Before you actually jump on to your data to start cleaning, analyzing, visualizing, there’s something you do first.
If you don’t know what you have to analyse then what are you gonna analyse? It sounds obvious but it does happen a lot. Starting without knowing what you’re looking for. Someone says there is some data that needs to be analyzed and they send it to you or tell you where to find it, and the first thing you do is actually jumping on it, check the data and start working on. Until you actually realize you don’t really know what they want you to look at. What’s the reason they ask you to analyse this data? What are they looking for? Main ‘problem’ most of the time is that the manager or the person that asks you to analyse something, often doesn’t know exactly what they want. Because they are not the specialist, you are. And therefore they (more often than not) hope for a surprising result that just pops out of your analysis. Unfortunately that’s quite a challenge, let’s say, almost impossible to make that happen. That bit in between what you can and what someone expects you to is the gap that needs to close before you start working on the data. And make sure you are the person to close that gap, by asking questions.
And then what?
Don’t expect someone to come up with a super strong and perfectly formulated hypothesis. If they know what they are looking for, it quite often comes in the form of a general question, or wonder about disappointing results from whatever you can think of (depending of the company you’re working for). Which is, again, you are the specialist. You think in a different way than they do.
So you need to be able to transform. You need to ask questions to get as much information as possible, but at the same time you need to use your own experience to transform their information in the right strategy or approach that you will use to answer their questions in the right way.
Also important is to ask understandable questions. If you’ll start talking in all sorts of code language, they won’t understand you, might give you the wrong answer and you end up doing the wrong thing. Which is kind of a waste of time. Make sure you get a broad overview of what they want, what they expect of you and make sure to communicate properly.
It might sound all very straight forward but don’t forget that after a while, every thing you know about data analysis becomes common knowledge for you. For an employer, a manager or anyone involved in your data, it simply is not. They don’t have that knowledge, because again, they’re not the specialist, YOU ARE. So make sure you ask easy, understandable questions without too much technical language in it. Unless you’re talking to a head of data science or something, it might be a whole different story. The only thing I’m doing is making you aware of it, because in data analysis not only analytical skills are important, also communication skills are very important part of the process.