IBM Watson: A Consumer App for Toy and Gift Buyers
November 18, 2015
When I read “IBM Watson’s New App Predicts the Must-Have Toys and Gifts,” I wondered if dead tree catalog makers would embrace Big Blue’s approach. The idea is intriguing. IBM Watson crunches data and generates outputs that guide the person looking for a toy for a grandchild or a gift for an office secret Santa party.
I learned:
The iOS app surfaces suggestions in a number of categories, like tech, health and toys and shows products that it thinks will be the most popular in each section. It also aims to predict trends so you can see whether an item will be popular throughout the holidays or if it will be a passing fad. Behind the scenes, Watson analyzes the conversations around products from social media, blogs, reader comments, product reviews and ratings to determine what will be the most on-demand items. It also takes sentiment (remember Watson’s personality identifier tool?) into account to detect exactly why something may be popular.
I thought about this. Amazon tries to suggest products, and it works once in a while. What confuses Amazon is that my wife purchases books and products which interest her. I use the same account. You can imagine the recommendations that combine math and ancient history books with mysteries written by depressed Scandinavian writers. The personal products recommendations are sometimes downright bizarre.
Microsoft also does the prediction thing with Bing. I wonder if Microsoft used Bing to determine how users of Windows 10 would respond to the “you have no choice” updates that are shoved down the Internet tubes?
For IBM, my reactions went in three directions.
First, IBM marketing and PR professionals are trying to make Lucene, acquired technology, and home brew scripts do something useful. In my own gift buying experience, I keep my eyes open, listen, and then use that information to purchase a personalized gift. For other gifts, I write a check. I am not going to fiddle with a mobile phone app. IBM is obviously aiming at a niche, which for IBM’s sake, finds this approach to buying useful.
Second, IBM Watson is not yet on my radar as a viable solution for search and content processing. IBM may be tallying huge sales, but I don’t hear about them. Even more telling is that my system for monitoring news about search and content processing snags wild and crazy assertions about how wonderful Watson is in curing cancer, making recipes, and, of course, selecting gifts. I find this difficult to believe. The sheer range of applications and capabilities attributed to Watson are difficult for me to believe.
Third, IBM has to find a way to generate substantial organic revenue. Reducing full time equivalents, buying back stock, and losing money for the Buffet machine are not inspiring confidence.
Perhaps IBM Watson will select the perfect gift? Amazon uses its recommendations to generate revenue for Amazon. IBM uses its Watson to generate public relations. Which is the better approach?
Answer: the one which makes money. I do not include the revenue IBM generates for its marketers and PR advisors.
Stephen E Arnold, November 18, 2015
Comments
One Response to “IBM Watson: A Consumer App for Toy and Gift Buyers”
-
- Subscribe to Beyond Search
Feature archive
News archive
-
Encontrará más
IBM Watson: A Consumer App for Toy and Gift Buyers : Stephen E. Arnold @ Beyond Search