How Do I Become A Data Analyst?

Whether you’re a graduate looking for your first role, or someone with more experience looking to get your foot in the door of this constantly developing role, we’ve got all the details you need to know about Data Analysts. 

‍

What Does a Data Analyst actually do?

A lot of articles out there talk about the role of data analysts and all they really say is ‘a data analyst analyses data.’ Which isn’t super helpful. To get more specific data analysts use data to optimise all kinds of things in all kinds of ways, from marketing strategies to recruitment. Two analysts at the same company could have wildly different roles.

But regardless of what they’re doing, most follow a similar pattern of work.

Find a Problem

• This could be something simple like ‘find out how many accounts were deactivated in the last month’ or something more complex like ‘come up with reasons why this many accounts were deactivated in the last month’.
• Either way, every task for a data analyst starts with the definition of a problem that needs a solution.

Explore the problem 

• With the problem in hand, you’ll need to query the data and gather information. 
• This usually involves SQL or Python, depending on the format of the data.

Collect insights 

• Then you need to take this data and put it all together.
• This will often involve a lot of tidying up and organising data, and can be pretty time consuming. A lot of your time will be spent on this section, it can get pretty tedious.
• Did you know that a lot of Data Analysts are looking into using AI and machine learning to refine this process? More on that later...

Visualise insights 

• This can be as simple as creating a graph with Excel's built-in tools, or it could require a dashboard such as Tableau or FineReport.
• Presenting data in a good format for the context is really important because...

Communicate findings

• Once you’ve used the data to reach a conclusion, you’ll need to be able to present your findings to the rest of the team. 
• Without this, the team won’t be able to do much with what you’ve uncovered, so get polishing those presentation skills!

‍

How do you Become A Data Analyst?

Whatever your background, there are ways to become a Data Analyst using your current skill set- you just need to be open to developing other skills. You need to know what’s needed, which you have, and which you need to work on. 
‍

We’ve put together a simple list of important data analyst skills to act as a guideline- it’s not extensive and we’ve simplified some areas, but this should be a good guideline to get you started.

General Skills

• Programming- absolutely a key skill, You don’t need  a super advanced understanding, but the more you know, the more of a competitive edge you have.
• Statistics- this is the other key skill. Without statistics, you can’t interpret your data!
• Math- more generally than statistics,  a strong understanding of mathematical principles will help you feel more comfortable interpreting data.
• Machine Learning- actually building new machine learning tools is a highly advanced skill more appropriate for data scientists, but learning how to use the tools others have made will help make your process way more efficient. 

Data Wrangling, Visualisation and Intuition- OK, there’s a lot here, so we’re going to break these down even further

• Data Wrangling is translating raw data into useful datasets. Clean, organise, and arrange data. In other words...wrangle it!
• Data Visualisation is illustrating those freshly wrangled datasets so they can be understood by people who are not data analysts (or even especially tech literate). 
• Data Intuition- is broader. As you develop, you’ll learn to intuit what is important and what isn’t- you could spend your whole life on one dataset otherwise!

Phew! There’s a reason that data analysis is lucrative. It’s a big payoff for a lot of time investment. But don’t be intimidated- remember that each skill builds on the other. As you learn one, the others become simpler.

Now, what about if you already come from a background that suits data analysis?

If you’re a programmer…

• Focus on learning about statistics and probability, and study linear algebra. A good understanding of statistics will help you understand and interpret the data that your programming skills will flag up.
• Learn to translate real-world problems into algebraic expressions and equations.

If you’re a mathematician…

• Basic programming principles (such as variables, control loops and functions)
• Debugging- programs almost never work the first time you run them! Learn how to fix that.
• Object-Orientated Programming (also known as OOP)
• Advanced Concepts (such as data structures, algorithms and software design patterns)- these will allow you to optimise your programs for efficiency

We’ve collated a couple of e-learning tools so that you can get started.

• If you’re a total beginner at Python, try this course from skillshare, which is made with the assumption that the audience knows nothing

• If you already know some Python and want to specialise into using it as a data analyst, try this 

• Learn basic statistics here

• For courses on statistics themed more around data analysis, look here

• Learn algebra at various levels here

• Calculus is also important to brush up on, try that here

‍

don’t be intimidated- remember that each skill builds on the other..

‍

What are some key Data Analyst Tools? 

Different companies will have different requirements for their data analysts, as well as different in-house tools and languages they like to use. We’ve found some of the more common ones that you should know as standard.

SQL

• To keep it simple, SQL is a programming language used to get data from a database.
• If you need to brush up on your skills, Codecademy has a tutorial covering the basics here

Excel

• Simple but deceptively powerful, most companies use either Excel or Google Sheets (which is functionally similar but makes it easier to work together) for their spreadsheet needs.
• You should have a decent understanding of most of the basic Excel functions, as well as how to use pivot tables and create graphs. If it’s been a while since you last used it, take some time to practice for a refresher.

Tableau

• Tableau is a specific data visualisation platform that is widely used by many companies.
• It’s great for displaying real-time metrics and can really support presentation of data
• Other companies may use different platforms (such as PowerBI or FineReport), so you may need to get trained up on those, but the principles generally remain the same.

Python

• Python is a super versatile programming language that’s pretty simple to pick up and learn.
• It’s one of the most popular languages for data analysts, so you’ll need to learn it to keep on the curve
• Some data analysts prefer to use R, so check and see if that suits your needs better.

AI & Machine Learning

• This is an area that is constantly developing, so brush up on your fundamentals to stay current, especially if you want to excel. 
• Machine learning is being used to help organise data, which cuts out a huge chunk of busywork that data analysts have to go through (Remember all that data wrangling from earlier?)
• Especially look into the cutting edge research about probability algorithms and mathematical optimisation. 

The Future Of Data Analysis?

Data Analysis is a constantly evolving field. What developments are shaping up to be the biggest deal going forward?

• Data analysts spend most of their time cleaning up data rather than collecting or modelling it. Machine learning tools will become even more important as a result. 

• Automation! Data Analysts were scarce a few years ago, so many tools have been developed to pick up the slack and decrease time spent on the cleaning work. Cutting down on ‘Data Janitorial Work’ in order to focus more on the actual analysis should make the role more attractive- will we see more people switching to Data Analysis roles? Or will the duties of a data analyst be blended into other roles?

• Artificial Intelligence isn’t a perfect solution. Can we always trust the results that an AI gives us, and are their conclusions always reasonable? It’s a known factor that AI can be as prone to bias and mistakes as any human- they just make the decisions faster. How will the industry adapt to this?

• Companies have spent a lot of time adapting to GDPR regulations. The full implications of how anonymised data can be used effectively- and whether data can ever be truly anonymous- have yet to be uncovered. It’s definitely something to watch!

Where do you think the market will go? 

‍