Okay terrific. So now we have a real treat For everyone. Paul Gram who was here yesterday. He started off with an interview of Charlie Rose is going to conduct his famous-- office hours with Michael, and Michael's going to explain how this is going to work. Paul is Why comment, one of the famous things, the office hours that all the partners, including Paul, conduct.
I love how you Kill time anymore. No you're awesome, because you're like we have no idea where Mike is so I'm going to talk until somebody finds Mike. But this time I was actually very close. Will you join us on stage. Thank you. So you guys You guys all know Paul Graham. Please have a seat Paul.
Is this mic, oh good mics on. Where should I sit? Where would like to sit? On the motorcycle? No. Are they going to stand We're gonna work this out but I think that's the backup but we're gonna try to mike them. In that case, maybe here. You sit there, maybe there's more one of them, I'll sit here to be on the safe side.
You're making me nervous, here. OK, do you want some water?
Nah, it's alright.
So, what we're doing here is office hours. Can you just tell the audience what office hours are Y Combinator?
OK, office hours are where the YC partners talk to the startups about what they're working on So were going to try and simulate that. But, it should be a pretty realistic simulation because these start ups are much like the ones we invest in. Except you don't do office hours in front of thousands of people.
Right. So the unusual thing is these guys' secrets are all going to be discussed in front of a whole room full of people, so we'll see how they feel about that.
It's an Well we warned 'em.
And they've asked for this.
Yeah.
So 130 companies have applied. And raise your hand if you're one of the people who've filled out the application. OK, and we're gonna pick by random, we've made a short list here of a few, I think there's 15 or 20 in And we're going choose them random, all at once. And get them lined up, and then you're going to have 10 minutes with each one.
I think eight So what I'd like is if we pick you, I'd like for you to immediately shout out so we know you're here otherwise we'll move to another one. And then I'd like for you to come up and sit at the front row here to my left, to your right, in the very very front, and then our sound guys will get a microphone on you.
Do you wanna do the honors? Okay. Ben Lavender of Dydra. Dydra Is Dydra not here? Shout out if you're in here. Not here. Alright. Okay, let's put that one aside. if you know these people and they're out in the hallway, yell. Go ahead.
OK, whoa this is a hard name him to pronounce. Sairam Chilappagari neighbourly? No, alright. From neighbourly, neighbourly OK.
Go ahead. Phillip[sp?] Sip man[sp?] of Data Curious.
You here? Alright. You're up.
Alright. Right place. Right time. So, just go ahead and set up there the sound guys will grab you. And let's do these in the order we're calling them for you guys just to sort of line up. Jody[sp?] Presty[sp?] of VidAppy[sp?]? Are you here?
Is that a yes?
That's a yes?
OK, come on up. I was getting worried, I said, what happens if none of these people are here?
We can just have a fireside chat.
We just talk, yeah.
Sixty companies in this batch, by the way.
Sixty-two. You're just letting anyone in at this point. No, no. Francois Laberge[sp?] of Motley. Motley? Emotley. Emotley, oh-- Oh I like this one. You do? Is he here? It's not that I don't like the other ones, but I remember Emotely[sp?] was kind of cool. No. Is Emotley [sp?] here? No, man. That's too bad, I was hoping to hear you give him advice.
Whose in charge of this? event?
Heather. Hey well, Ariana, I'm lost. Alright, Wow. Heather Rachna Singh and Vivek Handy of Hatchee Labs.
Hatchee Labs.
No.
No Hatchee Labs? You know what, they're out there, eating lunch. And they're not, they're going to, what did you just do?
This is a blank one.
Oh, that was a trick. That was the one where I got to say anything I wanted. I was going to pick these out. That's.
David Amsalem, a TV track or TV tack.
TV tack.
TV tack, is that you? Yeah, okay. OK, he's coming.
Make a noise. Like really make a big howling, like.
That's him. He's waving.
OK, he's going to be a quiet one. Alright. How many is that, by the way?
That's three, including TV Tack?
How many should we do, five?
Let's do five, and then if we have time, we'll pick one more.
Alright. Justin Riack of Flytivity?
Flytivity, come on up. Alright.
One more, right?
Yeah. Zubin Wadia of Secret Social. Secret Social here? Apparently not.
Here is the next one. Joel Fan, of LuckySheik.com? Okay, he's here. Oh, this is the one that didn't quite make it. We'll have this one, if you have more time.
That's the back up.
