VR analytics: the art of attention

New platforms cannot succeed without a system to collect data, and analyse patterns leading to meaningful insights; not just to understand the user experience, but to set up a solid advertising framework to be used by agencies and brands. So what do we have for virtual reality at the moment? How does VR analytics differ from traditional web or app analytics? Can we transfer existing knowledge to this new field, or we need to rethink our models? How will we measure the most important goal, attention, in virtual reality? 

The bread and butter of analytics: attention and action

How do we measure activities on a website today? Without getting into details, we measure attention and action. Let's start with the easier one in virtual reality: action. In VR everything counts as an action when the user activates something. Opening up a new scene, click-looking on a button or purchase a new product. These are easy to track. In fact, measurement can be done using the existing tools like Google Analytics. We can easily get the same figures as we do now: when, how often and who interacted with what part of the environment.

Attention is a bit more complicated, but in a way developers and advertisers will love. In virtual reality we know exactly where the user is looking, where's their eye focus, hence their attention. In web analytics we do have a similar metric called mouseover, where we can measure when the cursor hovers over a piece of content. You can build nice heat maps with a similar technology. You can argue this is simply guessing attention, and you are not exactly wrong. A true attention map requires eye tracking, something that you normally only do in UX research labs.

In a 360 virtual environment this is a bit different. A content heat map sounds like a good idea, but it's tricky. It's easy to quantify VR - it has pixels, just as websites have - but it is super hard to analyse data which is always changing. There's just way too much to track, and the process has to be simplified in order to develop actionable insights. 

Looking at something vs watching something

We truly believe that display advertising will thrive in virtual reality. Having virtual screens in a virtual environment is the first thing that comes to mind, but we have to learn our lessons from the not-so-glorious state of display ads.

There's quite a difference between looking at an ad and watching an ad. Both of these actions have our attention to some degree, but they are different in terms of our commitment. The insights about page depth and scroll depth on a webpage have the equivalent in VR, too: the amount of time and focus we spend in different scenes in a virtual environment. It can be quantified and measured easily. 

A very different user journey

The ultimate question of analytics is simple: have we met our goals by creating this platform? To answer this we define Key Performance Indicators (KPIs), set up KPI goals we need to achieve, and create funnels where we gently push users towards conversion. In the virtual reality KPIs and goals can be defined the traditional way. We might need new metrics, but the process is pretty much the same.

The funnel metaphor won't work though, as the user journey is very different. It's easy to understand why: in a virtual reality environment users love to aimlessly look around sometimes, creating an insane amount of noise. Sure, we can establish checkpoints, like a virtual screen that runs our ad, a virtual shelf in a virtual house with product placement, a button to trigger a new scene etc; but we also have non-interesting but spectacular places where the user can wonder around for an hour contributing towards none of our KPIs.

The answer is creating virtual zones, areas which are important in achieving our goals. Activities in these zones will be highly tracked, and anything that happens outside is less relevant. This can help us create a relatively noise free data environment and focus on the insights that really matter.