Be a Top Notch PM Using Data Science by Farfetch Product Leader, Sofia Pitta

Almost 90% of executives agree that Al represents an opportunity, but a mere 18% have tried to use the technology to general revenue. 

1. Why should you care about Data Science as a PM. 

Let's first understand what is a Data Science problem. 

 This a deterministic problem: for this pair of jeans, we will always get the same recommended box (it behaves in a predictable manner). 
Some problems (many!) are solved by Data Science. Some are not. 

2. When should you use Data Science. 
Not every problem related to data is a Data Science problem. A good thumb rule is "If a problem can be solved in Excel, you dont need a Data Scientist to handle it." 
Data Science/Al products are automated systems that learn complex patterns from historical data, and based on those patterns make predictions on unseen data, to make or recommend business decisions (customer-facing/technical). 

- Your problem is complex. 

-There are patterns to learn. 

-There's historical data to learn from/data is available. 

-Unseen data follows the patterns of the training data. 

-It's at scale. 

-Changes across time. 

As a PM, you should understand if it is worth investing in DS (i.e. if the cost-benefit equation makes sence). 

3. How can a PM contribute to a Data Science team/product? 

-Clearly define the problem & the why.  

-Measure the value added to the business/customer. 

-Help your team finding high-impact problems. (Where can you automate time-consuming manual processes? Where is the friction in your product? What are other companies doing out there?) 

-Run a feasibility check. 

-Start small (before knowing if you should go big). 
-Identify risks upfront. 
-Make DS products impactful. 

-Be the advocate & connect the dots. 

Recommended articles: 

-"What you need to know about product management for AL" 

(Peter Skomoroch and Mike Loukides, 2020) 

-"Everything We Wish We'd Known About Building Data Products" 

(Dj Patil and Ruslan Belkin) 

-"Spotify's Discover Weekly: How machine learning finds your new music" 

(Sophia Ciocca, 2017) 

-"Using Machine Learning to Predict Value of Homes On Airbnb" 

(Robert Chang, Airbnb Engineering & Data Science, 2017). 

Sofia Pitta, Farfetch Product Leader