Archive by Author | Semil Shah

Quora and the Quest for Long-Tail Search

One of the Valley’s favorite parlor games over the last year has been to debate what the future holds for Quora, and specifically, to speculate who will invest in the company moving forward. All rumors aside, and acknowledging my strong bias in favor of the product and company, I firmly believe that Quora is one of the most important, impressive, and disruptive social software companies that exists in the market and, therefore, presents a rare opportunity for venture capital and a potentially outsized return.

Before I lay out my arguments as to why I’m so bullish, I have to note and address the most common refrains lodged against the company and my responses. “They have no revenue!” That doesn’t matter yet. “I don’t use it anymore.” Try not to extrapolate from your own behavior. “I don’t like the boards.” The boards are just a tacti to refer traffic, and the only issue is that board-activity ends up in your newsfeed. I agree Quora should fix that, but it’s a small problem. “The newsfeed is noisy.” Well, stop following one thousand people or topics, or just search on the site. “It’s not growing.” Quora isn’t totally focused on growth yet, but they are laying the foundation for it, especially with one specific hire. They are just finishing their product. It was a long beta. “They haven’t nailed the social aspect.” I’d say this is actually a valid criticism — more on this below.

People know I’m bullish on this company, so I’ll share how I articulate my rationale when I speak with others: The market for search, if graphed against the size of a keyword search, roughly follows a power law distribution. Some refer to the shape under the curve as a “dragon,” with a head+body and a tail. In the “head+body” of the dragon, you have traditional search — the short keyword strings we enter into Google, Bing, Yahoo, and other major engines that have provided these companies with business models that exert incredible leverage. There is a big area under the curve for the head+body, and there’s stiff competition among those larger companies to manage fixed costs and maintain market share.

Now, when we move our attention over to the tail of the dragon — what people refer to as the “long tail” — we see it what on the surface appears to be a very thin line and small area under the curve. However, in power law distributions, the area under the curve in the tail is actually equivalent to the area in the head+body, which means the size of the market is theoretically similar, though there’s a reason Google hasn’t gone after long tail search and why Facebook hasn’t gone after questions.

The long tail is a hard nut to crack because (1) it’s hard to police and (2) it’s hard to build a business model around. Yet, we must remember that every question pumped into Quora is, in fact, a long tail search. By encouraging users to write out grammatically-correct questions, Quora has trained its users to revert back to searching by asking questions (as humans do in real life) rather than crude keyword arrangements.

The people that contribute to and/or follow a Quora thread make up the market for that long-tail search query. And, in order to address the root problems in long-tail search (policing and revenue), Quora has deeply integrated two features in response: PeopleRank and Credits. PeopleRank is a generic term to capture all the algorithms and administrative controls that provide a policing mechanism for the site. People monitor the site, organize topics, watch for bad actors, and surface the best or most relevant content. Quora’s credit systems creates an economy for all actions taken within the site to reward specific behaviors; eventually, one can picture the credits coming together to help monetize some of the long-tail searches that occur within the site.

Details aside, Quora also fits into a larger, evolving trend. The world is moving away from centralized systems and embracing the crowd. Whether or not you believe the crowd is wise, or rather if you feel that the masses are asses, there’s no denying this shift in many area. Look no further than crowdfunding, for instance, a phenomena encapsulated by the rise of Kickstarter and the recent JOBS Act legislation. At its core, Quora is a crowd-sourced platform, and if they’re able to enforce their policing structure and build a real business model around credits and other features, they have a chance to generate, distribute, and deliver excellent long-tail content to other places such as media brands, educational institutions, and any other entity that’s willing to pay for specific content.

I will admit, despite my overall bullishness, that Quora has made mistakes. Every startup does. Even though I use the boards, they were poorly introduced, managed, and designed. Credits are smart, but they weren’t explained well to users. Many new users who come to the site aren’t welcomed overall, nor do they know what to do. There are some legitimate concerns that the site won’t be as “social” experience as others. They haven’t built an API yet. They haven’t tested lightweight text ads against certain threads yet, though one could make the case it’s too early to try. They haven’t seeded content by going after key knowledge communities in other verticals, though they have seen nice organic growth of content in the major content areas of politics, finance/economics, sports, and entertainment (including movies).

It is easy to list out these missteps partly because there are such expectations attached to this particular site and service. While it’s important to note these missteps, they are — to me — trivial when considering the overall picture of what this company has already achieved. The site and its contents emit a feeling of authority and fidelity. It has a “permanence” about it. When I go to the site, I have the sense that the content on the site will be there, forever — some of it will evolve when the crowd feels compelled to update it, and some of it will just live in on for others to benefit from. It is a site that doesn’t look like any of the others out there — it is an outlier, what on the surface looks like a Q&A site but is much more, a slick engine driving a complex economy of information. In terms of investment, my belief is that Quora stretched their Series A investment to the limit and accomplished a tremendous feat. Whatever the market will bear for their Series B will have been hard-earned, and in my opinion, worth every penny for the outside shot at owning a piece of what could be one of the most dynamic human knowledge systems in the world.

