BigCompany A/B testing has this unfortunate tendency to demonstrate positive results to whatever you are testing. Like the A/B test that repeatedly demonstrated that after you buy a blender, your most likely purchase is another blender... (which Amazon still hasn't fixed nearly a decade later)

Avoiding triggering buyer’s remorse is something high end consumer business understand and Amazon emphatically does not. I just bought a kayak, don’t advertise $3000 kayaks to me. Advertise dry bags and paddles and dry suits, you numpties.

It’s because Amazon is an e-commerce store. They only win if you return and buy a more expensive one from another brand on Amazon. Brands however have a brand reputation to preserve, hence the need to avoid triggering buyer’s remorse.

Are you sure that buying another blender isn't common?

1. Maybe you didn't like it so returned it, but are still in the market.

2. Maybe it was a gift and you now want one yourself (or vice versa)

For something relatively expensive it may be the best expected value add even if the chance of repurchase is only 1%. Maybe the repurchase rate is 1% while the base purchase rate is 0.1%.

I'm not sure it is, but it seems entirely possible that this is actually a smart move by the ad engine.

> BigCompany A/B testing has this unfortunate tendency to demonstrate positive results to whatever you are testing.

No, it doesn't. It tests which one is more positive, not which is is positive.

> Like the A/B test that repeatedly demonstrated that after you buy a blender, your most likely purchase is another blender... (which Amazon still hasn't fixed nearly a decade later)

That isn't an A/B test.

> No, it doesn't. It tests which one is more positive, not which is is positive.

"positive" in the sense of "always answering questions with 'yes'", not in the sense of "positive numbers". Turns out when you start linking people's compensation to their datascience results, the results start to lean in whatever direction optimises for compensation...

> That isn't an A/B test.

A series of A/B tests. All the alternatives failed to produce better metrics than the status quo, so everyone became satisfied that their local minima was optimal

I believe you’re significantly overstating how good and thorough a job they do of such things. Also it’s not the sort of thing that’s particularly conducive to A/B testing.

Amazon has one of most ridiculous and terrible UI's that are outdated and difficult to navigate with so I wouldnt take anything from there to follow any standards

There's a big "if" here - IF Amazon is actually using them to maximize profit. There's a ton of reasons they might use them even if suboptimal - vendor agreements, advertising contracts, loss leading, idk who knows