I will reiterate, You stated this:
we've been high volume customers since 1998),
This is irrevelant and somewhat ignorant making this statement to the point already addressed. Amazon doesn't care how long you have been a customer since 1998 when considering a temporary ban. You mentioned you have returned a decent amount of items, but have kept the majority of what you have purchased. It doesn't mean that you can't face a temporary ban returning various items.
Which leads me to why I said: "There is no confirmed algorithm on how the consumer is banned". Meaning no one understands what grounds or formula on what they're basing the temporary bans on. From my understanding, there is NO evidence suggesting what triggers the bans. Hence there has not been a confirmed algorithm proving otherwise, but only enough data from others voiced experiences to make an educated guess, just like your suggestion "Everything" is factored in, when in fact, you have zero evidence proving that and others shopping purchases/exchanges have contradicted that otherwise.
I would also suggest that you read on the Internet about others who have made minimal purchases and/or returns to have been temporarily banned from Amazon for such reasons (Who claim to be long time Amazon members. This is not necessarily a new trend, but it's becoming more increasingly sensitive for those who are subject to being banned similarly to what's happening with all the retailers like Best Buy, Home Depot, etc.
i.e., it's either not confirmed, or _any_ metric/attribute might be part of their analytics - and as someone who's been in the tech sector for almost 30 years, with a amount of focus on analytics (including ML for some analytics in the Fed sector and HR markets, that deals with patterns and projections), I'd suggest everything is factored in: age of account, return cycles, types of product, purchase increments, product source, all sorts of granular elements with complex dependencies, etc.
I absolutely couldn't
disagree more with you. I'm not trying to denigrate here, however I think you're uneducated on how retailers are unified with companies like the retail equation, and there are hundreds, if not thousands of voices discussing how they have been banned for different/various consistencies/ranges from online retailers and in-store retailers for either minuscule reasons, few returns, returning AND exchanging Products of All price ranges, separated months from The returns, etc.
I believe you have those who are honest buyers who are victimized by the returns that clueless as to why, because it's been speculated that it's randomized on how TRE is making their selections without having definitive basis, which in fact others scenarios in this thread alone have stated this.
I'm not saying there isn't those who abuse the system that Shouldn't be banned, is there are those route to fraud the system or take advantage of return policies excessively. That said, there are also other customers who have made one or two returns that have been temporarily banned from online retailers like Amazon or in store like Best Buy.
i.e., it's either not confirmed, or _any_ metric/attribute might be part of their analytics - and as someone who's been in the tech sector for almost 30 years, with a decent amount of focus on analytics (including ML for some analytics in the Fed sector and HR markets, that deals with patterns and projections), I'd suggest everything is factored in: age of account, return cycles, types of product, purchase increments, product source, all sorts of granular elements with complex dependencies, etc.
Here is what I can add it from my various experience and/or knowledge:
I used to work in the loss prevention sector years back for a very large retailer, which I both worked in store and corporate. When I did work in loss prevention in store, we conducted 'Customer Audit' reviews Regularly with access to the stores online Citrix Sales system that shows all purchases/exchanges both online and in store from the company. TRE was the company that was affiliated the retailer and had access to the customers data through their drivers license information, which was already in the Citrix system.
Example A:
(Regardless of price point of the Products, considering time as a measurement in below scenarios.)
I can Attest there were customers who were banned from making six spread out purchases a year and out of those six purchases, two of them That were returned/exchanged were temporarily banned.
To contradict how you "Suggested everything is factored", which indeed suggests otherwise:
Hence:
Example B:
And then I had a group of customers Audited who would make over 150 purchases a year, and perhaps return exchange 30/40 out of 150 of those items and Were NOT banned.
