Anscombe's Quartet

You're seeing it when two product cohorts show the same average retention rate, but one has steady engagement while the other has a small group of power users masking mass churn.

You're seeing it when two investment portfolios show the same average return, but one is stable and the other is volatile with a single outsized gain pulling the average up.

You're seeing it when two factories report the same average defect rate, but one has consistent quality while the other alternates between perfect batches and disaster batches.

When your team presents metrics — average deal size, mean time-to-close, NPS scores — demand the distribution. Averages can mask bimodal realities where your business is serving two fundamentally different segments. The strategic response to each is different.

Simons built the most successful quantitative fund in history by insisting that models be validated against the actual structure of data, not just summary statistics. Renaissance Technologies' edge came from detecting subtle patterns that standard statistical summaries — means, correlations, regressions — would miss entirely. Simons understood Anscombe's Quartet intuitively: the same summary can describe radically different realities. For founders relying on data for decisions, the lesson is that aggregated metrics are starting points, not conclusions. Always look at the shape of the data before acting on the number.