Reading the Room on Open Rates

Averages are the most abused statistic in publishing.

Consider two lists with identical aggregate engagement: one dispatch that reached a thousand readers and one that reached three. A flat average treats them as equal. Weighting each issue by its own delivery count does not.

What we recommend

• Weight every issue independently, then average the weights.
• Count an issue with zero deliveries as a hard zero.
• Publish the denominator alongside every percentage.

The full methodology is documented openly, because a metric you cannot audit is a metric you cannot defend.