Cool Job: I Analyze Pinterest Data Using SQL & Python

Cool Job: I Analyze Pinterest Data Using SQL & Python

5 minutes

Say you’re looking on Pinterest for a mac and cheese recipe to try: There are so many delicious-looking options, so you start pinning the gooiest, cheesiest photos that you think you’ll come back to later. As an Ads Measurement Lead at Pinterest, it’s Scott Menke’s job to use data analysis to take a deep dive into why you chose the mac and cheese pins that you did.

As Scott puts it: “I do research to understand consumer behavior on Pinterest,” he says. Scott uses SQL and Jupyter notebooks to analyze Pinterest data, then puts together research reports to present to advertisers so that they can better understand how to reach customers. For example, a food and beverage company that sells mac and cheese products might want to see how Pinterest audiences are interacting with mac and cheese content so that they can tailor advertisements accordingly.

Scott actually got his start working in the insurance industry as an actuary, which is someone who uses math and statistics to analyze the financial costs of risk and uncertainty. While he used some SQL as an actuary, he’s a self-taught programmer with a knack for puzzles, trivia, and problem-solving. (Fun fact: Scott has been a contestant on Jeopardy! and the game show The $100,000 Pyramid.) Here’s how Scott went from actuarial science to data science, what it’s like to work at Pinterest, and his advice for people with similar aspirations.

Learn something new for free

What got me interested in the job

“I took some beginner computer science classes in college, but I was a terrible student. My first real job out of school was working as a financial analyst at a casino in Las Vegas, and then I was an actuary — that’s what my training is in. Even as an actuary, I was learning how to code a little bit.

I really loved learning, so I got really into these little online coding challenges, like: Print every multiple of 3 but not every multiple of 7. That’s something that really suited me, and I still think it’s fun to figure out how the logic works. I love puzzles and games and I’m a big math guy.

I always felt like I probably wasn’t good enough to work in a ‘real’ tech industry job because I was a hobbyist coder and I just didn’t have the chops. Now that I’m here, I’m like, Okay, I am capable of doing what they’re doing. It’s just that getting a job is hard.”

How I got in the door

“My wife does the same thing as me at another company. I knew just enough of the fundamentals of advertising and the tech industry, just by being married to her. One of her former managers went to Pinterest and needed more of a data-type person.

I knew how to code pretty welI; I knew SQL and I used that a bit in my actuarial job. Like, instead of building an audience of mac and cheese lovers, as an actuary you’re working on things like, which of these policyholders paid $10,000 in policy revenue for some time? The combination of that was enough to get the job at Pinterest.”

What I actually do every day

“I use SQL and Python every day. To use the mac and cheese example, I need to find those pins on Pinterest that are tagged as related to the food and drink coterie, and specifically, cheese and mac and cheese. Then I can start uncovering all kinds of interesting things about these pins. Is there a seasonality component to when these pins are most popular? Maybe search and pin activity ramps up leading into summertime BBQs. What about time of day: Are people more interested in looking for a last-minute recipe right before dinnertime, or is it consistent throughout the day? Which keywords are used most often when searching for mac and cheese pins? Are Pinners primarily looking for gooey, decadent options, or easy and budget-conscious versions of their favorite side dish? The crazy thing too is these insights change over time — trends start to emerge, and if you’re not paying attention, keto or vegan mac and cheese suddenly become very popular. Our annual report Pinterest Predicts has some great examples of the emerging trends we have been able to see.

Studying one topic can lead you to another topic entirely, too. If we look at related searches and pins, you can see some other topics you might not expect. For example, you might see a lot of sports content being pinned by the mac and cheese audience. They may over-index on content relating to Kentucky Derby parties, or the Dallas Cowboys. It could be that folks are trying to whip something up to watch a sporting event with friends. Zooming out, maybe we can infer that those interested in mac and cheese are also often interested in the holidays, or hosting during special occasions. There is an unlimited amount of fascinating trends that can be understood by diving into the data and using SQL and Python to unleash the hidden insights.”

Here’s what you need to get started

Scott’s advice for anyone who aspires to work in tech is to keep learning and be like a sieve for knowledge. “Continue learning in any way, shape, or form that you can, no matter how insignificant that may seem — because skills accumulate in a way that’s impossible to predict or anticipate,” he says. “Just be open to learning a little bit every day in different fields, and you will eventually find success.”

As a career-switcher, going from the staid actuarial world to a fast-paced tech organization was a transition, Scott says. In tech, you have to wear lots of hats and be adaptable to change: “The analogy I give people at Pinterest is that it’s like a pirate ship, in that no job is really well-defined,” he says. “You’re part of the ship, but what do you do? I kind of swab the decks, look at the crow’s nest, scan the seas to make sure everything is okay.”

Having strong data skills can help you add value to any type of organization. Be sure to check out the Codecademy course catalog to find data science courses in SQL, Python, and more. A great place to start is with the free beginner-friendly course Learn SQL, or the skill path Data Science Foundations. If you’re looking for the fastest (and lowest-code) way to work in data, consider taking the new Codecademy career path Business Intelligence Data Analyst. You’ll learn how to use Python, SQL, Jupyter Notebooks, and other business intelligence tools to build dashboards and reports.

Related courses

3 courses

Related articles

7 articles