Cutoff: May 12 (latest data) − 8 days =
May 4 — days ≤ May 4 use actual conversions; days > May 4 use a fixed reference rate.
Fixed conv. rate window: Apr 28–May 4 → 47 conv / 177 starts =
26.6% (same rate applied to every projected day).
Week of May 3 | actual starts: 178 | projected upgrades: 49.9 | upgrade rate: 26.6%
2 days settled (≤ May 4), 5 projected using 26.6% rate.
| Day | Starts | Upgrades method | Conv. rate | Upgrades |
|---|
| Sun May 3 | 18 | actual conv. | — | 5.0 |
| Mon May 4 | 36 | actual conv. | — | 12.0 |
| Tue May 5 | 31 | rate × starts | 26.6% | 8.2 |
| Wed May 6 | 27 | rate × starts | 26.6% | 7.2 |
| Thu May 7 | 32 | rate × starts | 26.6% | 8.5 |
| Fri May 8 | 22 | rate × starts | 26.6% | 5.8 |
| Sat May 9 | 12 | rate × starts | 26.6% | 3.2 |
Fixed rate window (Apr 28–May 4) is the last 7-day period where all conversions are fully matured. Using a sliding window for post-cutoff days artificially depresses the rate as it draws in recent starts with no conversions yet.