7 ms·
Congrats. I'm sure they're aware, but for others here: decide.com does a fantastic job at price predicting products, and is founded by the same person behind t
by mitchellh 14y ago
Congrats.
I'm sure they're aware, but for others here: decide.com does a fantastic job at price predicting products, and is founded by the same person behind the price prediction algorithm purchased by Bing ("Farecast") and he is also a tenured professor at the University of Washington specializing in machine learning. Pretty steep competition both from an engineering and business standpoint.
- joshu 14y agoI actually met Oren at TED. Super sharp guy.
- redorb 14y agoThe difference I see is priceonmics give me a real answer $200 for a iphone 3gs see here: http://priceonomics.com/phones/apple/iphone-3gs/ http://priceonomics.com/phones/apple/iphone-3gs/ While decide.com give me a .99 cent option and no other real value ideas of that phone https://www.decide.com/search/iphone%25203gs/1/22818276/RELEVANCE/ https://www.decide.com/search/iphone%25203gs/1/22818276/RELE...
- kruipen 14y agoFor anything from long tail priceonomics seems to show pretty much random data, e.g. this is the most most popular amplifier: http://priceonomics.com/amplifiers/eagle/ http://priceonomics.com/amplifiers/eagle/ (I tried amplifiers because I was curious to compare data with a toy site (500 lines of haskell for crawler+site) I built for audio equipment price trends: http://kruipen.com http://kruipen.com)
- rjb 14y agoDecide seems fine for new items, specifically electronics and appliances which is probably why the service highlights their price drop alerts. Priceonomics really impressed me when it was able to price a Leica M4P and a set of P90 pickups. Decide had neither in their system.
- zissou 14y agoThe economic (consumer choice) problem that Priceonomics is aimed at solving is the "when to buy a product" problem. Plenty of products already exist to aid consumers with the "where to buy" and "what/which to buy" problems. This paper at Microsoft Research goes through the theory: http://research.microsoft.com/apps/pubs/default.aspx?id=148337 http://research.microsoft.com/apps/pubs/default.aspx?id=1483...