Here's one of the simplest ways to simplify complex analyticsThis design "no-no" appears almost every time a new client [with a product that displays analytics] asks me to review their UI/UX.
More often than not, I'm not provided with any relevant user tasks/usage contexts by which I can do my evaluation, but clients still want my opinion on what could be better, or what they're doing wrong.
Inevitably, the UI will have some KPI (key performance indicators) presented on the screen, perhaps as big numbers or histograms. The data might be I/O operations per second. Or, number of new subscribers. Or, how much energy was saved last month. However, even if I don't fully understand the domain or KPIs, there is often a very good chance that something is missing in the UI that could make the UX better for the customers intended to benefit from this information:
Providing useful, neighboring comparisons.
Did you know our US National Debt as of this writing is $19.4 trillion dollars?
usdebtclock.org is a fairly ugly, but interesting website which gets a few things right, and a few things wrong. Here's a grab from one corner of the page that we can break down for a moment. For now, I'm going to ignore the interaction design, improper display of numerical data, and aesthetic issues with the site and talk for a moment about how they did and didn't use comparisons effectively:
- The national debt was broken down to "per citizen" and "per taxpayer." $19T is hard to fathom, but $59k/year is relatable. It's a decent, middle class salary in many parts of the USA. That's a useful comparison that just took some division.
- They could also have used a proxy for $59k, perhaps substituting a piece of merchandise such as "a new, mid-range BMW." How might the information feel more relatable if somebody told you the debt was like "every citizen in the US having a 100% unpaid loan on a new BMW they can't afford, and every taxpayer having 2.5 BMWs"? Are there places in your UI where a proxy might help people understand the information better?
- Remember how I mentioned "neighboring comparisons" above? There's something very fascinating about the national debt being 6x the federal tax revenue. However, this analysis is lost because the revenue and debt numbers were neither placed near eachother, nor were they visualized (designed) in such a way that one appeared 6x larger than the other without doing mental math. This is a fantastic example of the difference between data and information. You have to apply design to turn data into useful information.