An interesting discussion came up on Twitter, why would you use a dashboard?
It’s true that dashboards only present a limited view of data, so have the potential to be misleading. For example let’s say you created a dashboard a few years ago showing how happy people were with your telephone service. Since then you’ve beefed up your website and now most people use that instead of the phone to access your services. If your dashboard still only shows the telephone results then you are potentially missing out on what most of your users really think because they are going through the website (this assumes you are using the dashboard as your main source of information).
Also dashboards tend to be very brief in their reporting of figures, maybe a handful of figures presented in simple graphics. This has the potential to hide detail. Does the satisfaction rating of 75% mean that a lot of people are very happy and the rest are moderately happy, or does it mean there’s a lot of happy users but a significant minority of absolutely furious users?
This illustrate the dangers of relying solely on a dashboard. So why would you use such a potentially misleading tool?
A well crafted dashboard can give you a quick overview on some areas that you want to track. Let’s be honest, most of us don’t have the time or inclination to delve into all the data we have on our organisation on a regular basis. However most of us wouldn’t mind seeing a quick summary of how things are going once in a while. This happened yesterday, I wanted to know how the visitor stats for the Data Unity website were looking. Was I motivated enough to dig out the log files, load them up and do some aggregation? Nope. But I was happy enough to pull up the Google Analytics dashboard over a tea break to get a quick overview. If there’s a choice between a bit of information and no information, then I’d go with a bit of information. It might not be detailed, it might not give a complete picture, but at least I’m a little bit more informed than before.
Can we protect against the case where dashboards are hiding detail? Open Data can really help in this regard - allowing users to ‘click through’ (or ‘drill down’) into the results which make up the pretty graphic on the dashboard. We could easily get a better picture of the 75% figure we mentioned above if we let the users click through to a bar chart which shows how many users were dissatisfied, neutral and satisfied. Why stop there - could we give the user access to the raw data so they can make up their own decision (assuming the data is made anonymous)? In this view, the dashboard is at the top of the data pyramid and with Open Data we let users go down as far as they want.
At the end of the day, a dashboard shouldn’t be taken as the true picture of an organisation. Any real life organisation is going to be too complicated to reflect in a few pretty pictures on a screen. However a dashboard can be a handy overview and a starting point for insights into the data, not the final destination.