What it takes to be a Data Analyst
Updated: Oct 28, 2020
All my life as a professional has been around data.
It started by consuming it and make usage of it in web applications and continued with designing it.
From SQL scripts to ORMs (Object-Relational Mapping) frameworks, I managed data with the mentality of using it in Object-Oriented Programming, and sometimes that impacted the quality of the data model (a topic for another article I'm afraid).
The need of scaling web applications forced me to develop a broader perspective on designing the data model and the database solution chosen. Needing to take into account scalability, replicability, transaction load, changed the way how I perceive data.
Then Revolut happened. Even though I started as a Community Manager, Revolut is a data-driven company, so from day one, I was able to query it. They were basic select and joins to find active users, number of transactions, simple things.
The first part is not as random as it might appear, but it is instead me trying to paint a picture of my relationship with data, yet it does not contain the recent present.
Now I'm in the position where I'm responsible for making data make sense and as simple as it sounds it is quite an interesting ride to be on.
Now, with that in mind, what it takes to be a Data Analyst.
Frankly, I'm still figuring out some bits. Since it's such a broad subject, data, I'll stick to some basic principles I have recurrently noticed appearing when manipulating and visualizing it.
Structured Query Language
Also known as SQL, is the strongest tool a Data Analyst has, so mastering it is a true must.
It might appear as a "boring" language, yet it is extremely powerful.
So, it's simple. You gotta know SQL.
Structured and Logical thinking
It is no coincidence that these two are hand in hand. Separately they are vital characteristics needed for all roles that involve developing any kind of tech-oriented result.
However, it is crucial to have them both when working with big volumes of data.
Without structure in the way you process or create it, you will most likely neglect small but vital details which are "playing hide and seek", blending in the big volumes.
Without logical thinking, the output of your work will mislead, rather than providing support in taking decisions.
Data Analysts' main job is to show reality using the data; that is why these two characteristics are core ones, necessary to form it reliably.
Communication and Visualisation
It's not about effective communication but rather how to relevantly deliver the information from a technical format to a non-technical one.
Data tends to be scattered in multiple tables, with some column names, with some numbers or pieces of text so it's kind of hard to understand it.
A Data Analyst's job is to make be easily interpreted by taking into consideration two dimensions: How the data is visualized and how easy is for a non-technical person is to understand it.
Lines, bars, bars and lines, scatter, pie and on and on and on. There are a lot of ways to see how data looks like and sometimes conclusions may differ based on the chosen one and sometimes showing misleading directions.
The chosen visualization needs to be clear, easily readable and to be as self-explanatory as possible. The best way is to think that the person who reads your work has nothing to do with the topic and yet the information is easily understood.
SQL + Way of Thinking + Good delivery = Data Analyst
This is the base layer of the whole recipe of being a Data Analyst. On top of it may come a second layer, seasoning and some sauce to complete the whole course (data and food in the same post, go figure).