David Lieb on Understanding User Behavior with Dot Plots

In the quest to build products that truly resonate with users, understanding how people interact with them is paramount. Aggregated metrics like daily and monthly active users offer a high-level view, but they often fail to capture the nuances of individual user journeys. David Lieb, General Partner at Y Combinator, emphasizes the critical importance of looking beyond these broad strokes to truly understand how customers use a product. In a recent "Startup School" video, Lieb introduces a powerful yet straightforward tool for gaining this granular insight: the dot plot.

The Pitfall of Aggregate Data

Lieb highlights a common pitfall for founders: relying too heavily on aggregate data. While metrics like DAU (Daily Active Users) and MAU (Monthly Active Users) are useful for tracking overall growth, they can mask critical details about user behavior. For instance, a product might show high overall engagement, but this can be misleading if only a small segment of users is driving that engagement, while others are struggling or not finding value.

The full discussion can be found on YC's YouTube channel.

"What you don't know," Lieb explains, "is how they are interacting with your product, what features they are using, what the pacing of their usage is." This deeper understanding is crucial for iterating on the product, identifying areas of friction, and ultimately ensuring that users are deriving real value.

Introducing the Dot Plot

To address this gap, Lieb advocates for the use of dot plots. This visualization method involves creating a two-dimensional grid where each row represents an individual user and each column represents a specific time period, such as a day or even a sub-day interval. A dot is placed in a cell to signify that a particular user performed a specific, valuable action during that time period.

Lieb illustrates this by drawing a sample dot plot. He labels rows with user names (e.g., "Dave," "User 2") and columns with days of the week (Mon, Tue, Wed, etc.). He then populates the grid with dots to represent specific user actions, such as listening to a song on a music app or searching within a product. This visual representation immediately allows for a clear overview of individual user engagement patterns.

Uncovering Hidden Patterns

The power of the dot plot lies in its ability to reveal patterns that are invisible in aggregated data. For example, Lieb points out that by looking at a dot plot, one can quickly identify users who are highly engaged on weekdays versus those who are more active on weekends. This granular view can lead to crucial insights about user demographics, behaviors, and preferences.

"You can start seeing patterns," Lieb states, "that you would not have seen just looking at aggregate charts or looking at your user logs." This can help founders understand which features are truly driving value for specific user segments, or identify users who are exhibiting behaviors indicative of potential churn or high engagement.

Leveraging Dot Plots for Deeper Insights

Lieb emphasizes that dot plots can be customized to capture a wide range of user behaviors and attributes. For instance, different symbols or colors can be used to represent different actions or user states. This allows for a more sophisticated analysis, enabling founders to segment users based on their operating system (iOS vs. Android), their first usage day, or even their engagement with specific features like creating playlists.

He draws a parallel to GitHub's contribution graphs, which use a similar visual language to show coding activity over time. By applying this concept to product usage, founders can gain a much richer understanding of their user base. For instance, a dot plot could reveal that users who perform a specific action (like creating a playlist) on a Monday are more likely to remain active users in the long term.

Avoiding Common Dot Plot Mistakes

Lieb also cautions against common mistakes when creating dot plots:

• Don't chart the wrong event: Founders should focus on charting events that genuinely represent value or key actions within their product, not just any random interaction.
• Don't pick a time period that's too wide: Using overly broad timeframes (e.g., months instead of days) can obscure the specific daily or hourly patterns that are often most insightful.
• Don't rely on dot plots alone: While powerful, dot plots are best used in conjunction with other analytical tools, such as cohort retention curves, to provide a holistic view.

By meticulously tracking and visualizing individual user actions, founders can move beyond surface-level metrics to gain a deep, actionable understanding of their users. This, Lieb argues, is the key to building successful products that users truly love and continue to engage with.