P.S. One morsel to chew on: Check out this pie chart of the top 10,000 “Wikipedians” broken down by the amount of “work” they do. Some of these people do an incredible amount of work, entirely free and anonymous. This presents two insights: (1) Quora has built a judgement layer on top of facts (Wikipedia) where contributors may actually be known and earn reputation; and (2) a very small group of people working in concert together can potentially create massive economic value — Wikipedia is a Top 10 site.

Photographs > Video ?

I’ve been trying to think about all the interest around mobile video concepts today. From what I can tell, companies like Klip, Viddy, Vidly, and Socialcam (plus many others) all allow mobile users to capture and share video from their phones. With all the excitement around the phone camera (encapsulated by Instagram), as well as the quick scale that social distribution platforms can provide, there seems to be another rush to see if someone can pull off video. Whereas photographs seemed obvious to many, I’d have to admit I’m a bit skeptical about video. One, I’m not sure if it’s a core behavior as snapping pictures is. I wonder what percentage of media files on a user’s phone is picture vs. video I posed this question informally on Twitter today, and most answers came back from zero to around 10% (either music concerts or for their family’s children). Two, the current interfaces seem to deemphasize “capturing video” as a feature within the camera interface. And, three, even if core behaviors change and/or the software around the video camera evolve, there are technical hurdles that would need to be overcome in order for the average phone user to capture and share a video that would meet the requirements the audience’s time to consume that media. The bar is just higher for video. I’m not saying it cannot be done (nor that it’s not worthwhile), but my gut is telling me there are still more opportunities around photographs themselves. More on that later this week…

Hip Hip, Jorge

Today was opening day at Yankee Stadium. Long-time catcher Jorge Posada threw out the first pitch today, but this time in civilian clothes. In January 2012, he announced that he was retiring from baseball. After 17 years in pinstripes, with one team, through multiple playoffs, championships, and near misses, Posada was a constant fixture in one of the franchise’s most storied dynasties.

Posada started with the Yankees just as I stopped playing baseball. I grew around Yankee Stadium, so growing up it was hallowed ground. Baseball was the first sport I really got into, and I was obsessed, so much so that even my father is now a diehard Yankee fan, so much so that we will start our phone conversations about the most recent Yankee news. Like Posada, I was mainly a catcher, the only position on the field that is both involved in every play and is able to see the entire field. Because catchers manage pitchers and move position players, many of them turn into coaches. I ended up coaching little league <smiley face>; I think Jorge will be a major league coach one day, because he has a brilliant mind behind the plate.

Jorge gave me and my father incredible memories over his solid, 17-year career. I know there are Yankee haters out there, but no one can deny that Posada was one of the best offensive catchers in history and lights out in the playoffs.

Most important, however, is that Jorge was “the man.” He hid behind clunky catcher’s gear, often referred to the “tools of ignorance,” crouching behind home plate and managing some of the best pitchers in the country. He caught and applied the tag after the “flip.” He didn’t wear batting gloves. He didn’t get caught up in the media circus. He was supportive of his teammates and coaches as if they were his family. And, away from the diamond, he was the family man. His son, for instance, was born with a severe cranial condition which required 7-8 surgeries to repair. From this, he used his wealth and celebrity to create a foundation solely devoted to finding other children with this rare disease and providing support for their parents.

I can’t find the interview, but I remember seeing a clip with then Yankee head coach Joe Torre recalling one of the Posada family’s charity events, which they hold each year in NYC. Torre said that, to the shock of the crowd, at the end of dinner, there was a surprise keynote speech. It was Posada’s son, Jorge JR, who had finally finished his last massive facial surgery. He stepped up to the podium, all of 8 years old, dressed up in a tux, and addressed the crowd, thanked them for their support, and declared that he was excited to fight onward. Torre remarked the entire crowd fell silent, sitting on the edge of their seats, people wiping their eyes, and when Jorge JR was done, the room erupted into a standing-ovation.

I don’t care if Posada has awesome stats, or if he ends up in the Hall of Fame. It doesn’t matter. He was one of the classiest, most durable baseball players of his generation, a throwback of sorts, and who used whatever good fame and fortune for good means. So, thank you, Jorge. Hope to see you in Pinstripes again.

The Quietly Persistent Entrepreneur

Alain Mayer doesn’t look like your typical Valley entrepreneur in today’s celebrity parlance. He’s a not a college dropout. He’s from Europe! Switzerland, to be exact. He came to the United States, grabbed a Masters in Computer Science from Brown and then his PhD in Computer Science from Columbia University.

