Zen, sticks, ants and analytics

March 14, 2017   |   Digital Marketing
Nudge Digital
Nudge Digital

A famous koan has Master Shuzan declare - If you call this a short stick, you oppose its reality. If you do not call it a short stick, you ignore the fact. Now what do you wish to call this? As a koan, the correct answer is not to attempt an answer, but there's still a lot to learn from old Shuzan.


How can something be two things at once or have two different meanings? In the digital world, where everything is measurable, we want website performance to be something we can nail down and talk about with confidence. The idea that a website KPI might have a reality that opposes its own fact is at best disconcerting and, at worst, a pain in the backside.


So what can Zen buddhism teach us about web analytics? This occurred to me a short while back, as I found myself renewing my Google Analytics certification around the same time I completed an analytics review for a client.


What struck me was the highly reductive way that Google approach the subject of web analytics. The Google course teaches how to set up reports that tell you about your site's performance. This can be very powerful. If you have an ecommerce site, monitoring where users fall out of the checkout process is invaluable; if you have a site to deliver your content strategy, measuring what articles are being consumed, for how long and in what depth is an absolute must.


However, the assumption underlying this is that a metric means what it appears to mean. Or that its reality does not oppose the fact, if you will.


The analytics reviews we provide at Nudge are a little different. Instead of looking at individual elements of the site's performance and simply measuring them, we try to look at combinations of metrics to learn about how the site is being used.


The difference here is reductionism versus holism. If you're not familiar, the former reduces systems to base components and looks at each to understand what's going on, whilst the latter takes the view that those components can only be understood by looking at the system as a whole.


In his seminal book Godel, Escher, Bach: An Eternal Golden Braid, Douglas Hofstadter neatly captures the difference between the two by differentiating between an ant colony and the ants within it.


Ants aren't individually intelligent but an ant colony is - yet ant colonies are made of ants. You can't see what an ant colony is doing except by looking at the actions of all the individual ants. And you can only make sense of what individual ants are doing by knowing what the colony overall is trying to do. Looking at only one level will give you information; to understand it, you need to look at both.


The same applies to web analytics, to an extent at least. Trying to isolate metrics too much and then draw solid conclusions from them can be risky.


Let's say you're measuring pages per visit: if it goes up, that means user engagement has improved, right? Perhaps. But what if average visit duration hasn't increased? This means that users are actually spending less time, on average, on each page. In that case they're actually engaging with your content less deeply than previously. If users are spending less time per page but viewing more pages, are they more or less engaged?


The only way to discover this is to look at what pages they are looking at, whether that's a change from previously and whether any changes you've made to the site might have made a difference. In other words, to avoid opposing the reality or ignoring the facts of your pages per visit data, you need to remember that you can't understand the whole colony by looking at just one ant!


If you want help understanding your web analytics and turning information into insight, please get in touch.