Methodology
12 Mondays. One honest comparison.
This analysis compares Active and Non-Active Speed Layer cohorts across 28 McLarty properties.
What was measured
The cohort covers 12 Mondays from May 4, 2026 to July 20, 2026. It is not a consecutive-day comparison.
- Active
- Users identified by the AG SpeedLayer Active audience.
- Non-Active
- Total users minus users in the Active audience.
- Lead
- ASC form submission, click-to-call, and communication submission events included by the source workbook.
Why Mondays
Mondays carried a meaningful live split, with roughly 39% of users Active. Tuesday through Sunday showed only about 5-8% Active share and appeared to include audience carryover, so those days are not used for the comparison.
The result holds under a stricter read.
2 properties had no valid Non-Active lead conversion baseline. Excluding them leaves 26 properties with baselines on both sides.
What this does not claim
This is an observed Monday cohort comparison, not a laboratory experiment. The results support a measured rollout decision, but they do not prove that every individual outcome was caused by Speed Layer.
Provenance
Calculations come from the July 27 McLarty Monday cohort workbook and PageSpeed workbook supplied by Alex Oleynik in Ben Hadley's Touchless Slack DM thread. The CRO opportunity library is derived from the supplied 20-page mobile audit.
Sanitized data version: mclarty-2dd042bd