are the ones asked at the beginning of a project! That is, if you're smart. Far too many analytics projects fail due to a lack of a robust process of framing the right questions at the outset.
In the words of Peter Drucker,
“There is nothing quite so useless as doing with great efficiency something that should not be done at all.”
Inexperienced managers often act under pressures of execution and rush through what is possibly the most crucial factor in determining the outcome of an analytics effort:
Clearly articulate what question the analytics project is meant to answer, and rigorously examine the question to make sure it is valid, relevant and impactful.
Why is this so difficult despite seeming so simple and obvious? There are a few reasons.
First, this process requires diverse perspectives, an environment that fosters constructive conflict.
Second, it needs time and dedicated focus.
Third, the questions need to be framed in the context of well understood goals.
Fourth, the goals and questions must be of immediate importance to key stakeholders. Does the organization have the willingness and ability to make a change NOW based on the answers to the question?
If this process is run properly - with the right alignment, careful selection of participants, and proper focus, framing of analysis questions can be a very enlightening exercise that avoids considerable confusion and wasted resources down the road. Don't skimp on the questions!
Beware of the technology trap
Another common malady is to succumb to the swarm of jargon, technology and infrastructure that is so prevalent in the Big Data ecosphere, without a clear idea of your analytics goals.
Paraphrasing the exchange between Alice and the Cheshire cat, "If you don't know where you're going, any road will take you there!" Big data and analytics technology is a lot like this.
I have seen teams spending inordinate amounts of time debating over the merits of various platforms, tools, and technologies without first having clear alignment on their analytics goals and driving questions.
As we'll see, the process of framing the right question often influences:
a) the type and depth of analysis required,
b) the data sources and hence the technologies needed,
c) the intended audience and visualizations that are expected.