How do you usually interview a data analyst candidate? In this article, I'll share some of the guidelines and areas you can focus on when interviewing a data analyst candidate. These serve as pointers to aid with your interview.
What does a data analyst usually do?
Before we go into how to interview a data analyst, first we need to define the job scope of the data analyst.
Generally speaking, depending on the actual role, the data analyst will do a few things:
Extract data and build recurring reports for different business users.
Perform exploratory analysis: extract and look at the data, organize them, and try to draw meaningful, actionable insights from these vast amount of data.
If the company doesn't have a data engineer, the data analyst sometimes will have to act as a data engineer, writing custom ETL scripts to extract and move data into data warehouse.
Important Skills of A Data Analyst
Aside from typical softskills, here are some key important skills that a data analyst should possess:
1. Business Sense
People think data analyst is a technical role, but this is not completely true. A good data analyst should understand how the business operates.
He/she needs to know what's important for the business, which metrics makes sense.
2. Data Sense
This is like business sense, but more specific. A good data analyst understands how data comes into the picture. They are not fully reliant on the data, but only use data as a quantifiable measure to support their decision makings.
A person with good data sense is someone that can tell you if something feels right, or wrong when looking at a chart/table/numbers. They can tell you that they think this conversion rate is too low, or that bookings number doesn’t make sense given what they understand about the business.
3. SQL Skills
Every data analyst must know SQL.
As data and data-related business requirements get more complex, people tend to move to using more technical tools to do BI, instead of relying on drag-and-drop interfaces (only for simpler reports). And in the world of data analytics, SQL is the data language. It has been, and will be for a long time.
So even if your candidates said they’re experienced with some well-known BI tools, make sure they’re well-versed with SQL as well. This is especially important for Data Analyst role that involves with Product, or complex online business (like ecommerce, marketplace).
Data Analyst Interview Questions
Below are some questions to ask a data analyst to test them on different skills as above. Note: feel free to suggest more in the comments and I hope to update this list with more questions.
1. What do you understand about our business, what do you think would be important in our business, and how would you measure the business’s performance?
This is more of a conversation than a single question. Pay attention to the questions they ask you, and which direction they’re venturing towards.
A good data analyst has the ability to ask great questions.
2. Please share some past data analysis work and analysis/insights that you’ve come up with.
This is a common question that’s asked. The important thing is to pay attention to their communication skills, how they communicate the insights/analysis:
- Do they have strong clarity in the communications?
- Is there anything actionable coming out of the analysis? If so what was the results?
- Ask more details about their analysis, debate with them, see how thoughtful they are in their analysis.
3. Problem Solving: An experiment measurement question
One of the jobs of the data analyst is to assist in measuring effectiveness of a particular feature or campaign(A/B Testing, marketing promotion) and quantify the results.
Take a hypothetical feature of your product, or other common products, and ask them how would they define what makes it a successful feature? Some examples would be:
- If you run an ecommerce site, how to measure the effectiveness of the Search feature of an ecommerce website
- If you run Facebook, what would you choose to be the metrics for Newsfeed feature?
Tie them with the events tracking question below.
4. Data Design: Events Tracking and Funnel Tracking
If you’re an internet business, there is high chance that you have a sales/lead funnel and some events tracking in place. Ask questions around your funnel, and your events tracking.
A good data analyst should be able to understand these 2 concepts well, and they should also understand how the tracking is done even at the engineering level.
5. How would they design a sales funnel, how would you implement some tracking to measure conversions by different source?
Take some common lead funnel, and ask them what would be low/medium/high conversion rates at each step (free → paid, traffic → trial). This tests their numbers sense.
6. Ask knowledge about database, data warehouse and data pipeline
This might be a little bit of a stretch, but ask them on their knowledge of database and data engineering. A good, experienced data analyst will have worked with some database or data warehouse platform before, and I would expect them to at least have some understanding about that.
7. Give them real dataset and work with them like a hired data analyst
Depending on your company policy and the interview’s progress, you might want to consider doing this or not. But this is a pretty powerful exercise that can tell a lot about the candidate’s ability to deliver.
This is like an onsite assignment, where they are given a computer with access to your data warehouse. Treat them like a hired data analyst, work with them on some key areas that you want some analysis on, and see what they can come up with within 1-2 hours.
Some suggestions would be:
- If you’re in ecommerce, ask them to look into user retention and purchasing behaviour.
- If you’re in travel, ask them to look into promo code behaviour
Essentially, the idea is to treat them like your team’s actual data analyst, and spend 1-2 hours working with them to see if they’re up to your expectation. Again, pay attention to their line of thinking, and the questions they ask.
p/s: Do discount for the complexity of the business, since they might well be spending all the time trying to understand your data, instead of asking meaningful questions and writing queries to extract insights.
A Good Data Analyst Is Hard To Find
I think a good data analyst is even harder to find than a good software engineer, because a good data analyst has both technical skills and business sense, which is usually hard to come by.
It’s rare to find someone that has a combination of:
- Business sense (understand how business operates).
- Data sense (understand how data can come into the picture)
- Technical skills (SQL, Python, etc, have the necessary skills to get the data needed)
So if you find one that fits these 3 skills, don’t pass on ;)
p/s: how do you usually interview data analysts? Share your thoughts them in the comments!
Subscribe to Holistics Blog - The Data Analyst Guidebook
Get the latest posts delivered right to your inbox
From SQL Queries To Beautiful Charts
Connect to your database and build beautiful charts with Holistics BILearn More
"Holistics is the solution to the increasingly many and complex data requests from the operational teams"
Tang Yee Jie
Senior Data Analyst, Grab