The responsibilities of a data analyst vary drastically from business to business. Unlike careers in medicine or accounting, there are no state or national licenses that a data analyst can obtain to prove their qualifications.
It can be really difficult to get a solid understanding of what it is that a data analyst does day to day. We’ll help you understand the kinds of questions a data analyst answers, the tools they use, and show you how you can break into an exciting career as a data analyst.
The demand for data analysts is growing rapidly as companies look to leverage their vast data sets to gain a competitive advantage. There is no better time to learn more about data analytics.
At their core, data analysts turn messy data into actionable insights for a business or organization. It is up to the data analyst to take raw data and turn it into a story that can be told to business leaders.
While the tasks of a data analyst change company to company, there are responsibilities that are common to most data analysts. If you’re interested in high level details of typical data analyst projects, you can browse resume examples of data professionals.
Retrieving, cleaning, and preparing data is how a typical data analyst spends ~80% of their day. Companies collect all sorts of data about their marketing, their customers, their product, and everything in between. This data is typically stored in a database. Think of a database like a really big spreadsheet with millions and millions of rows.
To access data from a database, you need to use a programming language called SQL. When you go fishing in the ocean, you need a boat to access the fish. Think of SQL like the boat you need to access data from a database.
Even small companies have databases that are millions of rows. It is the job of the data analyst to figure out exactly what data they need to answer the question they are working on. Then they need to write the SQL code to get that data and transform that data into a structure that can answer their question.
Reporting is one of the primary, and most important, functions of a data analyst. Data analysts work very closely with other departments like product and marketing to understand their reporting needs.
To build a report to shed light on whether a new marketing initiative is working, for example, a data analyst needs to know what metrics indicate whether the initiative is “working”, how to calculate those metrics, where the data is stored to do those calculations, and finally how to display the data so it’s useful and actionable for the marketing team.
After a report is created in a business intelligence tool like Tableau, the data analyst is charged with maintaining and updating the report as needed. One thing is certain for data analysts: reports break due to changes in the underlying data used to generate the report. You’ll have to diagnose and fix those issues to ensure the report is displaying accurate results.
Much like reporting, answering high level business questions with data requires collaboration with other business units. Data analysts never work in isolation. They have to deeply understand the business problems they are trying to solve and that is accomplished by talking to colleagues. Data in a vacuum is useless. It is the job of the data analyst to tell a story with that data.
“When we’re hiring data analysts we try to understand if they’ll be comfortable making decisions in ambiguity. We want analysts who can take data and make actionable recommendations while also articulating the risks associated with their recommendations”, says Neal Taparia, the CEO of Soliatired.
Let’s say your company launched a new product feature a month ago. Now your boss wants to know whether that product launch has had a positive impact on the business. This is where creativity is an invaluable asset for a data analyst. You have to come up with a list of possible factors that can impact whether the launch was a success, then quantify those factors and arrive at a compelling answer.
For example, did this new product feature positively impact customer conversion rate? Did it improve customer retention? Did it have an impact on the volume of customer support requests? As a data analyst you’ll have to come up with these questions then combine the answers to these various questions to decide whether the product feature helped or hurt the business. At the end of the day, data analysts need to make concrete recommendations backed by intuition and data.
Unlike a data scientist, data analysts usually don’t build predictive models that are used real time in a product or website. For example, data analysts won’t build the functionality to power a “recommended items” on an e-commerce website while a data scientist might.
After you’ve worked with your business partners to understand their reporting needs or business questions they’d like answered, your focus as a data analyst turns to gathering the requisite data to solve the problem.
What if the data you need to answer a question is not currently available?
For example, what if you need to know how long a user was signed in to their account but as a business you’re not currently logging when someone signs out of their account? If you only know when a user signs in, but don’t know when they sign out then you won’t be able to know how long they were signed in!
Understanding and communicating gaps in a business unit’s current data collection is another important responsibility of the data analyst. Once a gap is identified the data analyst will work closely with the engineering team to implement a solution to close that gap.
To gather data, analyze that data, and present findings data analysts need a few tools in their toolbelt. Here are the top skills employers are looking for in prospective data analysts:
Like we talked about before, SQL is the programming language that is used to get data out of databases. Since nearly all companies store their data in databases, it makes sense that 90% of data analyst job openings require SQL skills!
Next, you need a tool to analyze the data once it is retrieved from the database with SQL. This is where Excel, Python, R, and SAS come in. These are all tools (Python, R, and SAS are programming languages) that can be used to analyze data. With these tools you can assess trends, perform statistical tests, and visualize results of your analyses.
Once you gather and analyze data you may want to display your results in a report that automatically updates and can be accessed by business stakeholders. To create those easily accessible reports you need to use a Business Intelligence (BI) tool like Tableau or Looker. These tools make it simple to create compelling data visualizations that automatically update over time.
Now that you have a good idea of the day-to-day responsibilities of a data analyst and you know the tools of the trade, let’s put it all together and analyze the steps of a sample project you might work on as a data analyst.
The marketing team for your monthly chocolate subscription company just started experimenting with paid TikTok ads and they want to understand how those ads are performing so far.
When a user sees an ad for your product, they go through a few steps before they become a paying customer. You’ll likely want to create metrics around each step of this process.
Before you can start gathering data, you need to understand what metrics you need to calculate to answer the business question at hand. Now that we know what questions we need to answer, we’ve come up with a list of metrics we want to calculate.
Now we have a concrete list of metrics we want to calculate over time to see whether this TikTok ad campaign is performing well, we need to gather the data to calculate those metrics.
You quickly realize that the data to gather these metrics comes primarily from three data sources:
Once you know where you’ll get the data from, it’s time to actually get it! We’ll export data from TikTok and Google Analytics into Excel to calculate those metrics. To get data from the database you write the necessary SQL queries then dump that data into Excel as well.
Finally you have all of the data you need to assess the performance of the TikTok ads. You calculate all of the metrics of interest and analyze the results to make a recommendation. You find that while the click through rate of the ads is low, people that do click the ads convert into paid customers at a higher rate and at a higher revenue per customer than other people who come to the website.
Although right now the return on investment (ROI) of the money spent on the ads is very low, you recommend this can be drastically improved by testing more ads to improve the click through rate. Overall, you think TikTok advertising is promising and worth continuing to test. You send your analysis and recommendation in the form of an email to the marketing manager.
Data analysis is a challenging, rewarding career that pays well and has great career prospects. To break into the field you’ll need to have a strong foundation in statistics, business, and the tools of the trade. A data analytics bootcamp is a great way to gather these requisite skills.
When you’re ready to apply to jobs, you’ll need a strong data analyst resume. Here’s a sample to get you started:
Here are the quickest ways to make your resume stand out and increase your chances of getting an interview:
Are you interested in data and how learning about it can either help you switch careers or improve in your current field? Check out our Data Analytics Bootcamp program which will help you become a true data expert.
What is the difference between a data engineer, a data scientist and a data analyst?Read more...
The best data science cheat sheetsRead more...
Learn the basics of data analytics: Intro to SQLRead more...
What is Python? Learn the top 3 best uses for Python programmingRead more...
Data science vs. data analyticsRead more...
6 Reasons To Learn About Data AnalyticsRead more...