How To Better Understand Your Users

Episode Summary

Executive Summary: The speaker argues that founders over-rely on aggregate metrics like DAUs and MAUs, which hide how individual users actually behave. He introduces dot plots—2D grids showing each user’s activity over time—as a lightweight but powerful way to reveal usage patterns, retention issues, feature impact, and risk signals in both consumer and B2B products. He positions dot plots as complementary to cohort retention curves and useful even at large scale through sampling.

Main Topics: Why aggregate metrics are insufficient (Priority: 5/5): DAU/MAU and similar charts obscure per-user behavior and can look healthy even when users are not getting value from the product. Dot plots as a user-behavior visualization (Priority: 5/5): A dot plot maps users by rows and time by columns, with dots indicating meaningful product actions, revealing detailed usage patterns over time. Patterns revealed by dot plots (Priority: 5/5): The visualization can expose weekday vs. weekend usage, onboarding/retention problems, and differences in user segments that aggregate graphs miss. Extending dot plots with richer encoding (Priority: 4/5): Dots can represent different events, states, platforms, geographies, or demographic attributes, and rows can be sorted to surface meaningful comparisons. Business value for consumer and B2B products (Priority: 5/5): Dot plots can diagnose product engagement in consumer apps and flag churn risk in B2B accounts when seat activation and usage are weak. Best practices and pitfalls (Priority: 4/5): Choose a value-creating event and a fine enough time interval; avoid vanity events like app opens and overly coarse weekly aggregation. Complement to cohort retention curves (Priority: 5/5): Cohort curves show whether users stick around in aggregate, while dot plots explain how and why users are behaving, making the two tools jointly powerful.

Key Arguments: Founders should not rely on aggregate metrics alone because they can hide whether users are actually deriving value from the product. The most important signal is how individual users use the product: frequency, pacing, feature usage, and behavioral patterns. A dot plot makes individual usage visible while still allowing a high-level view of the whole product. This visualization can reveal retention issues, segment differences, and feature-specific patterns that suggest causality. Dot plots scale by sampling users, so they remain useful even for very large products. In B2B, dot plots can show whether purchased seats actually activate and whether an account is at risk of churn. The chosen event must represent real user value; shallow events like app opens are misleading. Dot plots work best alongside cohort retention curves, not as a replacement for them.

Data Points: Users in example DAU graph on day 1: 2 - Illustrative aggregate usage for the Spotify-style example Users in example DAU graph on day 2: 3 - Illustrative aggregate usage for the Spotify-style example Users in example DAU graph on day 3: 2 - Illustrative aggregate usage for the Spotify-style example Users in example DAU graph on day 4: 2 - Illustrative aggregate usage for the Spotify-style example Users in example DAU graph on day 5: 2 - Illustrative aggregate usage for the Spotify-style example Users in example DAU graph on day 6: 1 - Illustrative aggregate usage for the Spotify-style example Users in example DAU graph on day 7: 0 - Illustrative aggregate usage for the Spotify-style example Users in example DAU graph on day 8: 1 - Illustrative aggregate usage for the Spotify-style example Seats purchased in B2B example: 10 - A customer bought 10 seats for an ~$80,000/year contract Seats activated in B2B example: 3 - Only three seats ever used the product Contract value: $80,000/year - Referenced in the churn example with a name-brand customer Scale mentioned at Google Photos: more than 1 billion users - Dot plots were used there via sampling at massive scale Time to build a dot-plot tool with AI coding tools: about 10 minutes - Speaker says modern AI coding tools can generate it quickly Time granularity recommendation: day or sub-day granularity - Speaker warns against weekly aggregation for this visualization

Pivotal Quotes: "One of the biggest mistakes I see founders make is relying on aggregate user metrics instead of understanding how any individual users use their product." — Dave: Opening thesis on why per-user visibility matters "The number one thing is to just chart the wrong event." — Dave: Warning about choosing vanity metrics instead of value-creating actions "Cohort retention curves and dot plots are, in my experience, two of the most important tools that you've got to understand your users." — Dave: Closing summary of the recommended analytics toolkit

Implications: Founders should instrument product analytics around meaningful actions and review per-user patterns, not just topline growth charts. Dot plots can uncover retention, segmentation, and churn risks earlier, especially when paired with cohort curves.