The psychology behind web analytics

Digital businesses rely heavily on data. But data can be meaningless unless it is analysed and interpreted properly – and not just to reflect what you want it to tell you. With so many metrics to consider, often we can get lost in the sheer quantity of information and not think about the human behaviors behind every action.

I recently had the opportunity to take part in an episode of the Digital Analytics Power Hour podcast, with Michael Helbling and Tim Wilson, and Michael asked his listeners some poignant questions: “Have you ever stopped for a second as you hovered over that order button? Does anyone care how you feel as you are doing so?”

So how can analysts, marketers and website owners better influence the customer journey, based on factors that may not be seen from the data? Traditional web analytics tell us a lot about what has happened, when what we should really be thinking about is what motivates that behaviour and why it happens.

To help go into this topic a bit further, here is a summary of some of the points discussed:

MH: How did you get to that intersection between digital behaviors and psychology?

If you think about it, we are not really selling products or services, we are selling experiences. And at the base of everything we do, we are dealing with human behavior. Every interaction – the clicks, scrolls, the hovers – they are all individual representations of a customer’s thought process. I got into this field by thinking: “OK, so we have the ‘what’, but to really understand the customer, we have to think about the ‘why’.”

TW: It sounds a bit like behavioral economics, where instead of thinking logically you bring more human psychology into the analysis?

Exactly. Most modern enterprise businesses will have data scientists or an innovation team in-house, but the problem I see on a day-to-day basis is that all of their analysis is about pulling linear regression, and that does not provide the sort of insights that they are necessarily looking for. We need to look more at why people behave in certain ways, any external factors, emotion, mood, etc.

TW: This sounds a lot like what I’ve learned on Voice of Customer over the years. Is it fair to say this is a cousin of those theories?

I think there is one fundamental difference between voice of customer and the work that Clicktale does – there is a huge gap between what people will tell you and the motivators behind their behavior. Beyond the environmental and emotional influences – mood, the company we’re in, etc. – and any cognitive bias that may be at work, there is one more reason why we need to look beyond just what customers say.

In the same way as we have non-verbal signals in human communication, we have similar non-verbal signals in the digital arena. You can look at how fast or how slowly a visitor scrolls or clicks through the pages and understand their mindset without actually having to ask them.

MH: You’ve also spoken about is conversion cycle vs conversion point. Can you explain a bit more about that?

However you define a conversion, the average conversion rate is about 3%. What about the other 97% of customers who do not convert? Do we infer from this data that they had a negative experience? Of course not, so we need to think differently about conversion.

Two people can interact with the exact same stimulus and they can have two completely different experiences. If I can detect a visitor’s intent when they arrive at the site, I can understand whether their experience was positive or negative. This isn’t done on an individual level, but using our algorithms we can assess hundreds of experiences and discover the intent of their visits to an accuracy of around 85%.

MH: To what extent are the demographics of the user involved in this process? Obviously, so much of digital is anonymous.

We have to create new demographic segments. For instance, are they goal-oriented, are they browsers? These are mindsets, and they can change as they go through the website too. You can detect these personality traits as you analyse their interactions and their journey through your site.

MH: How can people start to bring some of these ideas into their tactics?

Firstly, I would suggest that analysts start to think like psychologists and start to question every dataset they are looking at. Humans are not rational and we need to think about the factors involved in their decision-making processes. Purchases are emotional, so we need to provide ways to rationalize human decisions. Always ask the ‘why’ behind the behavior and not just settle for the ‘what’.

Also, every customer is different so we need to stop using the idea of best practices. If we think about it, a best practice is a common denominator that can be applied to different types of people. But the ultimate goal is personalization. And with that comes the realization that we need to provide different experiences for different people.

Listen to the episode in full below

Listen Now

If you would like to know more, or to subscribe to future episodes, visit the Digital Analytics Power Hour podcast website.

Talk to us to explore how customer experience analytics can improve your business