Analyst Interrogative #2
How to figure out how?
In the first article of this series, we stated that using How was critical to asking a constructive analytic question. How then is the key to asking a constructive question. But what is the key to How?
It starts with framing.
How do we define framing? Using etymology we can turn back the clock to the 1200s, then look to Old English, German, and Scandinavian. The former defined frame as a composition, plan. The latter had frami — meaning advancement. This knowledge quickly colors our understanding of framing.
Framing represents a process that can be employed for both analysis and synthesis. In synthesis, this takes the shape of a structure, a plan for organizing, advancing, and eventually modeling data. In analysis, the plan is one for decomposition or loosening(et.). A suitable analogy might be an architectural plan, the building’s frame (unsurprisingly), or even the scaffolding which rings the building.
Framing the problem may be the most critical step in analytics.
It is certainly the first. But often, it is also the second, third, and more. Framing a problem is based on one’s level of understanding. That level is determined by your understanding of the core question, the acceptable assumptions, and the practical constraints. If that level of understanding is not high enough, framing becomes iterative.
Iteration, or reframing, is a required practice of every analyst. Each iteration peals a layer of the onion (another analogy). Each step provides the understanding needed to advance the framework toward our final well-framed structure.
Let’s Step Back
Since I touched on etymology, analogy, and iteration, it seems only appropriate to bring in another favored technique recursion. We have stated that using the interrogative How is a better practice for analysts because the questions will be more constructive. We followed by offering that How questions often lead to an exercise in framing. But How do we frame the initial question?
The answer begins with the other interrogatives. Try to incorporate them into your question framework. Try to think like a journalist or 1950’s style detective. What do you want to accomplish? For who? Are there constraints on where or when? Feel free to write a beast of a question, you can synthesize it later (or not).
How do I create a model to simplify the major demographic insights from the portfolio of customers who subscribed to our new widget program in a way that will help marketing adjust the voice and content of our early communications so that it is more relevant to those customers?
This question is nowhere nearly complete. Hopefully, you are asking yourself questions like what kind of model? and what type of communication? Or even better, questions like How do I define relevant?, How do I determine the appropriate model?, and How might marketing’s communications be constrained by timing, channel, and frequency?
A well-framed How will become both your starting point and your basis for iteration. In the beginning, this will take a good deal of time. As you, your team, and your partners develop, you will quickly find that old frameworks can be easily borrowed for new challenges. So perhaps scaffolding was indeed the best analogy?
Why not Why?
Tune in for another exciting addition of Analyst Interrogative
Analyst Interrogative is a series created by Corsair’s Institute to increase the reader’s understanding of the analytic process. For more articles like these — please subscribe to Comprehension 360.
