The importance of asking the right questions
We live in a world that values answers. We were taught in school to learn how to answer questions in exams, we were conditioned to go to work knowing that we need to have the answers and our society, by and large, focuses on finding the solutions rather than figuring out if we are asking the right questions. Just like most people who have been through the traditional education system and started working in a corporate job, I was trained to have the answers, I was taught that my contribution and value lie in my ability to solve problems by knowing the right answers. While I do think that problem solving and the ability to find the right answers is a valuable skill to have, I would like to shed some light on the importance of the skill that precedes it, the skill of asking the right questions.
Questions and answers are by definition linked together but they are a very different skillset. Seeking answers is a process of elimination through research and experimentation, trying to piece together different information and narrow things down to a solution. But asking questions is a process of expansion through critical thinking and imagination. It is understandable why as a society we don’t value the cost of asking the right questions because in some way the more questions we ask, the more work we need to do and the further away we are from finishing what we need to do. This creates a systemic problem that favours short term patches over long term solutions.
This problem exists everywhere, but it is especially prominent in a solution-orientated field like data science and engineering, where the professions are built on top of solving difficult problems. Data scientists are prone to jumping right into solution mode before we even fully understand the problem because we think we have the answers. If there is a conversion problem, we can do personalised targeting. If there is a retention problem, we can build a churn prediction model. While those might all be valid solutions, they might just be addressing the symptoms and not the root cause.
It’s probably natural to think that a solution is a solution even if it only solves the symptoms. Especially in a world that favours immediate actions and quick fixes, we never seem to have the time to dig deeper. I am not going to argue that for every single issue that comes up, you need to deep dive into whether you are asking the right questions, but I will stand by the fact that if we are asking the right questions, more often than not, we won’t be in situations fixing problems that shouldn’t come up in the first place.
In the context of data science, the cost often comes in two main ways, wasted resources and unintended consequences. The first one is relatively straight forward, if we don’t ask the right questions, we end up wasting time and effort building a solution that is not fit for purpose. The second one is more nefarious because if we are optimising a system based on a wrong metric, we could be systemically worsening the situation. For example, if we treat a churn problem as an isolated incident and focus on improving retention without asking the question of why people are churning, we might miss problems that go back to customer acquisition, through to user experience and engagement before it becomes a churn problem.
There is value in having the right answers for every question that comes up, but the cost of not asking the right questions is a bit more subtle and long term. If we always prioritise and value quick fixes, we are subconsciously encouraging problems to occur.
