Good wasn’t it?
For those of you that didn’t do the background reading, HappyOrNot terminals provide organizations with real-time feedback on employee and customer satisfaction. They have an intuitive design, with colour coded faces capturing a range of emotions from Happy (green, smiling) through to Angry (red, frowny). Simply make a chance by hitting the button, and keep walking. I’d seen the HappyOrNot terminals in action at Heathrow Airport (as have many people according to the profile); I nursed a coffee while watching a woman pace back and forth in front of a terminal slapping at the big red face to indicate her general dissatisfaction with something. She was not having a good day at all.
As you know, at interstitio we dig all things organizational measurement and measuring, and love popular descriptions of novel tools and approaches. There’s always something to learn.
We particularly enjoy examples of what you might call ‘friction free’ measurement. Don’t get us wrong, we love a good survey, moreover we pride ourselves on being able to design efficient and effective survey vehicles to get at your issue, without having the instrument having to be a 72 page, 300 thumb scroll, “experience” … but we all know that irrespective of how well a survey is designed, it’s intrusive. There’s a cost. A friction, if you will, associated with completing it.
Enter ‘friction free’ assessment.
We’re going to use this term to mean when an action is so, so, so, easy to make as to have no discernible cost to the individual (i.e. pushing a button or ‘liking’ something on Facebook), and when the action itself can be made without you having to deviate to radically from your day. The HappyOrNot terminals are wonderful examples - single choice, simple button push, without a break in my stride as I head off to something else.
The use of metadata, data providing information about other data, can also be thought of as a ‘friction free’ assessment. From the inference of communications patterns from metadata embedded in email / IM / SMS / phone logs. This data exhaust is particularly powerful as it’s connected to a behaviour - the sending of an email / IM / SMS or the placing of a call. There’s an intent to communicate, with a specific actor, about a specific topic, at a particular time.
We know that Facebook likes are predictive of ‘private traits and attributes’, and stock market activity, but, as part of some recent research, in collaboration with some our academic friends we have demonstrated that demographic and personality factors can be reliably and accurately predicted from only email subject line data. This is a remarkable finding, because while the action of ‘liking’ is a seemingly small gesture - it is attitudinally oriented towards the object being ‘liked’. Our recent work on email subject line data suggests that your way of being, through text, in something as simple as a description of the remainder of the content in an email message, when looked at in aggregate, contains enough of your essence as to be predictive of your personality - and all that goes along with that.
How we interact with our devices, and the trail of metadata that can be interrogated and connected to behaviour is of great interest to a lot of people (see this recent NYT article on firms exploring possible connections to physical and mental health).
Of course just because we can, doesn’t mean we should. While we appreciate the ability to bring objective analytic tools in the service of an organization’s operations, we don’t do it lightly, or with the intention of creating a surveillance state! Appropriately managing the ethics of data collection and analysis is a much broader topic for another post. Until then dear reader, Thanks for listening.