Lessons For A Career In Analytics From A COVID-19 Briefing
Warning It Won’t Be Pretty…
I started watching this video because it was shared in an online debate. I am sharing it now to educate aspiring analysts looking to build a career in analytics. It does not matter which side you are on… if you can’t see past that, don’t pursue a career in analytics! It isn’t for you. You need to stay objective.
I am including the video below. It is nearly an hour long, but watch it! If you can’t watch a one hour video — don’t pursue a career in analytics! It isn’t for you. You need to have focus. If this bores … you get the idea.
This video comes about as close as you are going to get to a corporate debate in a boardroom. Assuming — that boardroom is filled with a leadership group that is not exactly thrilled with the findings being shared by experienced and honest analysts. Doubly so if they are calling into question recent executive decision making. This happens a lot — for great analysts. Not very often for the less than great variety (but I covered that yesterday). Here are more than a dozen quick points to consider. It is your choice as to whether to read first or watch first.
DATA
Presentations that are contentious typically start with a presentation of data. Answer First is nice, but never goes over well when the audience has a predisposed opinion. This conference even gives us disparate data sources, something more typical of the business world. You will note that data without recommendation or judgement sure starts to feel like it has both. Watching this, you can hear the discontent before the questions even start. That IS the power of data.
PROCESS/LOGIC
With the statistics on the table, it is important to lay out the processes involved. If you don’t have clear understanding of the process, the data is just noise. This is typically where the audience, who also often has a sense of the process, begins to object with passive but probing questions. Know your craft well. If you botch a process question — it will haunt you. No — probably undermine you completely.
Also — be careful about how you discard thing you find non-relevant. I think in this example — Dr Erickson misjudged the “testing for all” question. It came back numerous times like an annoying fly. It likely unhinged him more than anything else in this debate.
BENCHMARKS
Comparisons and benchmarks are critical, they will draw more poignant objections, but don’t let that stop you. It would have been much better for these two gentlemen to have a little more statistical back-up than Sweden is like Norway. Some demographics, densities, etc could have tied that up a little more nicely.
As for the Flu benchmarks? More numbers might have helped drive that home… but the deck was likely stacked for them on that one. It is hard to use the very benchmark the other side has already co-opted. This is NOT the flu… analyst beware. I would have buttressed that argument by helping people better realize they underestimate the flu. This does reappear later but without numbers, just behavioral arguments.
TIME
This argument enters every analytic debate. No matter how educated your audience, they will NOT understand the time sensitivity of findings or how they evolve. Get over it — the world thinks this stuff is wicked easy. Run a test. Read results. Test everyone. What is taking so long? … sigh
COUNTERPOINTS — ANECDOTES
Breath… and get ready. Here come the counterpoints. It is all starts with anecdotes. Someone knows someone, saw something, or had a moment and it doesn’t exactly align with what you are saying. It does support their own confirmation bias. So now you need to hear about it. Try Artin’s approach — get them to quantify it… likely they can’t.
COUNTERPOINTS— TESTING
Anyone can run a test. Unless you are in an organization where they tried that theory and then chose to hire you, you will get this a lot. Everyone wants more testing when the current data doesn’t align with their preferred outcome. They will still want more when the test don’t agree either. Are you sure you ran them right? I heard from my uncle that someone else is running them better…
COUNTERPOINTS — BUT SOMEONE ELSE SAID
Which brings us to the next and final counterpoint — it is Appeal To Authority. It starts with — “but someone else said!”. Soon it transforms into — “are you saying so-and-so is incompetent?”. These guys did a great, almost cringe worthy job of deflecting that. And you HAVE TO deflect that… if you get trapped into the reality, which is “yeah, maybe!” — you will lose the battle. Stay calm. Stay polite. Don’t blame!
CONFLATION — CO-MORBIDITIES
You will likely never discuss co-morbidities in most boardroom settings, outside the medical/pharmaceutical fields. But co-morbidities in this discussion is analogous with conflation. Things do not have single causes very often in the world of business. This is going to cause incredible pain. It may be the most frustrating part of the day.
TONE
If you listen carefully to this discussion, the folks asking the questions have their tone soften markedly across the hour. I don’t know that ANY of them left with their minds changed, BUT they did at least show some acknowledgement that this was a well-played argument on the part of our two “analysts”. This is recognition of them sticking to solid fundamentals like data, benchmarks, and no finger pointing.
PATIENCE
It is probably obvious, but you can only do this if you stay patient. You can’t get frustrate. Which brings us to more contentious observations…
THE SUPPORT GUY ALWAYS SOUNDS SMARTER AND MORE CALM
This so true! You can see this unfold here. Dr Erickson almost loses it a few times. Artin (Dr Massihi) gets to run support. That role is always easier. I almost guarantee there were times in this hour when Erickson was thinking “damn, why did I take lead?”
FRUSTRATION AND PRINCIPLE
As an analyst, you always need to fall back on principle when things get tough. As a presenter, in this environment, if you fall back on principle — you are going to lose. Your audience will see that as a sign the data isn’t holding. You are simply being stubborn, contrarian, or partisan. You must avoid it.
On a related note, as the analyst — you can’t counter your audiences objections with anything other than data. You are the prosecutor in this “legal system”. You have a higher burden of proof. Don’t point to hypocrisy — stick to data. No the other side didn’t do that … but the decision was already made. They get to be the defense. Accept it. By the way — Artin is masterful on this.
THEORY VS REALITY
This is another really hard and frequent point of contention in so many of these debates. But the models say! But the prevailing wisdom is! All theory. All simplified, sanitized, and invented. Reality isn’t always easy, fair, or simple BUT it is reflected in the data and measurable via probabilities and statistics. This is hard point to get across without belittling and finger pointing. Kudos to this duo. Though after they felt like they won this — they fell back on it a few times they should not have. So tough…
ALWAYS REMEMBER, THE AUDIENCE IS DOING THEIR BEST
Reporters, executives, leaders, or the general public — don’t lose sight of the fact that most of these people have good hearts and good intentions. They aren’t staring at data each day. They aren’t running tests. And they aren’t the ones being paid to make the argument. For that matter, they owe it to themselves and their company to challenge anything that seems “off” to them. Put simply — they are just doing their jobs!
Now, you may want to use those same arguments. They aren’t staring at data each day. They aren’t running tests. And they aren’t the ones being paid to make the argument. To ask — why won’t they trust the guy (or gal) that is??? It doesn’t work that way. Maybe when analysts die — there is a special heaven where people trust you? But what fun would that be?
As a final matter, there is always the possibility your data is wrong. The goal of any analytic endeavor is NOT action — that is something sales people and book authors sell you. The goal of a meeting like this is education and learning. If the other side makes strong points, embrace them and follow-up. The goal is not to be right for the sake of being right. It is to be right (or wrong) for the sake of learning. I would love to tell you that being wrong politely and with analytic discipline will save your job… it may. It may not.
Thanks for reading! Stay safe. More important, stay informed.
./
