The question looks more or less OK, except the ambiguity part. Hower, perhaps after "25 years in Big Tech" one shouldn't invent a process that discriminates against "loyal" customers living in timezones unaligned with the logging server. As usual, users come last.
I think unaligned customers are discriminated in favor of--more likely to be browsing around the server's midnight and get marked as visiting two days.
I don't think you can decide if it's a net positive/negative discrimination with any degree of certainty without knowing the date-stamping machine's location and the demographics of the userbase. However, there most certainly will be a non-zero amount of users discriminated against (browsing during a span around their local midnight that falls onto a single date on the servers), and that's what matters most in my opinion.
At the end of the article he elaborates the problem in a way he seems to feel is too complex to mention in the interview itself: what if there's a third day?
If there are three server days of logs and a loyal customer continues to be defined as one who visits on two different days, the timezone problem essentially goes away. More fully:
- if a loyal customer is one who visits regularly;
- and we sample several days of visits;
- then loyal customers will be detected regardless of their timezone.
If you want the concept of a "loyal customer" to match the detection threshold, the weirdness related to timezones will still exist, but if you think of the detection threshold as a tool that's good enough to detect loyal customers, then it won't.
