More MBAs are interested in data science and analytics than ever before. To stay relevant, MBA programs are rapidly expanding their curriculums. However, recruiting for an analytics role with a business background can be tricky.
As a recent graduate, I had difficulty contacting hiring managers and establishing credibility. I was declined dozens of times before I found a good fit. And at many points along the journey, the best way forward was a counter-intuitive one.
Here’s how I secured a product analytics position at Google. Hopefully, my experience can shed some light on the process and help you secure a fulfilling data science position.
Develop these non-negotiable skills
There are many skills that you will need to develop to be competitive. While most will be included in your regular scheduled university programming, some remain glaringly absent. Don’t let these catch you blindsided.*
A data scientist that doesn’t know SQL is like a chef that doesn’t know how to use a knife. To slice and dice data, you’re going to need SQL. Moreover, it’s liberating to not have to rely on someone else to hand you a CSV. Expect to be tested on this.
Sadly, SQL training remains one of those critical data skills that are not yet de rigueur at MBA programs. It can seem intimidating at first, but it needn’t be — the syntax is intuitive and you can usually get by with just a few basic functions. For a gentle primer, I used Mode’s Introduction to SQL series.
If you want to have a concrete impact on business decisions at your company, you’re going to have to be persuasive. Slides overwhelmed with charts and covered in uninformative metrics are common at most companies; it’s a quick way to lose your reader and demonstrate your inexperience — you don’t want that.
I found that an easy way to distinguish yourself from the average data scientist is to master the art of storytelling. While many data scientists are tempted to tout their command of new tools and techniques, an MBAs advantage is balancing rigor with consumption. Tufte’s The Visual Display Of Quantitative Information delivers a masterclass on thinking in data and helped me get started here.
Color outside the lines
Data science as a career, especially for MBAs, is young and moving quickly — this means that there are no hard and fast rules or just one right way to do things.
My favorite part of this process was the eye-opening opportunity to learn the latest techniques from practitioners and professors. I took as many data-heavy classes as I could, but I avoided sticking to the script and limiting myself to what a “track” had to offer. By pursuing PhD courses and courses in the Computer Science department I was able to gain breadth and build confidence.
I also followed my curiosity and allowed different interests to intersect. In doing so, even a theory-heavy class with a lot of reading can become a “data” class. For example, one of my favorite courses was in Public Policy, where weekly assignments were essays but the final was a heavily-weighted project on analyzing SF demographics.
Like a game of Calvinball, data-related titles are notoriously confounding and ill-defined. When my search for second and third-degree LinkedIn connections for “Sr. Data Scientist” left me hanging, I started peering out further. Other common titles I looked for included Head of Customer Insights, Director of Decision Sciences, or Business Intelligence Lead.
This is another opportunity to color outside the lines — today most roles are “data-adjacent.” A well-connected Product Manager, Program Manager or Operations lead will also be in a position to help you understand the data culture at a firm and get your foot in the door. When I was looking for internships, I didn’t stress about my title. I looked for any role where I could apply what I had learned and parlay that into a full-time role.
Expect to recruit differently
Unlike recruiting for consulting or banking where business schools are well-oiled machines, I found recruiting for analytics roles to be more challenging. After dozens of applications, I had first round interviews at twelve companies and received offers from two.
Finding your fit
When looking for companies, seek those that treat data as a first class citizen and not an afterthought. Typically these have centralized data teams, not embedded ones. They tend to have strong partners in engineering and data infrastructure. Their leadership goes beyond paying lip service and has a demonstrable history of making bold unconventional data-driven decisions.
Prepare to be patient and flexible. There is no data science recruiting cycle, most startups put up a job description as their need arises. Get your name out there by hanging around the hoop and reaching out to connections early so that they have you in mind when the time comes.
Tips for the take-home
In the otherwise gray and protracted recruiting process, I found the take-home to be a bright spot and an opportunity to shine. With a strong submission, recruiters are quick to usher you forward. An easy way to distinguish your submission is to do just a little extra — like not using default plots, including references that demonstrate enthusiasm or insider knowledge, and adding a “so-what” that takes a decisive call or recommends a future line of inquiry.
It’s worth remembering that every interviewer is secretly rooting for you to succeed so you should emphasize all the things that you were able to accomplish rather than worry about what you didn’t.
If you are coming from a background without technical skills, it is easy to feel insecure — do not let that concern overshadow all that you bring to the table. A business-savvy technical person or a technically-proficient business person is an asset to any firm.
You will not be graded by your RMSE. As an individual contributor, you will be recognized for your ability to take an ill-defined problem — structure it, devise metrics, and confidently socialize the results. And as a manager, you will be recognized for your ability to identify a problem — spearhead an initiative, secure resources, and lead others in execution.
If our goal as data scientists is to drive change, then business skills are an edge, they will serve us well wherever we land.