That one's the back up. That's the one that's sort of unfolded. I'm gonna leave them for you. I'm going to leave you and then, if you guys have mikes? Do you have mikes? Nothing. OK, first guy. Okay, so come on up stage. I'm going to leave you. Okay. Yeah, I tend to try to be the center of attention.
I'm going to not be here and just tell us who you are again, and what's your company and you guys do your thing.
Okay, here-- I'm Ben Lavender with Dydra, we're here pitching a cloud-based graph database graph data base in the sense of like graph theory graph? Yes, exactly. Or cartesian graphs? Graph theory database. Ok, so who uses it. Well, great question. People who use it are two main segments. Startups are talking to us about it for a social network analysis and more traditional segments are talking to us about it for things, like finding similarities between customers and product, where all of the aspects of a given Product, not know in advance.
So you guys, I assume, it's for searching gigantic graphs, or something like that. Is that the idea? Not even necessarily for searching gigantic graphs. A lot of people have smaller datasets, they're simply better represented as a graph.
Hmmm.
And even these can be much more efficiently computed.
Can you give me an example like, of all the people in the world, who needs you the most right now. Like, what problem most needs the software you built I really hate to use a name from somebody who presented here, but I'll say somebody like Rexley, for example, who was here yesterday, social music discovery Basically, what kind of music might I like, based on what my friends are listening to, what they're doing?
Okay, so some kind of collaborative Right. That's the low-hanging fruit. People who need to do collaborative filtering What the core at social network analysis, right? And social network analysis looks like a graph of friends and how they connect to your friends, and so if you want to see what your friend are doing, you can build a database that represents that natively and computes that, say, before the heat of the universe. but who has trouble with this problem now?
Who wants to do computations that would otherwise take too long? lots of people. I've been here for the last few days. Social is changing a lot of start-ups. Maybe 3/4 of the people in Start-Up Alley that I've talked to are doing something related to a graph.
Yeah. Your advantage probably is that you can run the queries faster, right? It's also much more easily representable for a programmer in a traditional database query language, something that might take 6 or 7 joins and be very difficult to tests really difficult to comprehend, really difficult to maintain can be represented much more simply in a graphic interface.
So people are going to run your code on their servers as a of their apps, is that how it works?
We're cloud-based on the Heroku model: database as a service.
OK, so you upload all the data into this great big sitting somewhere on the net and then you send it queries. Exactly. Is the latency going to be a problem? We're looking at that. The default standard protocal is over HEDP and that's about 15 milliseconds But we have a couple customers who want to test over other protocols that are much faster.
Fifteen each way shouldn't-- be a huge problem.
Cause I was thinking one thing you could use this for--one thing where you'd wanna do a ton of queries and get the answers back really fast as if you were using social graph type stuff to target ads. Right? Because like your just about to show an ad you're showing the page, what ad are you going to show on it?
You need the answer really quickly, right?
That's exactly a target market that we're talking to. We talk to e-commerce people who need to find similarities between customers. This old customer converted, how are they similar to this new one? So, what do I show them in the shopping cart to try to up sell them at the last second, for example.
Yeah, so is there anybody who said, "yes"? Is there anybody who actually wants to use this software now?
We have a couple of big testers developing products on it, yeah.
Okay. So these like aren't you friends?
No. These are not people I've met before.
They're people you don't know, who actually need to use this and do.
People who emailed me and demanded to use it, actually.
After reading some article in the press or something like that After examining their problem, realizing they needed a graph database No, but how did they know about you? Because we're the easiest.
No, no, no, they must have read about you somewhere, for example.
We got a little bit of buzz in our little segment, yeah.
OK. OK, so, can you tell me who the beta users are? In real office hours that's what I'd want to know. Who are they?
I'm allowed to talk about one in the real world. It's an Australian company that's trying to build a sort of competitive rap to plug into a CRM, say, so that you might be able to do something like, have a salesperson following up a lead who a competitive graph?
Right. So a prepopulated CRM, so you might be tell a salesperson that the CTO of the the company they're trying to get a lead built into a customer on used to work with the CEO of a competitor you need to be careful with bits of information. Oh, this is like LinkedIn. This is like what people use LinkedIn for.
Right? To find your friends of friends. To find the contact to reach somebody. LinkedIn is in a graph database, for sure.
So but why wouldn't they just use LinkedIn? Why do they need you? That's what I'm trying to figure out, what's special about you? lots of problems getting the graph databases, lots of business applications that are core to a specific business, business logic that's specific to you, need to be represented in a graph database.
