EBay Geographical Sales, A Possible Answer
If you sell on eBay, have you ever noticed to you sell a large portion of your items to one particular area ? Maybe you’re getting a cluster of sales in Yorkshire, Cornwall or South Wales. Have you ever wondered why this is?
EBay’s best match algorithm works in a similar way to Google’s it’s based on relevancy. When a user enters a search term on Google, Google will display what it determines as the most relevant results. For example if you enter the search term “ red shirt “ on Google you would expect to see a first page result with sites that all have some connection to red shirts, maybe they make them or sell them but they are all relevant to the search term. You would not expect to see any sites that make or sell blue shirts.
One of the ways Google judges how relevant each site is is by looking at past results. It remembers that search term being entered previously and it remembers which site that user/s clicked on. That site is then deemed more relevant and appears higher in Google’s listing. However, that still wouldn’t break down the results enough for Google to return an accurate result. Thousands of users probably searched for “ red shirt “ and probably clicked on thousands of different sites.
Google has to break these results down even more and one of the ways it does this is by looking at the location of the user. If you search on Google for “ local wine bar “ look at the places results. They will all be local to you and of course, you will get a different result if you try it in a different location.
So, what has this got to do with geographical sales on eBay? EBay does the same thing with it’s search results, however the results it returns in best match are based on sales from a listing rather than clicks on a website. For example search on eBay for a red shirt and will get many results, the ones at the top of page one are the listings that have sold more red shirts to people who have searched for red shirts. These listings are judged by eBay to be more relevant and so it gives them a boost in the listings.
Ok, here comes the theory, what if eBay breaks these results down even further ? For example, someone in Yorkshire searches on eBay for a red shirt and best match not only looks at the search term to provide the most relevant results but it also looks at the geographical area the user is in. In this scenario the user in Yorkshire would be shown results from sellers who had previously and recently also sold, not only a red shirt, but a red shirt to a buyer in Yorkshire. So the buyer in Yorkshire would see a skewed result in favour of sellers who had to sold to other buyers in that area. and would be more likely to buy from the same seller.
This would result in a cluster of sales going to one area which would have been “ set off “ by the first buyer making a purchase.
This effect would be amplified in areas of low population. I am led to believe that some eBay buyers live on the island known as Scotland and that some even live in the area of this island known as “ The Highlands “ Joking aside, there is comparatively sparse population in The Highlands compared to it’s actual size. Compare it’s size and population spread to London or Manchester for example. So, a buyer in the Highlands searches for a red shirt and eBay has to judge the most relevant results that that buyer would want to see. EBay search looks for listings that include the term “ red shirt “ but it also recognises that the buyer is in The Highlands so it shows listings that buyers in The Highlands have previously bought from. If the buyer was in London this would still return a lot of listings because of the dense population in London but due to the smaller population in The Highlands it will return fewer listings and possibly even give a higher weighting to the geographical aspect of the search. This would result in the same listing/s being shown to this area and gaining more sales.
It’s important to remember that this chain reaction in search results will have been caused by the first buyer of the red shirt being in The Highlands. So, how did the very first buyer of the red shirt, who lived in The Highlands come to buy from the seller who subsequently saw a spike in sales to The Highlands? Simple, the first buyer used a longer keyword phrase. Maybe he looked for a red shirt with white cuffs or a red shirt with red buttons and this longer search phrase narrowed the results down to a specific seller. That seller then saw a rise in sales to the original buyer’s location because the listing is seen as more relevant to buyers in that location.
To summarise my theory; eBay search has to break down it’s results to provide listings relevant to the user. It looks at the search term and says “ what have people who searched for this item before actually bought “? But eBay also looks at the buyer’s location and says “ what have people who searched for this item and that lived in xxxxxx actually bought “ This geographical chain reaction is set off by the first buyer using a longer keyword phrase to find a particular listing.