Since then, he’s been a quiet but persistent startup junkie working in highly technical capacities. He started at SMArts, which was acquired by EMC, Bell Labs, became the CTO at CenterRun (now part of Sun/Oracle) and CTO of Red Seal Systems. After Red Seal, he wanted to do his own thing, and he’s long been fascinated with technologies that can help discover and summarize news — not unlike what Google News tries to do. He went back to Columbia and found some old professors who were working on NLP technology, and licensed it from them.

And, thus, his startup — Dygest — was born. Mayer teamed with Alan Caudill, a former colleague from his startup days. Dygest started very small with precise entry points. They used the NLP technology to create four native iPhone applications focused on specific news verticals: Apple, Technology, Celebrity, and Gaming. All did quite well, and as you may imagine, the Apple one did quite well. My wife used the Celebrity Dygest app to keep up to date on all the Hollywood gossip.

The cool thing about their technology is that they could pick a topic, seed it with 50-100 hand-picked blogs around the topic, and then the system would surface the Top 10 or Most Popular articles at any given time, synthesizing different blogs on the same topic and distilling a news story out of existing news stories. We all know this problem in tech blog circles. There are 25+ blogs that announce the same bit of news; here, Dygest went to all those and created one article about the event. The unfortunate thing about this is that the apps were federated, and while communities of users were being built. Dygest didn’t know much about them and couldn’t find the right engagement layer.

And herein lies the rub for Dygest. They did everything right, but hit a wall. Over the course of half a year, they made some key decisions. They began to create one unified app with roughly 10 news channels, which included Apple, Technology, and the other legacy apps. They decided to use a right-hand menu bar, which visually offended me but Alain pretty much ignored my advice! They added politics, too — so he was listening, sometimes. And, they had to get more social. So they rebuilt the system to have more identity around users.

Ultimately, the key decision in the new build was to figure out what to have the readers do once they’ve read a news story. The answer is to “play the news.” Just like we may want to bet on March Madness games or trade stock in Hollywood celebrities or Intrade political futures markets, Dygest decided to let its users to interact with the news around polling questions that could hopefully provide insights into sentiment, forecasts, and opinion. Yes, you can call it a game mechanic, but in this case, it really is, and the crowd so far loves it. Read a summarized article to catch up on a topic, get prompted with a short question, the option to say “yes” or “no,” and then see how the crowd reacts around that yes/no question.

With Dygest powering the underlying technology in the app, the company needed to turn toward consumers. This led to a series of quick changes that Alain and Alan executed quite precisely and quickly:

  • The unified app became “Scoople,” a play on news “scoops.”
  • I was able to convince Sahil Lavingia to help Dygest redo its iOS logo, and of course, he delivered [see here]
  • Then, the guys closed a seed round of $750,000, which they really had to grind out.
  • Without much effort (and some kinks need to be ironed out), Scoople quickly amassed a footprint of 100,000+ iPhone users, and I’m working with the guys to pour gasoline on that.
  • Then, Scoople was covered by TechCrunch.
  • You can follow a Twitter feed of their mobile poll results here.
  • And, now, they have a shot to build a bit more viral loops into the product, clarify their brand and messaging, and hopefully step on the gas to get more users.

So, let’s see what happens with Scoople. I don’t know if Alain is sitting on a gold mine, but I do believe he’s on to something. Scoople is able to capture statistics of a certain set of mobile phone users, and the amounts he’s collecting are very close to meeting tests of statistical significance. He talks about it as a “Nielsen for mobile.” News is obviously shared with lightening speed on social networks, so Scoople’s challenge is to harness a small slice of that massive sharing traffic and pull some opinion data out of the users. And, I think they can do it.

I’m really rooting for Alain more than the technology or the app. He’s really been grinding it out for a long time. He’s put a ton of thought and elbow grease into the work, like many of you. It’s really hard to have a breakout app. There’s Instagram, there are apps that have large web audiences, and then there’s everyone else. But the initial signs are promising. You should give it a whirl. You can download the iOS app here.

The Hidden iPhone Battery Issue

Over the past month, innovation at the application layer that makes more use of the iPhone’s location sensors has prompted a wider discussion about battery performance, battery extension packs, and what users will be willing to give up in exchange for battery drain. Just like the carrier’s business models that charge for voice and limit data hamper innovation and adoption as iOS continues to grow, the issue of battery strength and fidelity are of great importance to users. When iOS5 became available, nearly everyone complained about immediate battery degradation, the result of default settings that instructed the phone to constantly pull and push information to the notification bar.

