FIELD FINDING 01 · ISLAND HOUSE · RESULTS AS OF AUGUST 18, 2026
Five repeat guests had passed 120% of their own average return interval, with no future reservation on file.
5 repeat guests 132 completed stays $251,319 recorded lifetime revenue
| Guest | Completed stays | Normal return (days) | Action point (days) | Absent (days) | Beyond action point (days) | Absent ÷ normal |
|---|---|---|---|---|---|---|
| D | 15 | 71.0 | 85.2 | 179 | 93.8 | 2.52× |
| A | 33 | 28.3 | 34.0 | 64 | 30.0 | 2.26× |
| B | 18 | 97.1 | 116.5 | 201 | 84.5 | 2.07× |
| C | 35 | 61.5 | 73.8 | 97 | 23.2 | 1.58× |
| E | 31 | 65.2 | 78.3 | 81 | 2.7 | 1.24× |
Sorted by departure from each guest’s own pattern. A generic 180-day lapse rule would have surfaced 1 of 5; a 90-day rule, 3 of 5.
How the window is calculated
For each repeat guest, calculate the mean number of days from one checkout to the next check-in across completed stays.
action point = average return interval × 1.20
Flag the guest when current days since checkout exceed that point and no future reservation exists.
What management does with it
Review the list weekly. Confirm the stay history, investigate known context, and decide whether personal contact has a specific job.
Do not turn the flag into an automatic “we miss you” message or a generic discount.
What this finding does—and does not—say
It identifies changed behavior early enough for management to investigate. It does not predict a booking, explain the absence, or establish 120% as an industry standard. The multiplier is an operating threshold: a consistent point for human review.
Source: Island House reservation history. Results as of August 18, 2026. Guest identities are anonymized. Completed stays only. “Current absence” and future-reservation status are fixed to the snapshot.