I think I look at this a little different than most. At the very minimum, you need mid-level Excel skills, curiosity, critical thinking and the ability to learn quickly. Excel is pretty much the most basic tool at your disposal as a data analyst, so regardless of whether other languages like Python could be better suited to handle larger data sets, knowing Excel still remains a core skill that cannot be ignored. Excel is vastly used for quick analytic tasks and lighter databases, and being able to use a quick lookup function here and there is always a must.
From there, you want to know how databases work and how to code SQL (Structured Query Language) so you can dig deeper into the data sets that you have. SQL is an industry standard that is required whenever datasets are too big for Excel to handle, and knowing how to handle it is an absolute necessity if you plan to work with Big Data at all. Almost every organization needs someone who knows SQL—whether to manage and store data, connect multiple databases or build or change database structures altogether.
It is also important to have the ability to visualize analytics results into an easy-to-read format. Being able to tell a compelling story driven by data is critical to grab the attention of everyone who’s going to actually put those data-driven insights into practice.
Graphs and charts must present all findings in a clear way, but they must be concise as well — and that’s something data analysts often tend to forget. My rule of thumb is that it should take no longer than three minutes to look at a visual and understand what it’s trying to say.
What will make you stick out, though, is if you also have some knowledge on the following:
- Data cleansing
- ETL (extract transform & load) processes
- ELT (extract, load and transform) processes
- How to code Python
- Ability to use APIs to extract data
- Ability to use a statistical programming language called R.
While data analysis and exploration might look like the most interesting and enticing part of this work, nearly 80% of it is made of rather boring data cleaning and data aggregating operations. However, these steps are necessary, so having a good degree of patience and the critical thinking required to streamline this stage and make it more efficient are extremely valuable skills.