Fashion Marketing Group (FMG)
Google Shutters Fashion Platform ‘Boutiques.com’ + Interview with CEO Munjal Shah

Interview and reporting @Kristayloryoung

Editing/Commentary @JedWexler

After a much ballyhooed launch last November, Google has quietly shuttered its online fashion platform Boutiques.com.  In the wake of producing high-profile launch events, exhaustive designer collaborations, press junkets, and numerous celebrity & blogger-curated shopping ‘lists, ’ the mega search engine is now folding the service into its Google Product Search (and to some degree, Google Catalogs). Here is the new Boutiques.com redirect page;

Email (excerpt) received from Google on September 23, 2011;

“We wanted to let you know about the latest plans we have for apparel shopping at Google. We launched Boutiques.com last year, and it helped us bring a new set of technology to improve the apparel shopping experience online. We learned a lot about how people like to shop from Boutiques and today are incorporating the best features and technologies from Boutiques.com and related properties into Google Product Search.”

Style/tech site,  Signature9, was one of the first to report that something was afoot on June 13, 2011 with an article titled, “Did Google give up on Fashion?”

While it’s no surprise that Google thrives on acquiring companies  (57 so far in 2011) for the sole purpose of gobbling up technology, talent and data, the pace of this in-house wind-down may raise hackles within the fashion/digital/media cognoscenti – especially considering the resources and relationships Google applied towards leveraging key players to get the project off the ground.

Shortly after the original launch FMG’s own Kristin Young caught up with (now former) Boutiques.com CEO Munjal Shah – an intriguing read in light of recent events.

INTERVIEW WITH BOUTIQUES.COM CEO MUNJAL SHAH – BY KRISTIN YOUNG 

Google paid $100 million for Like.com’s visual search technology that uses images to drive searches rather than text. The Silicon Valley behemoth has been testing the technology through its fashion shopping site Boutiques.com since last November, when the site launched. It has the ability to learn an individual’s style through algorithms prompting some to call it Netflix for fashionistas. But is Boutiques.com making a seamless transition into the world of fashion retail? We sat down with Munjal Shah, Like.com’s co-founder and CEO who is now the director of product management at Boutiques.com, and asked him more about his approach to shopping fashion on the Web.


 Boutiques.com has often been referred to as the blending of “Geeks and chic.” Who’s behind the company?  

 Boutiques.com represents a pretty unique set of people. On the one side, there are a whole bunch of PhDs and computer science people specializing in computer vision—which is face recognition and optic recognition—out of places like Stanford. The other half of our team went to fashion school or were designers or stylists in the industry. So we are tech nerds and fashion nerds and it’s the combination of those two that we used to build this product. 

 You originally developed the technology for Like.com. How did you end up applying it to fashion?

Fashion is actually the area of e-commerce that has gotten the least attention. If you think of the e-commerce experience that people have been building on the Web for the last 10 years, it’s mainly selling books and electronics. If you think about the way that you bank today, it has completely changed. If you look at the way we shop today, there have been some changes but it had not been fundamentally transformed. The fashion industry hasn’t known how to translate their business online. Their experience is not about creating Websites. It’s about creating some kind of experience, a fashion show,  a line of clothing, a dream or a lifestyle. Our vision is to make the experience better. There is so much opportunity to improve it. 

 How is Boutiques.com different from other online shopping aggregators like Shopstyle or Polyvore? 

Shopstyle.com was the best generation-one high-end soft goods shopping search experience. But it was just that, it was generation one. It did not build in any personalization.  

Polyvore.com provides you good context but it’s context without the content. I think a lot of people go there and they use it as a design experience and as a fun tool to express your creativity. But that doesn’t mean you go there to shop.

The visual search technology is what commanded the $100 million price-tag. Why is visual more attractive than text searches? 

You can try to describe a carpet, for instance, you can draw a picture and use as many words as you want. But you can’t fully communicate what that carpet is like. Linguistics are just inherently imprecise for describing aesthetic things. It’s so much easier to just point and say, show me something like that. Our brains have been reading for only 2,000 years. It’s a relatively new thing for humans. We’ve been seeing things for 100,000 million years. In visual search, “view all” is our number one feature. You can see everything in one second.

How can you “teach” the site your style?

The first piece of technology we needed was the ability to classify clothing including genre classification. So we built an algorithm using literally 100 fashion stylists who organized and tagged each one of 50,000 items into genres. And we had it done twice because one person’s “edgy” is another person’s “classic.” We used that set of information to train an algorithm.

 So if you’re style is “street” and you’re searching for a black dress—and if you’re wearing it out clubbing—you’ll get options that are tighter with a higher hem. Whereas, if you’re “classic,” you’ll get looser fitting pieces.

 We also added fashion rules, such as heavily-patterned dresses don’t go with heavily- patterned bags.

What was the inspiration for added features like “visually similar items” or the “love” and “hate” buttons?

We studied salespeople in the Macy’s. We wanted to give you the inspiration and context that you need. When I hover over an item, I can find you other dresses that are similar, gives you a little bit of range. So if you don’t like everything about it, don’t like the buckle, I can give you other options real time. Half of wearing clothes is what you wear, the other half is how you wear it.

It’s about discovery not just search. How many items in your bag did you know you wanted to buy before bumped into items? How do we create that product to bump into not just spear fishing?

What’s next on the horizon for Boutiques.com

Men’s is coming but it will be a different format. Men don’t like to shop. Men don’t like to be seen as primping and fussing over what they wear. And there are no plus sizes on the site. We will add that.   ***FMG

Editor’s note;

****Google is obviously well aware that Fashion is the #1 revenue-producing E-commerce category online (Forrester Research)  – so we’re certainly intrigued with what they really do have on the horizon.

.