If linked doesn't build an API that you can plug into your CRM, then you end to build your own.
I see, so that's the problem: LinkedIn has the data in it but it doesn't have an API that lets them run the kind Query they want and so they have to get the same data into you. And they have the data already about all of the connections of all their customers? Well, that's this particular customer's business, their building that database.
I see, I see. And building the interfacers that make it accessible to N users. Oh, I see. We're just the engine. They're sort of building a linked in competitor and they're going to use you as a component. I wouldn't call it a competitor. It's about, Sort of similar.
Yea.
Somewhat overlapping product, I see okay.
But it's just one use. I said, it's infrastructure, it's a steam engine . It's what you plug into it that makes. Are any of these people ready to pay you yet? Some of them very much want to pay us. They do? Oh they want to pay you because they need the product that much. So if they wanna pay you, are they paying you?
We're more concerned about engagement. We have about four more months in private beta, as we estimate.
But it seems like it would be good to take their money because you never know if they're serious until you try to get it.
Fair enough. And me and my competitors have had nice discussions about this. And I'm sure they're giggling about Right, and it will impress investors too. right? If you're actually making some money versus none. Cause investors--it's not so much that they care about the money--they care care about the money as evidence that people really want what you're building. Cause an investor's gonna hear graph database, what's that, right? and they are going to worry if you describe it that way actually they are going to worry that this one of the solutions in search of a problem when someone like things I could build Graph database would do these searches really fast by doing a bunch of interesting stuff in parallel.
Actually, when I came up here I was going to ask you about gonna pivot our marketing, which is the feedback I've gotten from Yeah.
many of the people I've talked to in here. Would you focus on a segment focus towards start-ups which are doing something like, social network analysis? First, or would you pivot towards something more established? The way you decide that is who is actually going to start using your software In production first right?
Above all, you need to get people actually using it. Because you won't really know if you're--you won't really know if it's good until they're actually using it right? I mean, good in the sense of actually helping them out. It could be an interesting piece of hacking, right?
Right, right, right.
So so like the way to figure out who is going to use it first is a combination of how desperately they need you and how quickly they make decisions, right? Cause even a giant company, even if they need you desperately it will take them six months before they can get Their shit together and actually have you running in their stuff, right?
So who is going to start using you the soonest? Obviously a startup, so you would say pivot that way? No, no Depends; its a combination of how fast they move and how desperately they need you, right?
Well start ups are known for moving quickly.
I know, I know. But are they also the ones who need you desperately?
Some of them. Some of them have problem It's the one I just gave you, their core problem Okay, so they desperately need you?
They need somebody like us, and they found us and decided we were.
Because they can't do what you're doing themselves. At least not well enough, fast enough.
Well they Rather not host on of the other databases. They'd rather just buy a service.
Yea, how we doing?
His office hours are over so Oh okay, let me, let me Yeah, boy that went sour. I'll be the bad guy here and I'll keep everything moving. Do you have any other things you wanna say to him?
Okay, my last words are. Like, find out who those users To use, and then describe, like figure out what kinds of problems they're solving and when you're describing yourselves to investors, describe yourselves more as solving that kind of problem, than as a graph database, right? Alright, sorry, that was only eight minutes.
Okay.
Thank you.
Thanks a lot.
Alright, bring it up, yep. I've also changed it to 6,so that we can actually get everyone.
Okay, everybody's got to eat lunch. Oh, it's got a timer, how good. Alright, 5:58. Hi, pleased to meet you.
Nice to meet you.
Okay, so you guys, I read descriptions of some of the, of all the start-ups in advance. I think you guys are making the thing for people to apply for jobs by video.
Yeah.
Great. And the idea is that from most jobs like, not very highly paid jobs, you don't need a resume. You just need to see the guy talking and see if they're like.
Yeah that's it. So like great big industries, right? Like retail, grocery, food services right?
Right.
Hiring managers will tell you that all the applicants look really similar on paper.
How do they hire people now? They don't use videos in interviews?
Walk-ins, right? It's Craigslist, it's, you walk around the mall and you've got 25 applications.
And the problem is, walk-ins don't scale, or something like that?
Well, walk-ins are massive interruption of time, so if I manage a Radio Shack let's say, all of my applicants come in at noon on Saturday and guess what's going on at noon on Saturday? The place is booked. I'm dealing with customers, etc. A really well-dressed, well-spoken applicant leaves my store, walks down to Zales: gone.
