I was thrilled to be given the chance to work as a Product Manager at Yelp in San Francisco on my fourth work term. It was an intense term that had me doing plenty of both technical and non-technical work, and taking on a ton of responsibility. This experience has not only made me a better product manager, but has also improved my abilities as a software developer by giving me a better picture of all the work and considerations that go on beyond the code.
Within a few weeks of starting, I was the PM in charge of the Search Data team, an amazing team of engineers dedicated to improving Yelp’s data through machine learning. I was responsible for planning out the team’s projects, from the high level of quarterly goals down to the low level of individual tickets. I routinely participated in the team’s work, helping improve the team’s machine learning models with additional signals. I coordinated my team’s work with the work of other teams spanning the United States, Germany, and India. Ultimately, we shipped multiple products, including improved detection of chains of businesses, automated deduplication of businesses, and automated categorization of businesses.
I strove to make every decision I made over this term data-driven. I used SQL and Python (Pandas and matplotlib) to extract insights from Yelp’s massive database of business, searches, and user generated content. When quantitative data was not available in the database, I drove the effort to reach out to Yelp users with surveys to figure out what they were looking for from our service and how we could better help them find it. My survey was distributed via Yelp’s ad network and received over 8 million impressions. The data gathered was immediately put to use in redesigns of Yelp’s business photos page.