Kevin Lehrbass of My spreadsheet Lab posted this video Sunday night and there are aspects of it that need to be highlighted while it’s still hot out of the oven.
PEOPLE PROCESSES TOOLS
In keeping with a recent theme of People-Processes-Tools, Kevin’s video is a superb example of the People component of that trinity: the wary analyst.
- Kevin doesn’t show us a doggone thing about HOW to recreate what he did. And that makes the video so much mo’ better!
- The entire 5:29 is focused on the thought process and insights
- How can “close game” be defined?
- How can we visualize the multiple legitimate ways of defining “close game”?
- Are any surprises uncovered? Any intuition that was proven wrong or supported?
- Does a close final score correlate with an exciting game or, were there a good number of dull one-sided games with amazing rallies at the very end.
WHY THIS MATTERS
“Interviewing for Excel skill” is a hot topic and I think we’d do better to find out if people can think in ways that Kevin exhibits in the video. The actual doing can be taught. The fact that it’s in Excel isn’t so critical, either.
Kevin sets up the challenge, recognizes that there are multiple interpretations and presents them. That’s being a good analyst and responsible data steward.
REAL WORLD APPLICATION
Let’s say you have a business that sells and ships products (books, candy, helmets, whatever). Now you want to analyze shipping costs and methods. Kevin’s example can be extended to have us question: how do we define ‘shipping’?
We should consider:
- Should we include orders where shipping was refunded?
- What about free-shipping promotions?
- Is there a way to flag and separate refunded shipping and promotional free shipping?
- Are we only interested in customer shipping or also shipping between our own offices?
- Are there parts of the ordering process that will make the data messy, requiring cleanup before doing the analysis?
The answers to those questions are, “it depends on why you’re doing the analysis.” The point is that we need analysts who ask those questions. If they don’t, they run the risk of doing a bunch of math that presents a single and wrong view of reality which then:
- Leads to a horrible decision or,
- Results in indecision because the results just don’t feel trustworthy.
I see the indecision scenario a lot! And the next move tends to be to look for a better tool or put the project off for another day. For a decision-maker whose expertise is not in data analysis, this is a sad sad situation. The decision-maker needs to be able to lean on a wary analyst.
ANALYSTS: THE GOOD, THE BAD, THE NEFARIOUS
Good analysts may not be so good at Excel immediately, but they have the intellectual curiosity. They try to get at the purpose of the analysis so that the right data can be gathered and analyzed in the right way.
Bad analysts sometimes have the technical skill, but they just do the math.
Nefarious analysts will purposefully create confusion and give you 100 different slices of the data. Those types of analysts are going to Hell for violating the 11th Commandment:
Kevin Lehrbass did an excellent job with that video. On the surface, it’s fun, smart and fascinating. At a deeper level, it’s an example of the type of thinking that we need to identify in potential analysts. Because the video doesn’t get technical, it doesn’t distract from the concepts. Thus, someone who needs an analyst should be able to follow along and not ask, “do I have someone who can DO this?” The better question is
“do I have someone who THINKS like this?”
Phantoms of the Brain image by richworks