So instead of having my day interrupted by potential candidates, huge industries with massive turnover, Right?
When you need people, you just check into your website and you can click through a few people's applications.
So forget the idea of, what's your work history, where did you go to school, etc? Because it's not relevant for the job.
You just wanna see if the guy is a reasonable person.
You know in fifteen seconds whether or not you and I can work together.
Yeah.
So present me with a basic question like, what's a good reason to be late for work today? Or, tell me about yourself.
How do you know what questions you are going to ask them?
Well we've got ten years of domain expertise.
So you guys already know the kinds of questions that the hiring manager would want to ask. What are they?
They're simple and they're portable. So, tell me about yourself . Is it ever appropriate to lie to a friend? A customer left too much change on the counter, what do you do? So how you answer those questions are character revealing, maybe. Not necessarily skills based. So, you know, if I want to take an application that, or an interview, say.
And see I've got these basic questions. I want to apply it in my job at Radio Shack, Best Buy True value, etc., portably and asynchronously, right? So. It's good for the people applying too, right, they just do it once. There's so many inefficiencies in this system that haven't changed since I worked in fried chicken, right?
And still have not, to this day--so what do you do--you apply to a job on Craigslist?
Well, part about what you're doing is you're trying to create a marketplace. Like, one you're established, and all the people who want jobs know that this is the place to go, all the people who want to hire people know--then you're set, but how do you get there? It's always so hard to start.
I think the marketplace exists, right? We just want to organize it a bit better, right? Big players like Monster.com, right? Well, yeah. They're built around professional communities. That there really isn't a retail grocery application for what they're doing. But how do you get all of the--whose, first Well which one is the harder quantity to get; the people applying for jobs or the people hiring?
So, I think the people hiring have the need. Right?
Yeah, but who's hard to get?
Who's harder to get to?
No, who's harder to get on your site?
To get on our site are definitely going to be the people applying for the jobs, because there's effect involved.
Really?
Absolutely, right So jobs exist. But they're maybe hungrier. Are you sure? Well I don't either problem is insurmountable. I think you have Which one is more difficult, it's for sure going to get people to come to the website because they got to exhibit some effort, right? It's not just here's my resume, drag and drop in an email thanks, so long.
It's you have to answer some questions And two different minutes in our site.
Is this thing launched yet?
Well, that's one of the questions I had for you.
No, but do you have any users yet? Is anybody like?
We released it in a private beta to some big names that I can't necessarily mention. Big names, big retailers.
Big retailers. Right, so thinking mall stores and things like that. And so they came back to us and said tighten up x, y and z. And we're looking at the product being in the space Space, like job boards forever. I want to build this absolutely elegant beautiful solution. Maybe the way you get this started is to go to some of big retailers Hire people and say just tell people, go to this URL to apply, right?
And essentially their job applicants become your users. So, they probably won't ask you, hey There are these people that are going to apply to other competing retailers with the same application, right? Here's the thing. If they do then maybe You can make a special version where that one video is only limited to them.
Here's my question to you. There are big anchor retailers like say Home Depot. and everyone walks in there to get a job, right? Because there the benefits are great, right? Its a quality employer. But they're not going to hire everyone who walks in the door. But they're great collection for all of these potential candidates because if you're good for Home Depot, are you great for True Value?
Well you know what you could do; I don't know if you want to do this, but here's an idea. What if you let other stores pay them for their unused leads.
That's a great idea.
Right?
Yeah.
You can create, it could literally be like a marketplace.
Right.
Where, whoever can generate applicants, Yeah.
Like, anyone and they don't hire. They can just sell and you take a cut of the transaction, right?
Or as if you would have met my largest problem of getting people to come and interact, right? I zero cost of acquisition if I take the ones that Home Depot or Walmart don't want. And I turn them over to the rest of the retail public, right?
Who ever thought though that they could use recruiting as a way to generate money.
Generate revenue, right? Brand awareness, right? It's socially and economically important to help people get jobs, right? So don't just walk through a mall and collect All these applications, right? Apply at one place, if you're not a fit for there, let the other folks have at you.
Well eventually what happens is, people go to you, right. If you become super successful they don't think they're applying to Wal-Mart. They just like. They put their video on Vid-Ape [sp?] and like people hire them. Right?
It's world domination.
Yeah. I mean you could get there fairly quickly.
Sure.
You just have to get it so that someone who is thinking of getting a job knows you exist.
