should “skate to where the puck is going to be, not where it has been”.