Qwiki Founder: How I Launched A Talking Search Engine -- And Why You Should Avoid VCs
Qwiki's on-stage demonstration at September's contest between 27 rising startups not only earned the company a $50,000 cash prize, but it has quickly become a YouTube sensation.
The startup's first product is a multimedia answer to Google's flagship text-based search tool. Users type in a search term and Qwiki pulls information from the Web and narrates the results in a computerized voice while displaying related media.
Last week, Qwiki's CEO and co-founder, Doug Imbruce, came by The Huffington Post to show us how Qwiki works. Imbruce, 27, also explained how he convinced the "father of the internet search" to join on as co-founder and CTO, and why entrepreneurs should "avoid venture capital firms at all costs."
Huffington Post: How does Qwiki turn information into an experience?
Doug Imbruce: Qwiki's core technology evaluates text, enriches it, and turns it into a interactive multimedia experience that's compatible with any device. The platform is able to take structured data -- text, pictures and media-- as input and literally create, on the fly, a narrative.
Watch Doug demonstrate the tool here (Video reporting by Ben Craw):
DI: Qwiki as a technology has been in development for two years. But my partner, Louis Monier, has been thinking about these themes for probably a decade. I developed the initial prototype, and then we showed that around -- which is how we got Louis interested.
In fact, when I first met Louis, he was supposed to come on as just an adviser. Bobby Yazdiani, who is our seed investor, introduced me to Louis. He said: Doug, this guy is going to laugh at your product, but see what he says anyway. Louis and I wound up talking for eight hours. It was a pretty incredible experience, as a young entrepreneur, being put in the same room with the father of the internet search, and then emerging eight hours later with a business partner.
HP: How are you integrating with other platforms like Facebook, LinkedIn and Yelp? And how is Qwiki going to help small businesses?
DI: So Qwiki's first product is a reference tool; the second one will allow you to merge your Facebook and LinkedIn profiles and create a Qwiki describing you; and the third one will allow small businesses to aggregate all the data that exists about their business on the web, whether it's on Yelp or Citisearch, and also augment their photos.
HP: How much control are small businesses going to have over the Qwikis about them?
DI: We're still deciding and we're going to have to prototype it and figure it out. But we really do want Qwiki, from a consumer perspective, to have an unbiased view. So if a restaurant has a certain Yelp star rating, we want to represent it. But we also do want to give individuals and small businesses the opportunity to promote themselves, because this is a great medium for promotion. So we have to stay somewhere in the middle, and figure out what that means.
HP: So when it looked like Qwiki was going to be a big hit, did you just pack up your bags and bounce to Silicon Valley?
DI: Well, I was in New york, and I was a somewhat credible entrepreneur. I had built an educational content network called the U, which still helps about a million kids a year determine their choices for higher education. And that lead to me meeting all the players in New York.
You know it's funny, as much as New York is well-versed in creating new media businesses and new forms of content, everyone told me i was nuts; that I should go back and be CEO of the U and build that revenue.
So I was disappointed with the way the New York community received Qwiki. Fortunately, I met Bobby [Yazdiani] after spending most of my own money trying to create the technology. He said: There's real science that you need here. You need a real scientist. Come to Silicon Valley or else you're not going to be able to find him/her, because the talent just isn't in New York. Or, if it is in New York, people are working for banks or they're busy with their own startups.
So we did just that. I got rid of my apartment, and moved out to Silicon Valley. At first, I lived in a retirement community in a one bedroom apartment that was especially depressing compared to the Hudson River views I had in New York. So I just set up shop, started recruiting and all the other chips fell into place.
HP: Is Qwiki going to be free for users?
DI: Yes, we will never charge users to experience Qwiki. However, there might be offerings where we allow small businesses or the like to augment or promote themselves in different ways.
HP: So how are you going to monetize Qwiki?
