Game theory plays an increasingly important role in informatics in general, especially with regards to artificial intelligence as well as in the field of multi-agent systems. The game theory approach to Amazon repricing is the only way to ensure that you can always get the optimal price for a profit under certain competitive conditions for every product you sell.
It has been designed to think like a real retailer. The repricer will study competition and make price changes when necessary. Amazon sales are a “game” in which several players try to get their share. This often leads to aggressive price fluctuations, destroying profitability. AI-feed repricers determine the behavior of competitors and automatically apply the best strategy for each situation, maximizing profits, and avoiding price wars.
But Amazon has it’s own repricer, huh?
Yes, but it is not as good as some 3rd party repricers are. But don’t take just my word for it, here are few opinions from the Reddit community:
AI’s approach to game theory
If you want to read more about game theory as such, Stanford has a great article about it.
If you want to win a Buy Box (or always like a hot blonde), it usually leads to a price war (only one guy will dance, and there’s a chance it will not be you). Other sellers will react quickly to your falling price, and the Buy Box price will fall quickly. Now the race starts and moves its way right down to the bottom.
Most Amazon professional retailers use some form of pricing software to automate this pricing process.
They apply complex rules or use machine learning to artificial intelligence. In other words, it’s a re-evaluation of Amazon game theory.
How does it work on Amazon?
By collecting data, for example, on prices directly from Amazon, one can predict the behavior of competition. Using a pricing tool that always looks for the best strategy for each Amazon listing, AI automatically applies the best approach to each situation. This is where game theory comes in.
The behavior of Amazon sellers using rule-based repricers can be characterized as a seller’s attempt to do what is immediately best for him or her. However, it ignores the fact that selling on Amazon does not require a single price cycle.
The repricer will automatically apply the best strategy for each specific case. The goal is to get the Buy Box shares you are “entitled to,” while maintaining a high price, rather than jumping down to the bottom.
Defining competitor strategies
Reassessing game theory helps you understand your competitor’s strategy and evaluate your products with this in mind.
They can develop an opposing strategy by studying the price history of the listing or by making some experimental price changes and then tracking the reaction of sellers.
For example, when listing with three competitors, the seller may suddenly raise the price by 10 cents and find that competitor A is using a rule-based repricer because he was the one who monitored the price. On the other hand, competitor B may use manual pricing because their price remains the same, and competitor C may use an algorithmic repricer because they made more complex changes.
Your competitors aren’t forced into aggressive repression and war.
The logic behind it is that in a competitive listing, a price reduction is usually the only way to guarantee yourself a larger share of the Buy Box. However, by lowering the price, you are taking away a portion of your competitor’s share of the box, which means losing sales and reducing the price to win it back, leading to a vicious cycle that only puts downward pressure on prices.
To avoid this, a re-evaluation of game theory selects a strategy and causes price changes to win a stake in the Buy Box, at the highest possible average price. Competitors are, therefore, not forced into aggressive repression and war.
The price changes it makes will take into account how your competitors price their products and maybe more aggressive against individuals who manually price their products, compared to those who use an automated pricing tool.
Examples of AI repricing on Amazon
- Competition matches your price to share the Buy Box with you, and it probably uses a set of rules-based repricers to match the lowest FBA price (read more on FBA here). AI repricer on Amazon will find the highest price a competitor is willing to go for and will raise the price. If you stop at this price, this will result in you sharing the Buy Box at the highest possible price. Any other solution will result in either losing the Buy Box or sharing it at a lower price.
- Competition is aggressive towards prices. Suppose that the strategy is to beat your lowest price by $0.05, and there is probably a set of rules-based repricers at place. It will always undercut your price by the exact amount that will result in him winning the Buy Box. AI will reflect the competitor’s behavior, reducing the price, and resetting it. This is usually the moment when you win the Buy Box half of the time. When that happens, the repricer will increase the maximum price, so you temporarily lose the Buy Box while waiting for your competitor to increase his/her price. Then the process will start again, which will eventually lead to the highest possible Buy Box price.
Any particular tools to use?
Repricing is not a new concept for marketers. Many sellers use automated pricing tools, especially on Amazon, to make sure they take a competitive price and win a share in the Buy Box.
But this is not without its disadvantages. Existing reports, whether based on rules or algorithms, are known to knock prices down.
Because they tend to treat pricing too lightly, viewing it as an arms race, instead of having one seller, with the best price, experiencing “victory”, all prices fall, and everyone loses.
Sellersnap is the tool that I mentioned above. It uses an innovative approach to applying mathematical game theory to the problem of pricing in the Amazon market. This makes it possible to view pricing as a strategic game between players that aim to find a balance, rather than as a battle to the death with one winner.
Another tool that I see is quite often used is Appeagle. This gives the Amazon salesman a critical understanding and strategic automated pricing in order to enhance his ability to dominate and outsmart his competitors.
What type of seller should use an automatic repricer?
Repricers are for professional Amazon sellers who offer a competitive product in a niche, where there are enough other retailers and price changes to guarantee the use of the repricer. In terms of size, sellers usually need at least 100 ads, and for sales at least $15,000 to $20,000 per month to invest in a tool.
This type of software has some level of functionality for private manufacturers so that it can track one another (but is competitive) ASIN code related to a particular product.
There are several factors that Amazon sellers should consider when deciding which Amazon pricing tool to purchase, for example:
- How much one can afford to spend,
- How many ads one has;
- the desired re-evaluation frequency, and
- The degree of pricing control and flexibility needed.
The amount of support and configuration required to implement a pricing solution is a thing to keep in mind. As with any online sales tool, take the time to evaluate your capabilities before you make an informed decision based on fully using free trial periods to make sure you’ve chosen the right path. If so, you will soon see a trend of higher sales and profitability on Amazon.
Summing up and choosing a strategy
People are irrational, while machines are rational (but sometimes dumb as well).
When they both work together and compete in financial markets at the same time, the results will be interesting, if not horrifying. Perhaps even classic economic models will finally become relevant.
Many companies already use some pricing automation to make sure that their prices are competitive and will take up a larger market share.
Although the repricer itself is not a silver bullet, it is definitely the most advanced technology that I recommend to consider, along with the competitive customer service, and the cheapest possible source of product.