eBay SEO Hack – Das geheime eBay Patent erklärt | So funktioniert die eBay Suchmaschine Cassini
Welcome to the YouTube channel of eBakery, the e-commerce agency. The search engine with the so-called Best Match algorithm was introduced twelve years ago and six years later a patent was registered with the name "methods and systems of deriving a score with which item listings are ordered when presented in the search results". We're talking about "Cassini".
In today's video, referring to Marc Steier's article, we want to explain how the Ebay search engine works based on the patent. Every search query that operates on Ebay returns several lists of items, which are assigned a ranking based on various scores. The most important scores are the relevance score, the business score and the quality score. We can find out what these say in the excerpt from the US patent. As can be seen, the Best Match Score is composed of these three important individual components: "relevance", "listing quality" & "business rules".
Accordingly, one can see from the diagram that the relevance score is made up of how likely it is that a purchase, i.e. a transaction, will take place if the search results page were to be displayed in this way. In summary, this means that the score indicates how likely it is that a purchase will be made if product x were ranked higher for a certain search term. This also explains the naming. What is meant is how relevant the product is for the keyword entered. Let's skip the listing quality and look at the business rules. This result is essentially based on revaluations and devaluations and also includes seller data, as can be seen in this patent excerpt. This means that not only the optimization of the product listing is decisive, but also that seller data is generally included in Cassini's ranking assignment. Marc Steier gave me a very interesting diagram of the listing quality, which includes typical parameters such as the number of views, the number of transactions, the price, the shipping costs, but also the length of time the offer has been online , which illustrates why it sometimes makes sense to delete and re-list Ebay offers that have been active for a long time .
This was previously referred to in a video that I will link to you in the top right corner. Here we see a formula, which in turn is made up of three parameters: the predicted "listing quality score", the observable listing quality score and a time parameter called "weight". This time parameter changes over time and has a variance radius of 0 to 1. This means that at the beginning, after the eBay listing has been posted, this value is 0. Accordingly, the entire listing quality is only composed of the value of the foreseeable LQ score. That is, from the price , condition, offers, shipping methods, and so on.
However, the longer the offer is online, the higher the time value "weight". This means that if this reached the maximum of 1 , then the predictable LQ score was multiplied by 0. Because 1 plus 1 equals zero. In this case, the entire Listing Quality Score is defined only from the value of the observable score. If no more clicks and no more transactions are generated, then the offer falls in the ranking.
That's why it sometimes makes sense to end offers that don't sell and put them back on Ebay. Hopefully you have now been able to understand to some extent how the Cassini search algorithm is fundamentally composed. Of course there will be some changes and deviations in the near future due to updates. This innovation can, for example, concern the personalization of the search result display. This means that person A can be shown something different than person B, even though both enter the same keyword.
If you want to know more about it, we will link you to the word filter article in the description. Nonetheless, the basic concept embodied in the US patent will still remain, so it pays to take the time and understand how it works using the patent and then incorporate this into your own ebay selling strategy. With this in mind, we wish you every success in the implementation! Feel free to leave a like so that more complex topics can find their way into our channel in the future and subscribe so that you don't miss out. Until then, your eBakery team!.
