My Amazon Experiment, Part 2

This is a continuation of Thursday's post.

With those observations, I had an understanding of how Amazon's prices change in relation to new sellers.  I thought that I could artificially create new clusters simply by setting my price equal to that of the fourth- or fifth-cheapest item for sale.  Any items whose sellers didn't match my price would be quickly purchased, leaving only two or three items matching my price, allowing me to raise the market price and repeat the process.

My method worked.  I was able to raise the low price for Sonic Adventure 2 from $25 to $35 (a forty percent increase) in a one week period, despite only two copies of the game selling.

A few market conditions which aren't immediately obvious prevent this from being a viable technique.  First, and most importantly, are the volume sellers.  These are the people (well, companies, I'm sure) that have feedback scores in the hundred thousands.  They make their money by moving a lot of products with small margins.  They undercut the low price by a fair amount to be sure they have a quick sale.  For example, when Sonic Adventure 2 was $35, a volume seller listed a copy for $20.  New sellers then see the low price is $20 and start inserting their copy between $20 and $25, undoing all your market manipulation.  What's even worse is when there are multiple volume sellers attempting to undercut each other; at one point the $20 low price was cut down to $12 so there were many copies priced between $12 and $20.

Another issue is the people who attempt to undercut your price by $0.01 to $0.05.  Numerous times I had the cheapest copy, only to see someone price there copy $0.01 cheaper than mine and sell it.  You can't beat these people at their own game because when you reduce your price they reduce theirs within a few minutes; I'm guessing they use a bot.  For a week-long period I had three different sellers use this technique against me (all three of there copies sold) until eventually a volume seller undercut as all by $20.

I guess the conclusion of my experiment is that you can influence the Amazon market for limited-seller items but only in the short run.  In the long run, there are enough sellers who don't care about getting the maximum profit for this one item so you can't inflate the price too much.  The most you can get is a couple extra dollars if you're lucky.