Musings on designing experiences & (re)engineering complexity

May 2022

Data Analysis is Weird

Something of an “off the cuff” title to this, but in thinking about a few items which have been in the inbox lately, it kind of makes sense.

For example, one of the items in the inbox is of a session am currently conducting where folks have more or less needed to learn the basics of a specific analysis tool. The customer wants to move towards targeted trainings for specific features, but hasn’t yet identified what data points make the most sense for realizing success. If you will, if the point is to improve on a tool, the tool should enable better decisions for the business. However, those better decisions haven’t yet been pulled together, so the expertise in the tool goes where? Exactly?

Another item in the inbox, really a few items, points towards the “tools for thought” genre. This is a genre of tools and methods by those who synthesize, analyze, or remix data for various applications. And similar to the previous paragraph, often what happens in this space is the elevation of a tool or method until it becomes the floor for the next thing. A few interesting approaches always (Flatfile, SketchSystems, Einblick), and Hepabase as some of the latest) but the methods and discipline to create one’s own shape-of-thinking is the sensemaking bit which often seems missing.

And maybe that’s where some of this feeling of weirdness comes from. When one boils down a lot of knowledge work, outputs are really made of “counting and thinking with shards of creativity and ego.” So, if you can help folks ask better questions, there’s often not much more than something to scribble on which is needed to discover the solutions.

And in a similar frame, the automated insights which many software and services companies call artificial intelligence really amounts to nothing more than filtered calculations. It seems intelligent because the audience is either oblivious to the questions which made the calculations, or speed at which the number of questions were turned into answers amazes. It feels magical, when it is simply a faster view to the same wide world. In other words, weird thing causing appreciation.

So if this is the case, then what? Is data analysis really an admission by a certain body of people about the weirdness of the questions they couldn’t yet ask? Is the data really all that valuable, or the ways in which we think of it? Or, are we simply marveling at the limits of being able to reach into a box, so when the machine pulls it out there’s a “jack in the box” like surprise our dopamine receptors are responding to?

Weird, right?

Or maybe, just maybe there’s an overarching simplicity not just to the questions, but to the data itself. And as we canvas our respective inboxes, we open ourselves to the potential surprises of there being more to discover the more we are connected to. Which is cool. And weird.