Right.
Right? You have to get to just like people now like AirBnB is. Got a marketplace and they're now in the happy position where they've sort of crossed the threshold and people kinda know about them already.
I agree, that surprised by challenge, with only 10 seconds left, I don't wanna be rude, but when do I launch? Right? I've got a platform. I have folks that are interested in using it.
I would launch now in the sense of having your website actually work if someone goes there. But initially, juice it by going to these big retailers and acting as if you were just a sort of outsourced video interview from and for them right? But just make sure that when people sign up to give an interview, they're doing it, they become one of your users right, they're like user base.
So use these big anchor tenants to grow your user base. That's what I would do.
Super helpful. Thanks a bunch.
Bring on the next one. Please introduce yourself. And you can walk up on, while he walks off. Incidentally, this is a lot more like real.
Real office hours, they you would realize, we've just accepted 62 new start ups for the summer. The first conversations I'll have with them, the last time I talked to them.
Interview half the time, I won't even remember who they are. So the first conversation I'm having with everybody right now, it's like what are you guys working on again? Oh yeah, let's see what if you did such and such? I prepared all these questions, but. What are you working on? Which one are you?
It's called data curious.org, my name is Phillip . We're here in New York. You do data visualization? Data visualization. Okay. So who needs it? It's a trend-site for investors. Okay.It's pretty basic.
Is this day traders? Not day traders. It's more long-term investors. But individual investors, or are you going to sell this to companies? Individual investors, I want this to be public. Individual investors who use your software to generate graphs stock market stuff?
No. Potentially it could be a revenue source sort of kind of like consulting revenues for us. What we are hoping to do is create a Wikipedia or Investigated for numbers and transform hot companies that are the next year or two, Zynga, Facebook, Linked in, all those folks. And you're going to have research reports on them?
Essentially profiles okay. So it's not something where the individual investor uses your software on the fly to say, "oh let's compare, you know, [xx]." Revenue versus the weather. Or something like that. Yes, so they can look at each one individually so we're starting off. But are they going to do queries on the fly and generate Yes, exactly.
OK, but it sounded like you were gonna pre-generate research reports. Are you gonna do that as well?
Pre-generate pages on each one so you can go go to direct special.
OK, so there's gonna be a page without a company, and then you do queries on the fly about that company. And what kind of queries What do you think is the most common query that people do?
I think the most common ones are gonna be iOS versus Android growth and phase a number of users versus Google scope. Right. And I think people want to compare them both and so what the raw statistics that you have are about all the growths of these different things and people are gonna generate Graphs, comparing the growth of one thing versus another?
It's essentially like Google Finance, but instead of looking at stock tickers, you can compare other data trends. I worry. I worry that this is going to be, this is not going to have big enough sort of gravity well, to stay up in the top of peoples' heads. That it would be the sort of thing that someone would do occasionally, right, and then never come back.
So I guess my target is info junkies, and people who are making a lot of investment decisions. They don't have to be day trading. Not necessary, but they have to be, I have some Apple stock and some Google stock. My portfolio is very small, but, I'm just really curious about what's going on and that kind of like how fast they go.
But you would be trying to do some kind of analysis comparing. And it's got to not the like the kind of queries you can do on Google and Google Stocks, whatever they call it.
Oh yeah, you can't do a heck of a lot on there. If someone were to come back over and over, what queries would they be running when they came back over and over?
Well, they could compare, you know, obviously these companies the news cycle is very very fast. The tech crunch pumps out all of this new information about them.
Yes.
All the time. So the information constantly gets updated. And my charts are always getting updated.
Right, what charts? What charts are people going to be looking at?
So if I want to say, number of apps in the App versus the Android Store.
That's going to be another statistic you've got.
Sure, right, exactly. It's always changing, sometimes one is going faster, sometimes the other is.
I wonder how many investors there are, who are, you know, that sort of ended digging into statistics, or are going to want to come back, over and over again, and look at a graph of Apple stock price versus the number of apps in the App Store or something like that. I mean, I can see wanting to do it once.
The big worry I have is, people aren't going to want to do it over and over again.
Well, I think it's a question of making it really interesting. I think that's where the user experience has to be very, very compelling. Right? It's sort of like how many times would somebody want to use Square, right, or a Twitter, right? Like there's something.
Well you've got to use Square. Like Square and Twitter are good counter-examples, right. You use Square every time you do a transaction, you'll use it all the time. You use Twitter every time you want to send out a message or something like that, you use it all the time. But doing these queries, they're not the kind of thing you're going to want to do over and over again.