DI: Ultimately, we think there's a really compelling opportunity to develop a world-class rich media advertising network. This is based upon the fact that unlike traditional video, Qwikis -- because the way they're constructed -- can target the corpus of all the information. So users can know the exact text. In traditional video, you don't know the text unless somebody transcribes it. [With Qwikis], if an address is mentioned, you know what the exact address is. So there's a lot of exciting data that we can use to take rich media and target it in ways it's really never been done before. Also, the platform itself can be used to generate personalized advertisements on the fly. There are lots of ideas we're exploring.
HP: Do you have any advice for entrepreneurs looking to find financing for their startup?
DI: Try to avoid VCs at all costs possible. [Laughs.]
DI: Well, you know, there are a lot of VCs that can add incredible strategic value. But you're better off, both from a personal financial perspective and from a strategic perspective, I think, finding your product market-fit before you take institutional dollars. In our case, we've had institutional expertise provided by Bobby Yazdani, who is our seed investor and was an early investor in Sales Force and Google. Bobby has provided that role of company mentor and strategic thinker.
So I would suggest that more so than finding a [VC] firm, and feeling like you have to have a brand like a Kleiner or Sequoia invested in your business to actually create something that sets the world on fire, focus on finding an individual and incenting them to be a part of your management team as much as they can. At the same time, figures like Bobby and others don't have time to do the day-to-day. But you do want to spend a couple hours a week, at least, with a real mentor both for you as an individual and for the company. I think institutional funding at early stages is probably not that productive.
HP: So, we noticed the alarm clock function on the TechCrunch show. As I'm sure you know, it's literally straight out of a scene from the movie Total Recall. Is that currently available to users?
DI: The alarm clock function is probably about six months away. The reason we're able to provide that experience -- a personalized narrative that merges all your personal and social data into a morning greeting customized for you -- is because we've invested so much so far in our core technology and our scale. So we're able to handle all that audio -- the millions of users who generate tens and millions of sentences per hour -- because we have the infrastructure, and that's really what a lot consumers don't see. But that's where a lot of the early dollars and time were invested. And now we can start making user flow and pacing progress.
HP: Imagine, for a moment, that Qwiki goes global, and you start to consider offering your services in China. The Chinese government agrees to give users access to Qwiki on the condition that the government can censor searches on 'Democracy' or 'Tiananmen Square.' As a company in the business of making information universally accessible, how will you deal with that? Will you enter the Chinese market or will you say no to any kind of censorship and forgo offering Qwiki to the hundreds of millions of internet users in China?
DI: Look, certainly we have a set of ethics that we believe in very strongly. When you localize any service, there are challenges that you have to meter, and you have to understand local customs and local regulations.
I don't have the answer to that question today. But it is going to be a matter of balancing our company mission with the fact that you do have to sometimes change your strategy or change your mode of operating to develop localized versions of your product. I think Google set a very strong example, and we really respect what they've done and it's an example that we'd like to follow. But, ultimately, we want to provide people with access to information in the most unbiased, practical, and entertaining way possible.
HP: How is Qwiki advertising?
DI: Really it's been word of mouth so far. We have about 125,000 users waiting to access the Alpha right now. And we allow current users to invite new users. And the reason we're in Alpha right now ... [is] because we get so much feedback on a daily basis ... that we literally don't have the staff to read through 10,000 comments a day ... So we're trying to figure out how many people to expose the product to along with how to get feedback in a way we can actually act on it.
HP: How many users do you expect to have mid-year 2011?
DI: We haven't made specific projections. I think about this company 10, 20, 30 years out. And if we're not the ubiquitous information experience, and people are not waking up to Qwiki, watching Qwikis on their new colleagues and new friends, using Qwiki to help make dining decisions, than I haven't done my job and the team hasn't done their job. So our goal is absolutely nothing less than ubiquity. I don't know how long that will take. And it's certainly a lofty goal. But, you know, you've gotta aim high.
HP: Did you really go on a date with Natalie Portman, as you alluded to in your Tech Crunch presentation?
[Laughs.] No, just a little suspension of disbelief.