But, there’s one basic issue related to iPhone batteries and AT&T specifically that people often overlook. I usually keep my iPhone on the old Edge network for telephone calls, because those calls never drop. AT&T’s 3G is useless for a mobile device. However, in the last week, I can no longer connect to Edge — I’m only connecting to 3G, and it’s not consistent. Friends have informed me that AT&T is phasing out Edge in order to free up spectrum for 3G (and maybe 4G, whatever), but the result for me is that my phone is always TRYING to connect to 3G, and that connection has little fidelity. In reality, the phone is almost constantly trying to connect to the towers, and that takes battery power, perhaps more power than a service like Highlight draws. I don’t know the specifics around battery draw between persistant GPS vs trying to find a cell tower, but if my brief experience provides any indicator, AT&T’s choices also deserve a good part of the blame.

A Real Timehop: Votizen

A simple new service that I enjoy is timehop, which emails you little remembrances about your life it pulls from your social networks. It’s one of those ideas that’s so simple but provides a good deal of delight. Today, a year ago, I was with Sara and close friends up in wine country, and timehop sent me a great picture of her at the Artesa Winery.

But in the back of my head, I knew I had also done something interesting that day as well. As I was going back through some of my  first posts on TechCrunch, I wrote this piece about what I felt was going to be the next trend in social. At the time, we had just seen the beginnings of the Arab Spring spread through the Middle East and North Africa, and it occurred to me that we may see something analogous happen here in the U.S., of course in different ways. That proved largely to be correct. Products like Facebook became even more important for U.S. politics and organizing, Twitter has become the de facto commenting system during debates and primary results nights, and newer companies like Square, Klout, and Quora are working with campaigns and the establishment in D.C. to increase engagement on both sides.

In the piece, I mentioned a small company that I didn’t know much about: Votizen. Here’s a brief excerpt:

All of this activity got me thinking about what will be the next phase in the social networking revolution, what will reach mass consumer scale, be global, and generate real social and financial impact. There’s perhaps no greater market to disrupt. The fast-moving nature of politics today, whether in “mature” markets such as America or “new markets” such as in Egypt, have paved the way for individuals to express themselves and their interests in a political context. Governments and elected officials may ultimately have no choice but to monitor and cater to these activities. This could be the start of the next mass consumer trend, political expression and organization via social networks directly to elected government officials.

And, now, a year later, I work at Votizen. And, it’s changing real fast. If you haven’t yet, sign up for Votizen here and check it out. Tell me what you like and what you’d like to see. Both the good and what needs to improve. Check out how many voters you’re connected to, see which friends are on each side of the aisle. Endorse the candidate that you want to be President. It’s been interesting to see Votizen evolve with the political cycles, from legislation to election, and now, a year later, things are moving really fast. Timehop, indeed.

Tomorrow is a Big Day for Sahil (@shl)

In the summer of 2011, I became interested in a consumer website. My friends helped me get in touch with the company and I spent some time with the CEO to determine if there was an interest in working together. While it turned out to be too early for a nontechnical role, I got to know some folks at the company — and one in particular was outstanding: Sahil Lavingia.

You may have heard of Sahil already. He’s @shl on Twitter. He’s built a few products before that now live on their own. He designs iOS apps in his spare time. He blogs and shares his opinions. He built the iPhone app, and was part of the founding team of Pinterest. He was thinking of leaving a few months after we met. He wanted to do his own thing. I wondered why he didn’t want to stay at Pinterest, which was obviously taking off, was building a great brand, and was going to give him an awesome platform.

Nope, he wanted to do his own thing. After dropping out of college, moving up to the Bay Area, immersing himself in the community, and helping architect and design today’s fastest-growing social application, he wanted to turn one of his many projects into a company — and on his own. He asked for some help around meeting investors. I was very happy to do this. I sent a few intro emails that were really short — “hey, you gotta meet this kid, kick ass front end engineer and designer with a lot of ambition, ready to go out on his own.” My emails were returned within hours, and in a few weeks, he had his pick of some of the best early-stage and VC investors in the area.

In my short time here, I’ve helped a few people through this process, but no process so far has been as dynamic as Sahil’s. He always asked the right questions, checked in about ambiguous situations, and always followed up in the most polite ways. After he finalized his funding — which will be announced tomorrow — he asked me to advise his new company, Gumroad — to which I said, “Hell yes.”

I’ll be honest, I have no idea what will happen to Gumroad. It may take off and be one of the coolest innovations around links, or it may struggle to find its way. What I do know is that Sahil is a rare breed of creative talent — a maker — someone who just wants to build, and build, and build, and eventually one day, perhaps quite soon, he’ll again help build something that millions and millions of people worldwide simply want to use. So, congrats to Sahil for tomorrow’s announcement — and good luck!


Get every new post delivered to your Inbox.

Join 50 other followers