Time the market moves. I think it's new information that you can say, "Okay should i sell the stock or buy the stock? Right?" That's a decision that you have to [xx] i worry, you're just not going to be able to find enough, because you're a start up, and people will pick out the start-up if it has something that they desperately need but unless they desperately need it they sort of won't hear about you know.
create [xx]. It's sort of--the information is there, right, they could go
Dave the reason i ask day traders? Cause they do need to do queries all the time, they're sitting there trading all the time, right? Long-term investors only think about their investments occasionally. Sure, but let me, let me tell you this. A) journalists are always thinking about this, right? So they always have to kind of look at these--Journalists?
That's not a very big market. No, no no they're not buying. In anything, right?
No, there's just not many of them.
Sure, sure, but they could have used it as a source, right? So there's other way to get comparable out.
Yeah, but you can't start a company with where it's just journalists. no no no hold on hold on. But also,i mean info graphics are becoming really, really popular. So,u know I just stopped talking to themselves . The first graphic I did where one Number one on Y Combinator, right? And that kind of stuff really drive your traffic.
So you don't have to wait for people kind of get to you, right? There's other ways to generate that interest, right?
I see, you can use info graphic spam to generate those users.
Well, not spam but it's useful. I mean, it's useful information One last question? Alright. If this was real office hours we would keep talking longer because I feel we are still sort of at an impasse. here. Okay. So far as it goes, I'm worried. You're worried that's it's not big enough? Right. It's not going to be Have a big enough attraction to enough people, to overcome your obscurity and get people to come to this site... Yeah, we'll I think in the next few months we'll see... Okay, okay... Good Luck.
Good Luck. Thank you. Okay. So please go on up. Introduce yourself. And he's got a demo for you. I don't know. How the rest of us are going to see it. Hi I'm. Hi Paul, pleased to meet you. What are you working on? My company's. What does it do? We are launching enough that makes every television interactive.
It makes the television into what? Interactive. Every television interactive. So how? You are sitting on your couch. You see something interesting on television. You point the TV with your iPhone. Oh this is the thing. In one second we are able to recognize what show or commercial you are OK, so where is that going to be most used?
Like if you imagine,like a year from now, you've launched and your successful and people are using this. a lot, everyone who has a business like that, they can always say, "Oh, the biggest single case of people using this is for blank." So what is going to be the biggest single case of people using it There are, the idea is to be like the info button on the TV.
So let's say you are watching news You are going to look for other content that is related to news. I am watching a commercial you might be interested. during the commercial.
Now, what one does, what I do anyway, if I'm watching TV and I'm curious about something like some actor that I see and I wonder, "Where else have I seen that guy?"
You search on Google Or . Or something like that. Right. So you have to replace the Google Search. Let's say you are interested in a commercial. What do you do? You Google about the product that you just watched on tv. Right? Occasionally. Ya. Usually. So you get hundreds of brands. It's a Mercedes commercial, you ride luxury car, then you can get all kind of different brands.
You know The call to action on TV it is very difficult. I worry that the kind of stuff, you're going to decide what to show me when I click on the thing, right? When I do a search. If you're looking at a tv, tv show, or movie, or something like that. I decide something I want. Right. And often it's something obscure.
Like if I were looking commercial and do a Google Search. The search is probably not going to be for the actual product. Commercial is going to be for some peripheral thing like the actor who did the commercial, right. Where it was shot in some weird place and I'm like curious like more about this.
Little town or something like that, right? You know, Fubar, Texas and I'm curious. Is there really a place called Fubar, Texas? Right, and I'll like do a search. It's that kind of random stuff-- that I do. So unless you give people the same thing they're searching for, unless the information you're giving people is the stuff they're doing searches for, then they're going to look at this and they're going to say it's not right.
Right? Because you're controlling what they go look at. You're essentially doing a search query with the query predetermined and unchangeable. What if it's wrong? So now what would you do? Well, I would test this on some actual people a lot. Right-- makes see if you can, if you have your friends and family, make them use it.
And see if you can get them to the point where they won't watch television without having Unless you can get your friends and family they have to err on the side of being charged for this. Towards you, they will give it a shot, you know. But unless you can make it stick on them, it's not going to stick on other people.
Cause this is supposed to be. mass market thing, right? It's not for some kind of weird subset of people...
