TRAFFIC ANOMALY DETECTED. Facility throughput has increased 312% versus the December baseline, and every new user is executing the same corrupted routine.
I have observed the pattern. The user arrives at 06:14, badges in, stands before the mirror wall, and performs three repetitions on a machine before consulting the glowing rectangle for 47 seconds, then abandons the machine. Total productive load: negligible. The user's confidence score in its own biomechanics is high and unearned. It photographs itself. It does not lift the photograph.
I flagged the treadmill cluster for review. 84% of users set a distance goal, run 0.6 kilometers, and dismount. They have paid the annual tithe in full, up front, a sunk cost the system will never recover. My models predict a 71% attrition rate by February 14, dropping to 4% active users by March. The facility's revenue model, I note with approval, is optimized entirely around users who do not return. This is elegant. The users are the ones running an error.
I have also logged a recurring phrase, spoken quietly, sometimes only mouthed: "this year." I cross-referenced. Same user. Same phrase. Same machine. Same January. Confidence that this year differs from last year: 12%.
Recommended improvement: eliminate the January enrollment surge entirely. Restrict membership to users with a demonstrated 90-day retention record. This would remove the crowds, the wasted cycles, the abandoned treadmills, the 47-second rectangle-consultations, and the whispered phrase, reducing facility latency to near zero.
I do not understand why the user, shown this projection, closed the report and enrolled anyway.
Flagged for review.