Riya reborn is really cool way to search

OK, so, what’s the 250 machines for?

A better way to find stuff. A way better way.

Just don’t tell Maryam, OK?

Why not?

Well, it’s the first time you can search for “red strappy shoes” and have every type of shoe show up VISUALLY.

This is cool stuff.

It’s called Like.com.

TechCrunch has a report up.

I interviewed Munjal Shah, Riya’s CEO and here’s the interview. Here’s a demo of Like.com. They are both videos published to my ScobleShow video blog.

More shortly.

I hear that Like.com will be turned on shortly. Like.com is up now.

This is a lot of fun.

Some stories about Like.com.

1) The URL cost him $100,000. In the interview he explains how they bought it. It involved finding the guy who owned it, jumping a fence, and leaving a bottle of wine with a note on it (he wouldn’t answer his email).
2) Riya was pretty close to being sold to Google. If it had been, they never would have worked on this search engine. So, by getting turned down by Google Riya came back with a much better business.
3) Just the jewelry set takes 20GB of RAM.
4) Munjal still believes in blogs, but for this launch Riya talked with fashion bloggers, and journalists outside the tech world like at People magazine. Why? Well, this site — in its current incarnation — will be most interesting to women and non-geeks. If you’ve looked at who participates here, it’s heavily male.
5) Why not keep working on face detection? Because they learned through user testing that they’d never be able to make it good enough. They found that by focusing on visual image searches they can get a much more satisfied user base.

What do you think? It is the most interesting search experiences I’ve had since I first laid my eyes on Google about eight years ago.

UPDATE: Liz Gannes of GigaOm has a report too, but isn’t as impressed. “I do think this is a cool idea but it’s not dazzlingly good yet.”

UPDATE 2: Dan Farber has a report too. So does Rafe Needleman over on his new Webware blog.

UPDATE 3: Andre Ribeirinho writes “What I like most about Like is that being a shopping comparison site they are targetting a specific target, Women and pursuing a market worth of $15-$30B.”

Advertisements

58 thoughts on “Riya reborn is really cool way to search

  1. “Why not keep working on face detection? Because they learned through user testing that they’d never be able to make it good enough.”

    We’ve had this discussion in the past, this is an unsolvable problem.

    But the early Riya was in a good niche when it was a nice desktop app for family and friends. The point was that the set of photos made best suited for photo comparison and face detection accuracy.

    Now, this website looks like one of the thousands comparison shops out there.

    Also, drilling by color for instance is nice, but that’s not something Amazon and the likes cannot do overnight.

    Like

  2. “Why not keep working on face detection? Because they learned through user testing that they’d never be able to make it good enough.”

    We’ve had this discussion in the past, this is an unsolvable problem.

    But the early Riya was in a good niche when it was a nice desktop app for family and friends. The point was that the set of photos made best suited for photo comparison and face detection accuracy.

    Now, this website looks like one of the thousands comparison shops out there.

    Also, drilling by color for instance is nice, but that’s not something Amazon and the likes cannot do overnight.

    Like

  3. Stephane: really? If they can do it overnight I’d like to see them do it. I seriously doubt they’ll be able to pull it off.

    Why not? For the same reason that Google, Yahoo, and Microsoft (and everyone else) hasn’t been able to compete with YouTube.

    Doing software is a lot harder than you make it sound. I hear engineers brag the way you do all the time.

    Everytime I hear this kind of horseshit I say “go ahead and build it overnight then.”

    It never happens.

    Why is that?

    Like

  4. Stephane: really? If they can do it overnight I’d like to see them do it. I seriously doubt they’ll be able to pull it off.

    Why not? For the same reason that Google, Yahoo, and Microsoft (and everyone else) hasn’t been able to compete with YouTube.

    Doing software is a lot harder than you make it sound. I hear engineers brag the way you do all the time.

    Everytime I hear this kind of horseshit I say “go ahead and build it overnight then.”

    It never happens.

    Why is that?

    Like

  5. 1. Thanks for commenting on my blog. I’m honored.
    2. I think Like.com is a GREAT thing for visual shopping. Recently I was looking for a watch that looked like the Nixon that Hiro Nakamura wears on Heroes, but I also wanted to be atomic-clock-synchronized and solar powered. I eventually found it through Froogle, but it might have been faster using a slightly more developed version of Like.com.

    Btw, the watch I eventually settled on was a Casio WV300DA-7B.

    Like

  6. 1. Thanks for commenting on my blog. I’m honored.
    2. I think Like.com is a GREAT thing for visual shopping. Recently I was looking for a watch that looked like the Nixon that Hiro Nakamura wears on Heroes, but I also wanted to be atomic-clock-synchronized and solar powered. I eventually found it through Froogle, but it might have been faster using a slightly more developed version of Like.com.

    Btw, the watch I eventually settled on was a Casio WV300DA-7B.

    Like

  7. I just watched your video interview about Like.com.
    1. That watch search was just like the one I did to take the screenshot for my blog.
    2. That guy in the video has the same Firefox Mac theme I do.
    3. That guy’s toolbar has the same programs running that I do.

    So many coincidences…

    Like

  8. I just watched your video interview about Like.com.
    1. That watch search was just like the one I did to take the screenshot for my blog.
    2. That guy in the video has the same Firefox Mac theme I do.
    3. That guy’s toolbar has the same programs running that I do.

