You're a PM at LinkedIn. Your data science team reports that 64% of outbound messages receive no response. What do you do?

My first step would be to deeply understand the problem by asking clarifying questions. I'd want to define "outbound message" precisely in the LinkedIn context and understand the trend: Is this a recent, sudden drop or a gradual decline over time?

Next, I'd investigate external factors. Have there been any significant shifts in the competitive landscape, new industry trends, or major events that could impact user engagement?

Then, I'd pivot to internal factors. This involves slicing the data to identify patterns. I'd analyze response rates across different user segments (e.g., by industry, seniority, connection level), message types (e.g., connection requests, InMail, direct messages), and message content characteristics. Understanding which messages are failing to get responses is crucial for developing targeted solutions.

Based on these insights, I would formulate hypotheses and propose experiments. For example, if messages to a specific user segment are underperforming, we might test different messaging templates, personalization strategies, or even explore changes to the user interface for sending messages. The goal is to identify actionable levers to improve the response rate.