Right.
who have some sort of special interest, right? It's for everybody. So Make it stick to your friends and family. And if you can't, you've got to change something or it's just going to fall flat when it launches. You what you could do? You could try Making the advertisers drive the adoption. Right. You have this underlying technology for like Recognizing a t.v. show or a movie or whatever.
This doesn't have to be the way you use it. There could be other ways to use it besides being this thing that every consumer is just sitting there clicking on. You could, for example, you could use it, you could invent some kind of Nielson rating-like thing. Where a TV show, you know how radio shows say the first hundred callers get something or other.
They do that to drive up some number that show's how big their audience is, right? Maybe you can get some kind of TV shows to put something in them or to encourage people who are watching them to like check in. You could use this to check into your TV show, right?
Right.
You can prove your there,right? You can have some Iphone app and like you can say, they can run sweepstakes to get, to drive users to watch their thing, right? They would be gaming the Nielson ratings, which is, that's what everybody wants to do. They want to game the ratings system. They'd be gaining the ratings, right?
Exactly.
Check in by proving your watching the thing by pressing the app and then your technology isn't so much figuring out what store, what show they're watching in order to give them information, but verifying what show their watching, in order to enter them for the contest. How about that?
Actually, we have a lot of
And then you don't have to be useful for the consumers, right? Because if the advertiser Where the show producers would be useful for. I would try that. At least in parallel with what you are doing now. Go talk to some people who are producing shows to see if they would be interested in something like this.
And what about the next, afterward? That if is they are interested it works, then we have someone that is, if you get people like, if you get some big show to run some contest where people have to check into the show using your app, they will have to install your app in order to participate. That will get you possibly millions of installs, right?
So then you have a user base. Somehow you have to encourage them to keep using it. I think though if this thing worked. If it did actually work, to cause people to check into shows then allot shows the users. Right, why wouldn't they? If they can get more audience that way. Instead of the iPhone being this distraction that's getting attention away from them, it can be a method by which they get people to watch [x] right.
You know, a dying industry, fighting back using the new industry. Um, we could ride that for a long time. [x] Are we running out?.. You are out of time... okay... alright, alright... that was interesting... That was cool... Remainds me of into now a little bit... Yeah... [xx] vary. Sometimes all we do is beat our heads against the wall, and sometimes we have interesting new ideas for doing stuff.
Although OK. Are you the thing that introduces people in airports?
Absolutely.
OK, so the idea, let me see if I Remember this. The idea is your travelling in an airport, and you tell people who I should want to talk to
Absolutely
Is in the same airport so i can talk to You know business travel - not social. You sit in a lounge you don't talk to anybody.
Yeah cuz i don't want to.
Our concept is It's a sales dude
and I'm not talking sales more an executive type play. But, if you knew that somebody was here that you could see their profile, their LinkedIn public bio, and say, "Oh, that guy looks maybe I do want to talk to him. Otherwise you're just gonna sit there with your back next to you staring at CNN.
It's more general than just airports though. It's really when you're some sort of, when you're in some different place than you usually are, and there might be other people there too. Like here, probably somewhere in this place is somebody I've been meaning to talk to for a long time and I don't know they're here.
although that's not what you tell me. You tell me about people who are friends of friends or something.
No, no, no. It's breaking outside of your connections, it's not really who you know but who you might want to know.
So who do you decide who I might want to know?
Based on some very simple 3 profile questions that you answer when you sign up Like, what do you read? You will only introduce me to other people who are already users of the system. Yes. That's tough, right? I know. You'll go. And that's one of the questions I wanted to ask you. How do you chicken and egg problem.
Well, the way to beat a chicken and egg problem is to find some tiny subset of the market that is small, but much more driven than normal. Is frequent flyers top status level people, that subset, could that be that subset, you would look at flyer talk. No, I don't think so. I don't think so. Because all those people have in common is that they fly on planes a lot, right?
That doesn't mean they want to talk to one another. One could be a sales dude and one could be like an engineer who has to fly all over the place to service obscure equipment. That, I feel like airports, it's the danger of calling yourself too. I know it. Your name. us to the airport. Really we are looking at, you know, everywhere along the business trip.
The airport and in the But the real underlying technology is for introducing to people, people to other people around them its sort of like push LinkedIn, right? Because in LinkedIn, the data's there, you can figure out who Yeah, we're tapping into that.
Oh, and you have to know where people are, so I guess they have smartphones.
