brightly colored beads of an abacus
PHOTO: Crissy Jarvis

Admit it, we all do it. I know I will once this article goes live. I'm talking about how whenever we post something online, we can’t help but check back later to see how it was received. Thumbs up, likes, retweets, comments, downloads, page views. We all love metrics, whether it’s just “did anyone like the picture of my cat I posted on Instagram yesterday,” all the way up to complex reports about web-traffic, journey flow, click-through rates and all that good stuff it takes a data scientist to sift through. We have so much data available about customer interactions that the true meaning of them is often forgotten.

The problem is most of the metrics we capture today are a record of what someone did in the past, a physical interaction with our content either through a button click, or following  a link. They don’t tell us why the person did what they did. 

And knowing why is the most important part of understanding the customer journey.

Getting to the 'Why' (and 'Why Not') of Customer Behavior

There is an excellent video from Adobe, entitled Click, Baby, Click that shows just how reacting to clicks without knowing what is driving them can lead to some wrong interpretations of customer demand. If you haven’t seen it, I highly recommend watching it — it’s a fun lesson you won’t forget.           

So if action-based metrics don’t give us the information we need, do time-based metrics give us a better picture of what’s driving customer behavior? They are probably a step in the right direction, but have the same underlying issue in that time-based metrics are still a reflection of past action. We may now know how long someone interacted with our messaging, but we still don’t know why. For instance, time-on-page can be a false indicator: is someone engaged with our content because it's good and they enjoy reading it, or is it so obtuse that they have to keep plowing through it to find the answers they want?

Most people come to our websites or interact with our apps for one of two reasons: to get answers to questions, or to complete a transaction. So maybe we should be measuring how well we achieve those two things. Instead of having page-based analytics, shouldn’t we be focused on content and transaction based analytics, combined with search analysis and time reporting to determine how easily, or quickly, our customers are achieving their goals?

On top of wanting to know what people do during a customer engagement and why they do it, it’s equally important to know why someone didn’t do what we wanted them to do. Why is no one clicking on that beautifully designed call to action button we spent hours refining? Why isn’t anyone finding the high value content we just know would help them? This is where tools like heat maps can help, as we can track where people engage with our designs.

Related Article: Sorry, There's No Secret Shortcut for Measuring Customer Experience

Understanding Intent

So if the current metrics are a snapshot of the past of physical actions, how do we realign for a future where, as I speculated recently, our interactions with the digital world will migrate from the physical to the ephemeral? How do we measure voice based interactions? 

In many ways we already are, but for a different need. When you call a telephone helpline, or get passed to a call center representative and get that message about “your call may be recorded for training purposes,” chances are high that training is low down on the list of why the call is being recorded. Call centers have long used technology to record, index and analyze customer interactions not just for what was said, but also the way it was said in terms of tone and inflection.

Will sentiment analysis drive the next generation of metrics for the voice assistant driven interfaces? It will allow us to not only understand what was asked for, and what the customer wanted, but also, with the application of machine learning, allow us to start to understand not just sentiment, but also intent.

Once we understand intent, as opposed to past actions, we can start to deliver predictive customer experiences, and look forward instead of backward.

Related Article: Following the Intent Data Breadcrumbs

How Can We Help You?

The only true indication of a successful customer experience is: did we help the customer do what they needed to do in a quick, intuitive and helpful way? Did we make their day easier, or answer their question? The more we remove friction from the customer experience, the more likely those customers are to return and want to engage with us again.