

Oh interesting-- Yeah gaming the recommendation/search algorithms is another, related explanation. Like I know someone with an Etsy store who says that various things, like running out of stock or putting your store in “vacation mode” will hurt your store’s visibility, so people find ways to game it. Totally makes sense that the same kind of thing would be going on on Amazon.
Weird that they’d allocate ad spend at the same time that they were out of stock, but like I said, the algorithms are pretty coarse and probably just not that coordinated.






Seriously, if you’re motivated enough to do this, you should give programming a try. Python or Ruby or Javascript are ideal for this kind of thing, and you can solve problems like this in a few lines of code… just look up “word frequency in Python” or whatever language for examples.
If you want to see what the next level of this kind of analysis looks like, watch a few videos about how Elasticsearch works… not so much so you can USE Elasticsearch (although you can, it’s free), but just to get a sense of how they approach problems like this: Like imagine instead of just counting word occurrences, you kept track of WHERE in the text the word was. You could still count the number of occurrences, but also find surrounding text and do a bunch of other interesting things too.