We all know the story. Girl stumbles into a house, a bit peckish and – lo! – finds three bowls of porridge: one’s too hot, one’s too cold and one’s just right. After sampling the first two she sits down and tucks into the third. To a greater or lesser degree anyone working with content has been that girl, stumbling around and looking for something to satiate our hunger (or, in this case our need to understand our content through better editorial analytics. The metaphor had to break down at some point).
Stumbling around in the dark
It’s not that analytics went unseen until Google decided to start devoting time to its Google Analytics tools, or the various newspapers starting utilising those oft ubiquitous real-time boards in the bullpen.
There have always been ways to understand audiences: focus groups, letters to the editor as well as straight-up sales.
The realisation – which struck around the midnoughties with the dawn of Big Data – was that suddenly there was the capacity to delve into heretofore uncharted waters and unpick audience behaviour with the kind of precision and nuance previously only dreamt of.
It became possible to measure a vast amount of information. But, while the promise of limitless information was fulfilled, the delivery mechanism remained less promising.
In practical terms all that data had trouble finding a workable interface. Despite the proliferation of information, the overwhelm was the issue.
While the potential for harvesting reams and reams of data was being explored, social media giants were using likes as their own kind of measures of success. In newsrooms, the click became its equivalent.
With the emergence of Facebook (2004), Twitter (2006), YouTube (2005) represented a huge cultural shift, where it wasn’t just companies who could find out how effective their marketing campaigns were, but private individuals getting the grips with a new language of popularity, and new barometers with which to measure it.
We’ve reached a point now where the equilibrium is being found: it’s possible to balance exhaustive forays into big data with delivery mechanisms which retain simple metrics’ simplicity, without losing nuance or specificity. At the International Institute for Analytics, this new phase is termed ‘Analytics 3.0’. We call it content intelligence. Whatever its moniker the premise is the same: it’s about extracting meaning from the sea of data and applying it usefully to your individual business.
So, what are the things to consider in the search for that sublime
bowl of porridge analytics combination?
The first issue is that of mindset. Because our familiarity with data has developed in tandem with the rise of social media and the kinds of metrics we’re familiar with in our social lives, it can be difficult to think outside this paradigm. But extrapolating those values into a business setting just doesn’t work. Successful, progressive analytics work from the other way around: they look at the kinds of answers newsrooms and editors need and consider solutions.
To take an example, clicks and likes might be familiar, but what are they able to contribute to the editorial process? Are they able to inform strategy or recognise outlier content? What questions do these metrics answer? How many people read your article? If that’s the case, that’s not a answer: examining the property of the click it’s clear that the only thing a click tells you is if someone has clicked on an article. If you want to know that it’s been read the questioning needs to be better. How much did your readers read? How far through an article did people get? How much time did they spend on it? Whatever variant you select, already the focus on attention is more productive than that of a browser action.
Secondly, consider your audience. What kind of things do they read? How long do they spend on your site? Are certain readers more valuable than others? With that last point, it’s a bit of a trick question: of course that’s the case. Amedia make this point well: they’re a Norwegian company. Likes, hits and shares from New York are less valuable to them than likes, hits or shares from the Oslo suburbs. If you’re the marketing department of a tech company, the number of article reads or page follows that can be attributed to your friends and family is nice, but are those readers as valuable to you and your business as a tech startup well suited to your product? Of course not. Analytics might not be able to tell you exactly who’s reading what, but by tracking user behaviour you can get a pretty nuanced idea of where your reader loyalty sits.
Just like Goldilocks, try the options and find the best fit. The ‘just right’ solution is out there, and it’s considerably less dangerous for us than it is for someone embarking on a bit of Bear Home Invasion.