eBay’s Smart Search: A Peek into the Future of Best Match
Filed Under Best Match, Industry News & Analysis
eBay recently acquired the internet start-up Positronic and brought aboard the firm’s two founders, Christopher Payne and Dane Glasgow as Vice President of Search and Vice President of Engineering respectively. This move provides some indication that eBay’s recent and significant changes to their search and sort algorithms are but the tip of the iceberg. eBay is obviously committed to expending additional resources and energy to search in the future.
Both Payne and Glasgow were executives involved with the production and fielding of Microsoft’s Live search engine prior to leaving a couple of years ago to found Positronic. According to Wikipedia:
A positronic brain is a fictional technological device, originally conceived by science fiction writer Isaac Asimov. Its role is to serve as a central computer for a robot, and, in some unspecified way, provide it with a form of consciousness recognizable to humans.
This although it is not entirely clear what Positronic was specifically working to create, the firm’s name indicates their purpose was the development of artificial intelligence and “working in the areas of data mining, machine learning, and predictive models, all applied to search.”
Indeed, previously the Positronic web site had the following statement displayed on their home page (It now has a statement indicating the firm has be acquired by eBay):
Are you inspired by Isaac Asimov’s vision of thinking machines? Ever pondered how and when the break-through will come to make it possible?
Wow! That’s heady stuff! eBay’s management has obviously been “inspired by Isaac Asimov’s vision of thinking machines.” According to the corporate blog, “eBay has acquired Positronic to help with efforts at leveraging machine learning to provide a more predictive and compelling customer experience.”
Here’s some personal predictions on the future of eBay search & sort…
Prediction #1
eBay will monitor, track and store users browsing and shopping activity on the site in order to provide search results sorted in a manner relevant to the individual based on that person’s past actions.
Here’s a hypothetical example. Accessories compatible for a Dell XPS M1130 laptop computer will be listed at the top of the search results when I conduct a search using the query “laptop battery.” This is because the site will remember that I purchased that particular model on eBay a few months ago.
Or perhaps the smart search and sort algorithm will note that I have a penchant for the original Star Wars comics series published by Marvel vice those published more recently by Dark Horse comics. The algorithm will know this because I click on listings for Marvel Star Wars comics and ignore those from Dark Horse. So, in future searches, the algorithm will determine that Marvel comics are relevant to me (or anybody using my User ID) and sort them at the top of the search results accordingly.
Something like this is already happening on eBay. On the main page, the site displays a collage of listings it considers similar to one’s you’ve viewed recently.
This, of course, will raise all kinds of privacy concerns and will be difficult for eBay to implement without causing a ruckus.
Prediction # 2
Completing the Item Specifics form will become required when listing an item for sale. This is currently optional.
Item Specifics are the details sellers have the opportunity to provide when they list their product for sale on eBay. For example, a listing in the Book category might provide the seller the ability to enter details such as hard cover or paperback; new or used; fiction or non-fiction; year published; etc. The Item Specifics function varies by category and may duplicate information the seller has included in the listing’s description. This data allows eBay’s software to know specific information about the item in a format the site can use to help potential bidders find a listing.
Currently all eBay has to determine the relevance of a product associated with a listing are the 55 characters in the title and the details included in the item’s description. Since there is no standard format for details included in the description, eBay can’t effectively use that information to ascertain the details associated with a product.
But, then again – Positronic was working on data mining, so I could be dead wrong here. Therefore my revised prediction is Item Specifics will be required, then abused by sellers (much like meta keyword tags were in the early days of search engine optimization). Item Specifics will then fall out of favor, to be replaced by the data mining techniques introduced by the Positronic folks.
What are your predictions for the future of Best Match and eBay’s Smart Search?
Related Posts:
Comments
RSS feed for comments on this post. | TrackBack URI
One Response to “eBay’s Smart Search: A Peek into the Future of Best Match”
Leave a Reply
Amazon pioneered personalized search and recommendation engine for widgets. eBay has a big catch up game to overcome.