| @@ -512,11 +512,18 @@ | ||
| 512 | 512 | |
| 513 | 513 | /** |
| 514 | 514 | * Total page views counted across all published forms. |
| 515 | 515 | * |
| 516 | - * Paired with `forms_with_views` and the existing `total_entries`, this is what | |
| 517 | - * makes the conversion figure interpretable in the warehouse: a site with views | |
| 518 | - * but no entries reads very differently from one with neither. | |
| 516 | + * Read alongside `forms_with_views`, which is what makes this figure legible: a | |
| 517 | + * site with views but no forms recording any reads very differently from one | |
| 518 | + * with neither. | |
| 519 | + * | |
| 520 | + * Not a denominator for `total_entries`, and the two must not be divided. This | |
| 521 | + * total starts at the moment tracking opened and leaves out anyone who can edit | |
| 522 | + * the site, while `total_entries` is an unfiltered all-time count of every | |
| 523 | + * entry ever received. A rate built from the pair answers neither question and | |
| 524 | + * will not match the Conversion Rate column, which measures both halves over | |
| 525 | + * the same window with the same exclusion. | |
| 519 | 526 | * |
| 520 | 527 | * Counts only published forms, so views left behind by a trashed or draft form |
| 521 | 528 | * do not inflate the total against a denominator that no longer includes them. |
| 522 | 529 | * |
