This post has been reposted from the CompassRed blog

Background & Key Principles

Recently, I had the opportunity to talk with Masters students at Temple’s Fox School of Business about successful interviewing. Many of these students are in the Business Analytics MS program, curious about the interviewing process and what we look for in a candidate at our data analytics & data strategy company. At CompassRed, we’re passionate about hiring data team members who demonstrate creative problem-solving and self-sufficiency.

While the interview is a place to talk live with the candidate and align on fit, sharing project links and code beforehand is a great way to demonstrate creative problem-solving. University coursework and MOOCs (massive open online courses) are great for learning skill sets and following instructor guidelines, but there are many instances where we see submitted projects that look almost identical from candidate to candidate. To stand out we recommend sharing projects that demonstrate these three key principles:

A unique problem you’re looking to solve (and passionate about).
A creative solution that addresses the problem with technology.
Shared writing and code to demonstrate learning, thought process, and decision-making.

This can take many forms and doesn’t have to be a monumental effort! It could be a small visualization of your blog analytics completed in an afternoon or a predictive modeling side project you’ve been building for months learning new skills. It’s all about demonstrating creativity and passion for solving cool problems.

A Personal Story

Let’s go back to March of 2019 (a simpler time?). A young data analyst working in eCommerce named Ben Kates was looking for new opportunities. When I saw that a LinkedIn connection had liked CompassRed Analyst Lead Pat Strickler’s post searching for a Data Analyst I dug into the company and thought the opportunity sounded great. We had an intro call, talked about required skillsets, and set up an in-person interview with the team.

At the time my skills in R were mainly applied to Google Analytics data ETL, automated email reporting, web scraping, and some visualization. Since I knew the CompassRed team worked in R (and shared great blog posts on their R work), I thought I could use the time I had before the interview to learn a new skill: text mining! After taking a look at the CompassRed website, I thought it’d be interesting to mine it for the most used words and phrases.

I got to coding after work hours, spending most of my time setting up a script to comb through the CompassRed site using rvest and transforming the data into a clean format using dplyr. Side note: always plan for challenges here - of course, there were older blog posts that had a differently formatted URL! After getting my hands on the downloaded data, I started diving into the tidytext R package from Julia Silge and David Robinson and getting familiar with the common functions; tokenization, removing stop words, etc. The code, which was shared, was commented for analysis steps and decisions I had to make.

At the end of the scraping process, I had a tidy dataset to work with! My dashboard tool of choice was Google Data Studio for easy UX and its nice grid system. I also knew the team used it for client deliverables. I wanted to represent a main takeaway in the header making an overall recommendation (always have a TL;DR summary!).

Feedback and Closing Thoughts

I sent the visualization over a few days before my interview and it was positively received! Not only did it make for a great discussion item during the interview but I got to learn a new skill that I knew would apply to the job. It also prepared me for the kind of self-starting work that has been required of me for the last year and a half. While writing this post I asked my colleagues what was going through their heads at the time:

"I was immediately impressed with what you sent me. It definitely made you stand out among the candidates we were considering” -Pat Strickler

“This looks like a cry for help, we should rescue him from his current job.” -Ryan Harrington

I hope this gives you some insight into how to approach a project you’re sharing with a potential employer. Mine happened to be focused on the employer itself but think about anything you’re passionate about! Are you interested in climate change and want to analyze climate impact? Are you a music fan who wants to play with your Spotify history data? Here are some thought starters!

Unique Problem You're Passionate About

Is there something you do outside of work or school that brings you joy? Is there a dataset that could contain useful insight? If no dataset exists, can you create it with web scraping?

💡

Example: As a huge fan of (insert musical artist here), I want to see how my listening has increased over the years and see which month I listened to them the most!

Creative Solution

Is there a technology you’d like to learn more about? Already familiar with a technology and want an excuse to bring it to the next level? Is there something seemingly “difficult” that you’d like to learn but don’t want to try in professional projects?

💡

Example: Learn how to use the Spotify API, download listening history, and create a notebook conducting exploratory data analysis (EDA).

Shared Writing & Code

What did you end up researching the most? What Google/Stack Overflow rabbit hole did you go down to solve a problem? What surprised you about the project? What further research would you like to do?

💡

Example: Integrate a blog-post style narrative into the EDA notebook with links to resources you learned from, challenges overcome, and things you’d do differently next time.

Because everything we do is recorded there are unlimited possibilities for projects. When you’re done, personal websites/blogs are a great medium to share, giving an employer insight into how you think, work, and approach problems.

✍

Note: if working on a project like this outside of work hours doesn’t suit your schedule, use this framework when talking about a professional project in the interview!

The Importance of Portfolio Projects in Data Analytics