Workshops. On data. BIG DATA, even. What fun!
As much as I acknowledge the value of an education in statistical methods, it’s one of those things–like insurance and economics–that skims right off the surface of my brain and refuses to stick.
And yet, today I attended a statistics-based data workshop.
Most of it was an overview of the basics with an eye toward why you use certain techniques in certain situations and the ethics thereof. Kind of refreshing, actually.
Until we got to a section on dashboards.
Data dashboards, which seem to be big trend amongst the data-product companies these days. I know that in my area, there’s rumors that there might be dashboards in the works, and everyone is excited.
Everyone except our workshop presenter, that is.
She was not impressed.
Don’t be impressed by words like “dashboard,” she says. Look, here’s what a dashboard looks like. All that’s on it is some bar charts, a pie chart, and some recent history. There’s nothing new here. All of this can be made in Excel with minimal effort.
(Yes, this is true. Most dashboards aren’t the pinnacle of cutting edge data visualizations.)
I leaned over to the grad student sitting next to me and remarked, “But they’re not selling innovation. They’re selling convenience.”
CEOs must think dashboards are magic because they don’t know how they’re assembled.
I’m no data expert, but both of these attitudes strike me as out-of-touch.
Dashboards ARE about convenience more than anything. Would I love to have a financial dashboard that shows me a simple pie chart of budget categories and maybe a bar graph of actual versus projected expenses? You betcha. Because right now, to get that information, I have to log in to a remote server, run a report, clean up the data and hope it plays well with whatever version of Excel I’m running, and then create a few visualizations.
Is this difficult? No. Does it take time? Yes. But unlike academics, who tend to focus on their specialty like it’s the only thing that matters in the world, I neither have the time nor inclination to spend so long working on graphs that a dashboard could spit out at me–updated and in real time–in moments.
If that’s true for me, who is no longer at the bottom of the heap but who’s still at the “works with reality” end of the hierarchy, that’s so much more true for a CEO. And for someone at that level, having realtime reporting of actual data in the company is imperative. They have to have such a high vantagepoint that the data is critical–for me, I have a pretty good handle on how the budget is being spent because I reconcile all the purchases myself anyway. CEOs don’t do that.
Even so, I have a hard time comprehending how someone could possibly make it to CEO of a company that has the money to throw at a data-product company for a dashboard in the first place who didn’t have to fight his (or her) way up through the ranks and I would be willing to bet money that any of them could make a stupid bar chart.
Elementary school kids can make bar charts. It’s not hard.
But I forget how many people, especially permacademics, are brainwashed by the media, where the CEO AS BUFFOON narrative is in full force (see also: Trump). Even if you’re aided by a “good ol’ boys club,” you’re not going to make it long as CEO without a good bundle of smarts.
Enough about CEOs–back to dashboards. The remaining thing that entices me about dashboards is their (supposed) ability to bring together multiple databases worth of information. In all the universities that I’ve worked for, there are various databases that hold different types of information, all of which are extremely robust and none of which want to talk to each other. If–and I am aware that this is a big if–it is possible for a data-product company to create a dashboard that incorporates one or more of those data sources, that company would earn my undying gratitude and admiration.
Seriously, it’s awful having two robust sources of data with no way of bridging the gap. It makes you look so incompetent to people who request certain types of reports but don’t know the lay of the data landscape, so to speak. (Would you call that “data architecture”? Don’t answer that. I’ll google it.)
I feel like these people who get so caught up in process and method (which is their job) forget that there are so many other practical concerns, like data sources, time, and the fact that most of us want a computer to do any and all jobs that computers can do better than us, because it cuts down on human error.
I will be absolutely ecstatic when a computer can take over most of my budgeting duties. (I’m sure at some Fortune 500 companies, they already do.)
Maybe data dashboards don’t reinvent the wheel, but they sure are useful.