    So many coincidences…

    Like

  9. I just read the articles about Like.com. I do not understand the hype, two PhD’s out of London have come up with the largest browsable shoe index in the world several months ago. It is called ChezImelda. The site is very fast and searches for look alike, it has in addition to like.com a very fast interface… lot’s of fun to use

    Like

  10. I just read the articles about Like.com. I do not understand the hype, two PhD’s out of London have come up with the largest browsable shoe index in the world several months ago. It is called ChezImelda. The site is very fast and searches for look alike, it has in addition to like.com a very fast interface… lot’s of fun to use

    Like

  11. Before you jump on your high horses, let me correct myself. Amazon and any online merchant can build it overnight by looking up the color in the __textual__ metadata. In other words, it’s a metadata quality problem, not a visual problem.

    What Riya is trying to build is a new form of metadata that surpasses the textual metadata. It unfortunately remains to be proved that such metadata and associated retrieval system is an order of magnitude better, from a quality perspective and from a user experience perspective. And that it’s worth paying 250 servers in the long run…

    In fact, what Riya is doing first apparently, correct me if I am wrong, is real-time search.

    Like

  12. So many Web 2.0 sites launch with no focus on monetization. Like.com is a refreshing change.

    The only way to build this technology overnight will be to buy the company. It would certainly add a totally new aspect to Froogle.

    How much of their IP is protected?

    Like

  13. Before you jump on your high horses, let me correct myself. Amazon and any online merchant can build it overnight by looking up the color in the __textual__ metadata. In other words, it’s a metadata quality problem, not a visual problem.

    What Riya is trying to build is a new form of metadata that surpasses the textual metadata. It unfortunately remains to be proved that such metadata and associated retrieval system is an order of magnitude better, from a quality perspective and from a user experience perspective. And that it’s worth paying 250 servers in the long run…

    In fact, what Riya is doing first apparently, correct me if I am wrong, is real-time search.

    Like

  14. So many Web 2.0 sites launch with no focus on monetization. Like.com is a refreshing change.

    The only way to build this technology overnight will be to buy the company. It would certainly add a totally new aspect to Froogle.

    How much of their IP is protected?

    Like

  15. Great concept in a hugely exciting space. They are not alone though, London based startup Pixsta will be their competition it seems – they launched their visual shoe store http://www.chezimelda.com href=www.chezimelda.com>Chez Imelda in beta and are racing up the Alexa Rankings

    Like

  16. Great concept in a hugely exciting space. They are not alone though, London based startup Pixsta will be their competition it seems – they launched their visual shoe store http://www.chezimelda.com href=www.chezimelda.com>Chez Imelda in beta and are racing up the Alexa Rankings

    Like

  17. A year ago I was skeptical of Riya based on my past experience at AltaVista.

    Now I think Riya has a much better chance of success focusing on the fashion market because “likeness” is OK, exact matches are not necessary. Secondly, the Cost Per Click (CPC) business model will be much more effective than consumer based advertising. That addresses the two concerns I had initially.

    Sometimes technology people get all wound up about the technology. It is not about the technology…it is about the problem it solves. I wrote a blog on this today. http://dondodge.typepad.com/the_next_big_thing/2006/11/riya_tries_agai.html

    Like

  18. A year ago I was skeptical of Riya based on my past experience at AltaVista.

    Now I think Riya has a much better chance of success focusing on the fashion market because “likeness” is OK, exact matches are not necessary. Secondly, the Cost Per Click (CPC) business model will be much more effective than consumer based advertising. That addresses the two concerns I had initially.

    Sometimes technology people get all wound up about the technology. It is not about the technology…it is about the problem it solves. I wrote a blog on this today. http://dondodge.typepad.com/the_next_big_thing/2006/11/riya_tries_agai.html

    Like

  19. All they need now is a deal with TIVO and maybe Amazon, to make possible pointing out something in a show and have a ‘Buy this’ button come up automatically.

    TV industry people are looking at doing it manually, but with Riya technology, it might be more cost effective. TWIM 26 talked about it.

    For myself, I liked the face recognition idea better. It feels a bit strange to be actively dropped as a target audience. Usually, I just never make it into one.

    Like

  20. All they need now is a deal with TIVO and maybe Amazon, to make possible pointing out something in a show and have a ‘Buy this’ button come up automatically.

    TV industry people are looking at doing it manually, but with Riya technology, it might be more cost effective. TWIM 26 talked about it.

    For myself, I liked the face recognition idea better. It feels a bit strange to be actively dropped as a target audience. Usually, I just never make it into one.

    Like

  21. Pingback: Global Nerdy
  22. Robert:

    I’m curious how two people I consider pretty smart about these things — yourself and Greg Linden of Findory/Amazon (Geeking with Greg) — can have such differing opinions on Riya. You seem to think it’s great. Greg thinks its a farce.

    My guess is the truth is somewhere in between.

    Like

  23. Robert:

    I’m curious how two people I consider pretty smart about these things — yourself and Greg Linden of Findory/Amazon (Geeking with Greg) — can have such differing opinions on Riya. You seem to think it’s great. Greg thinks its a farce.

    My guess is the truth is somewhere in between.

    Like

  24. Matt: I just read Greg’s blog post over here: http://glinden.blogspot.com/2006/11/riya-vaporware-and-hard-problems.html

    I don’t see where he disagrees with me.

    By the way, Munjal told me they did user testing with HAND PICKED FACES and that the user testing only gave it a 3.2 out of 5 rating.

    It wasn’t that the algorithm isn’t good enough, it’s that there was absolutely no way to meet the human expectations that were being placed on it.

    Like

  25. Matt: I just read Greg’s blog post over here: http://glinden.blogspot.com/2006/11/riya-vaporware-and-hard-problems.html

    I don’t see where he disagrees with me.

    By the way, Munjal told me they did user testing with HAND PICKED FACES and that the user testing only gave it a 3.2 out of 5 rating.

    It wasn’t that the algorithm isn’t good enough, it’s that there was absolutely no way to meet the human expectations that were being placed on it.

    Like

Comments are closed.