Asking Better Questions
Specific Thoughts On The Questions All Analysts Should Be Asking
Nearly three years ago I wrote this article:
It focused on “how” to build an analytic question. Analytic questions are better questions for reasons we outlined in that article. But a good analytic question is not necessarily the “right” question. Or the “right now” question…
“How” is the “right” word for learning. Analytics being mostly scientific learning, it is always the right place to start. But analytics, at least the actionable sort, is also decision-making. No single word is likely more important to good decision-making than “priority”. Fortunately, priority can always be assessed with a “how” question (or two) of its own.
At this point, lets take a short aside to also remember that analytics has two more favorite words — iteration and recursion. The latter is about to come into play.
You see a good analyst first formulates a great “how” question. One that is testable, quantifiable, and actionable. BUT, before they do anything else, they need to ask another question entirely.
How will answering this question impact the bottom line?
Followed by:
How does that compare with other questions I could be answering?
It can be stated with utter certainty that these should always be the first two questions any analyst or organization should attempt to answer. AND that these are not actually the “right” two questions to ask. We will use the remainder of the article to explain “how” that makes perfect sense.
Let’s start with why these are indeed the “right” questions.
The answer lies in priority. Many a great question was contemplated only long enough to learn it was the wrong question entirely. Douglas Adam’s Hitchhiker’s Guide To The Galaxy made the number 42 infamous as “the answer to the ultimate question”. The problem being — no one knew what the actual question was.
Business is typically far more complicated than fantasy fiction. Embedded deep in the concept of priority — is the concept of opportunity cost. Every time an analyst opts to take on a question, they are also quietly opting not to take on many others. This can get both complicated and, to some degree, philosophical. But in reality, it is rarely either. The shear beauty of why these two questions are so important is that they are SO often — never asked!
This is one of those situations where you can be a greater analysts simply by trying… Getting the answers “right” is only a marginal improvement over simply asking them in the first place. This plays in nicely with why great analytics is iterative and our earlier emphasis on decision-making. Being sure to always ask our two questions is an iterative approach that benefits from making many small decisions as opposed to a few giant ones. The former enables learning and adaption — the latter is all glory or total failure.
Now — as to why these are not the “right” questions after all…
Both learning and analytics are predicated on the idea of simplification. Writing 3–5 minute articles about those two subjects — doubly so. A great analytic question, on the other hand, is typically wordy and filled with pointed but often endless details.
I did not give you the latter. Instead, I gave you a cliche’. The latter can not be given out with any sort of absolute certainty. The root to any business’ or organization’s “bottom line” is somewhat subjective and very much dynamic. It might be a matter of revenue, cost, profitability, or even persuasive impact?
There is also the further complicating matter of timing. It is never just about “is it right?”. It is almost always about “is it right, now?”. Of course, every business is operating on differing time frames based on their industry, their life cycle, their scale, and their cash flow. This secondary issue of timing is also why our second question is over simplified as well… though I did spare you additional cliche’.
To be fully transparent, there is rarely a clear answer to these two questions. Just as simply asking them may be enough, getting them perfectly “right” may well be impossible. The criteria for doing so is likely to evolve with the organization. The key then is establish them as a process. Just like that — learning and decision-making become an integrated, iterative, and often recursive process… much like these articles.
And so , there we have it. Thanks for reading!
