Antiquated Processes on a Big Scale

Continuing to Use Data to Drive Decisions

By: Julia Taitz (Wharton ‘19)

Data-driven decision making came along way before computers, the Internet, and social media. When one assembly line was faster than the other, produce was delivered on certain days, and if individuals needed the best directions on a map, they looked to population samples and the information in front of them. If we have been using data to make educated guesses all this time, then why are people seemingly so uneasy about analytics?

For the last decade, people have been using the Internet without any regard of what they were sharing. To most people, cookies were still only delicious snacks. When explaining to a friend that Snapchat pictures most likely are stored in a database after they “disappear,” she seemed completely bewildered. But there is not anything novel about the idea of tracking behavior to make observations. The only new aspect is the ability to do so on a mass scale resulting in great variety and velocity of information. Instead of coming up with a hypothesis and looking for data to answer a question, individuals now combine abundant data with exploration to determine actionable insights.

(Credit: Kartik Hosanagar)

The field of data science has emerged in the face of needs to better understand online behavior. For those of you unfamiliar, data science sits in the middle of a spectrum between software engineering and product management. The technical side comes from needing programming languages like SQL, Python, and R to aggregate large amounts of data. The logic behind most querying languages is fairly similar to that of Excel besides that data languages can handle trillions of pieces of information. For the remaining fifty percent of a data scientist’s day, he or she must determine what the key takeaways are and act on the data. While any data scientist can spit out a number on growth or retention, the real work comes from using analytics to drive product changes and communicating those decisions to non-technical stakeholders.

The rise in data science has spurred innovation across the board. On one hand, companies like Google’s BigQuery, Hadoop, and Hive serve to make processing and visualization easier. Other startups rely on analytics at the core of their products. If you were asked to think about a company like Spotify, you might give nods to Discovery Weekly, Release Radar, and Daily Mixes. What you might have overlooked though is that there is fundamentality no software changes across these core features. Instead, the company chooses to leverage data to provide new offerings with each update.

Why This is Important for Student Entrepreneurs

Whether people are ready or not, a wave of data-driven companies are here, and more are coming. At the WeissFund, we have invested in companies like Twine and BarnOwl that look at internal workforce analytics and ranching productivity respectively. Student entrepreneurs are leveraging data to create efficient and world class solutions to everyday problems.

But still, not enough engineers are learning data science. While some colleges like Penn now have minors in this field, most schools across the country do not provide access to this education. For most students, entry into data science feels less comfortable than software engineering when classes are spent learning about data structures and algorithms. In order to more individuals the correct skills, engineering schools need to integrate data science into their curriculae. But do not worry, this shouldn’t be a hard sell because there has got to be the data out there on why this is a good decision.

Julia Taitz is a senior studying Entrepreneurship and Computer Science. She has previously worked at Facebook and Spotify as a data scientist and enjoys skiing, traveling, and baking. Feel free to reach out to her at

WeissFund is dedicated to funding, promoting, cultivating, and supporting student entrepreneurship in the UPenn community. Working on a startup? Interested in partnering? Want to get involved? Drop us a line at