eBay is beta testing its shopbot, so how is it?



eBay unveiled its shopbot technology almost two months ago (Oct 18 to be exact). Of course, it is a beta version and naturally embedded in Facebook's Messenger app. As eBay admitted, it has plenty of features and functions still in development. The Beta version also serves the purpose of collecting user feedback, and "train" the shopbot to better interpret human intention thus offer better shopping experience.

 When at launch, the "bot" can only accept dialogues about six product categories: electronics, fashion, health & beauty, home & garden, toys & hobbies, and sporting goods.

 Though eBay claims that the shopbot is backed by artificial intelligence, it is clear the machine-learning algorithms are built on a decision-tree type of structure through which the bot guides a shopper to narrow down his choices until confirming what he wants. This linear conversational pattern doesn't sound fun or engaging and it may even kill the fun when people can browse selections and "discover" new things of interest.

 Another great idea by eBay is to use photo in a chat session to help find the right products. imagine you can link your Pinterest account with the eBay bot and by recognizing products on the photos, the bot can initiate a chat session when the products become available. Unfortunately, according to this Techcrunch article, the photo match function is also a hit-or-miss experience. Bummer!

 Looks like a true shopbot experience is at least a few months away. eBay is mute about shopbot's performance during the Black Friday shopping weekend. Maybe after the Christmas, we will hear more about it. Let's hope it is good news. Or maybe IBM Watson can help?

Comments

Post a Comment

Popular posts from this blog

Fake Reviews...Now a Headache for Amazon

Image
A recent Forbes' article highlights the latest headache that Amazon sellers have encountered: fake reviews that game Amazon's system by providing favorable visibility to those improperly-promoted products. The article is based on ReviewMeta 's findings, which show that the average review weight--a measure developed by the firm to calculate how trustworthy reviews are overall--has nearly halved during June, July and August of 2017, compared with those in October 2016. That month is particularly important because it was when Amazon banned the practice of incentivized reviews--people receive free products in exchange for writing reviews. Most such reviewers disclosed such "free for review" relationships in their reviews, but plenty didn't, which prompted Amazon to ultimately ban this practice. It is easy to reason that shrewd marketers now find another way to game Amazon's review system. How do they recruit reviewers? Simple, through Facebook groups.
Read more

Will Amazon's Exclusives Store Help Alleviate the "Commerce Pirates" Problem?

Image
In its most recent earnings report, Amazon.com for the first time shared the performance of its Exclusives store, a site where brands can test new products and buyers hoping for exclusive deals can find their likes. Amazon said that since launch early last year, the site has generated $50 million sales from about 120 brands that offer about 10,000 products in total. Amazon is apparently encouraged by this experiment, and plans to expand this store format. "We do expect the broadening of the product categories and geographical expansion," according to Peter Sauerborn, director of business development at Amazon who is in charge of this effort. It would be good news for retailers and small merchants that have truly unique merchandizes to sell. As the e-commerce market stands right now, the marketplace model opens up a great revenue opportunity for small and medium-sized retailers, but because a web store is so easy to set up, the marketplace model also invites what I c
Read more