Back at an old startup when I had a lot of data from clothing retail websites, including historical prices and sale status, I ran some queries out of curiosity

Back at an old startup when I had a lot of data from clothing retail websites, including historical prices and sale status, I ran some queries out of curiosity to look for this. I'm pretty cynical but I was still shocked at the number of results. This seems to be a completely normal practice. Price and being on "SALE" seemed to be optimized or A/B tested almost separately. And the "regular price" field on sale items was garbage data that correlated with nothing.