THE RESULT
What the data shows
The largest computed metric is 0.3167 for (100.0, inf]; the smallest is 0.04118 for (0.0, 5.0]. Metric: rate (Purchase-session share).
Page counts are measured over the completed session and can be consequences of purchase intent.
| depth | events | n | rate | wilson lower | wilson upper |
|---|---|---|---|---|---|
| (-1.0, 0.0] | 6 | 38 | 0.1579 | 0.07444 | 0.3042 |
| (0.0, 5.0] | 96 | 2,331 | 0.04118 | 0.03384 | 0.05003 |
| (10.0, 25.0] | 568 | 3,498 | 0.1624 | 0.1505 | 0.175 |
| (100.0, inf] | 229 | 723 | 0.3167 | 0.2839 | 0.3515 |
| (25.0, 50.0] | 509 | 2,507 | 0.203 | 0.1877 | 0.2192 |
| (5.0, 10.0] | 185 | 1,804 | 0.1025 | 0.08938 | 0.1174 |
| (50.0, 100.0] | 315 | 1,429 | 0.2204 | 0.1997 | 0.2427 |
THE METHOD
From source to answer
Predeclared page-count bands and binomial rate intervals.
The Python source exposes this study’s transformations. The complete project download includes shared preparation and evaluation routines.
THE NEXT DECISION
What follows from the finding
Test a specific navigation change with random assignment rather than assuming more pages cause purchases.
Where the conclusion stops
One observation is a completed session. End-of-session behavior cannot support an early-session prediction claim. PageValues is excluded from purchase models because it can encode downstream purchase information. Associations are not experiment results.
Related studies may reuse observations or holdouts. These are historical analyses; associations and backtests do not demonstrate commercial impact. Further model tuning needs new, untouched evaluation data.