If you have read Douglas Adams, you know that the ultimate answer is 42. However, you also know that no one quite knows what the ultimate question is! Most people starting out in analytics fields are excited about the chance to run sophisticated models, the opportunities to learn new tools, the possibility of demonstrating depth of knowledge based on years of stats courses. However, they tend to forget that in analytics, it is the business problem that drives all effort, and no amount of sophisticated models or advanced analytics tools will help unless the question has been framed correctly, and the answer derived using the right tools and data makes sense within a business context.
A key skill of a successful analyst is the ability to translate from a business context to a mathematical context and back. Only by framing a business problem in the correct way can we assess the right approach for addressing the issue, and understand if the derived results will help solve the problem. And only by translating the results back into a business context can we actually help the business implement strategies to address the problem.
Domain knowledge is therefore a very critical requirement in building a long term analytics career, as important as technical skills. For example, clustering is a technique that allows us to identify homogeneous groups of characteristics that drive behavior. However, clustering results and interpretation will differ widely depending on whether we are working to identify customers for a direct marketing campaign for credit cards, or are identifying churn factors for the telecom industry. That is where a good understanding of the particular industry and domain provide the advantage, allowing analysts to create truly appropriate clusters for the issue at hand.
Therefore, remember that just running a sophisticated model on available data without thought to the underlying issue being tackled will provide you with some “answer”, but it may not be the answer that you are looking for.