Fun with maths - grey market and waiting lists

[EverCloserUnion]

Originally Posted by GrailTime

I'm no expert on statistical modeling, but i'm not really sure how the "waitlist" can be predicted in anyway.

Way too many variables, especially when what an AD receives is sporadic and who they call is based upon spend history, relationship, potential future spend history, your job, the net worth that the sales assistant guesses you have and generally if your face fits.

Agree that there are myriad variables.
My assumptions were based on 'minimum standing'... so a profile that should be sufficient to get on a list in some way ceteris paribus.

Such a forecast can only ever be indicative.

Obviously there will be those who for whatever reason never receive a call for a watch that shouldn't be too difficult to get, and perhaps one person in a million landing something desirable as a zero spend walk-in, but those should be extreme outliers.

[PenDelicate]

Originally Posted by GrailTime

I'm no expert on statistical modeling, but i'm not really sure how the "waitlist" can be predicted in anyway.

Way too many variables, especially when what an AD receives is sporadic and who they call is based upon spend history, relationship, potential future spend history, your job, the net worth that the sales assistant guesses you have and generally if your face fits.

Oh, it can. Insurance companies (and now internet advertisers) are entirely dependent on massive multi-variable analysis like this.

Your initial model could be really simple: average wait-list time. Just adding-up all the time people have spent waiting and taking the average, then taking a one-sided normal statistical distribution about that average. This would be a much better model than random guessing.

And every variable you add to your model will conceivably improve its accuracy.

The model might be terrible at predicting one specific person's wait-time, yet be excellent at predicting statistical trends.

[jmgorman]

What's astonishing to me is how much money Rolex is leaving on the table. They surely have great manufacturing and delivery data at their disposal - imagine how much they could charge for a subscription to that! I'm betting tens of thousands of watch influencers would pay $149/month for access to a live dashboard of stats that might help them juice the ADs and better draw clicks telling us "How to get a new GMT today!"

[gamingdoctor]

Originally Posted by jmgorman

What's astonishing to me is how much money Rolex is leaving on the table. They surely have great manufacturing and delivery data at their disposal - imagine how much they could charge for a subscription to that! I'm betting tens of thousands of watch influencers would pay $149/month for access to a live dashboard of stats that might help them juice the ADs and better draw clicks telling us "How to get a new GMT today!"

They've just started their CPO program to capitalize on the artificial shortage they themselves have created.