You can trip it, it's like a tripit LinkedIn. so somehow you need to get a whole bunch of people to sign up for this because they want to meet one another, right?
Right.
I mean, that's the whole problem. If I were you, I'd be spending all of my time trying to figure out, don't think of anything except who this tiny market, even if it's just a couple hundred people, who really want to be introduced to other people who are in the same area, who is that, right? Well, Looped.
We funded a company that made something like this. Looped connect, right. You know, it kinda works. There are others out there, and our premise is that But you've got to do something different.
People actually do want to talk to each other, they just don't know what to talk about. They don't know how to approach somebody... I don't feel like airports is the secret... No?.. No, cause airports collect people together randomly... I feel like almost any... any other way of collecting people would be better.
For example, a convention. All the people, something like disrupts. Traveling to a convention? We're, this is just like an airport except we're not waiting But it's all these people that came here from somewhere else, and they are all in the room together. We are much more likely to want to talk to one another than people in airports Airports are the worse case.
I mean the worst cases. But I'll tell you that I saw somebody here that I saw in the airport. Yeah. Monday morning. Had I known that guy was going to be here, I would have talked to him. Well, if you had Geo-location in the future, then you'd be set. But if you had Geo-location in the future, there's probably even more valuable things you could do with it.
You just don't know, you don't know where people are going, unless you expect them to tell you all the time. And initially, you won't have the power to do that. So, you have to find a subset of people who really super want to talk to one another enough to seek out and sign up for some new thing that they're not already using, right, and use it.
I don't know, if I were you, I'd start over. How much time have you spent on this?
60 days.
60 days, alright, fine. Plenty of companies start over after 60 days. I mean, I don't think, I think the amount to repair this idea. It would require like 98 percent new stuff, right? You might as well just start with something different, right? So, what I would do is like, ask yourself, the very best start up software All the problem the founders themselves have; so I would ask yourself.
The trick is to say, "what do I wish somebody else would start a start-up to do for me."
Right .
Right? What is the worst problem in my life?
This is where I...
This is the worst problem in your life? That you don't know who to talk to in airports?
I travel 250,000 miles a year. I...
Whoa, and that's the worst problem. Who to talk to in airports. You waste people, you feel like your wasting your time when your sitting in the airport. I'm profiling, that guy looks good, I'll talk to him, wait, I should have talked to him.
Don't you just check your mail or something, when you're waiting in the airport? Yeah, but you've gotta do it anyway, that's what I do. Is this really? Well, maybe, maybe then, maybe it isn't so crazy, if you, if this is really honestly is, one of the biggest problems in your life, that you're sitting in airports and you want to know who to schmooze with, then maybe you're not the only one who has it, and maybe you can find, the question is, do the people you want to schmooze with want to schmooze with you?
No, but it requires, see, if you merely signed up all the people, if there were a bunch of people like you you who are sitting in airports and thinking, what missed opportunities, I can be schmoozing with all these people. If those people are also the people you want to schmooze with, then you're set, because you have both halves of the marketplace.
But if you don't, if you want to schmooze with people and the people you want to schmooze with Don't want to, they want just want to check your email and it won't work.
Yea by sending the message they don't respond.
How are you gonna get this thing? How are you gonna find like some little substance the market to get this thing started in. I don't know, I don't know, I don't know how to do it.
Something All right, all right, we're done.
We've got one more Somewhat of an impasse Can you handle it? Are you Oh, I'm ready. Are you guys doing OK? Is this interesting? I have no idea Okay, come on up. Last one, I am not trying to be I'm acting like this is really office hours and I'm trying to figure out how to make these guys look good on demo day pleased to meet you.
I am Jolthan from luckychic.com. We are a social shopping network for women.
I think I remember reading about you guys, and you already make a lot of money right. Well no, we don't make a lot of money. We make revenue top line, but we don't make bottom line. We are a [xx] so we have hundred thousand members, women. You'll see. Where did those hundred thousand members come from.
What drove them to you. A lot of hard work. Barter deals and ad words. This is good. What I am thinking is, ah scrappy not hapless. Right. Well scraps, definitely scrappy, yah. Ok. So what do you do for them when they get to your site? Well, you know the way that we're disruptive is [xx]. We looked at what Facebook does.
Facebook has profiles and a development platform that you can build a lot of apps on. And we've done the same thing for shopping.
So do people old apps on top of your platform?
Right now, just us. We build apps and game apps on our platform revolving around social commerce, so a lot of Did you say game apps?
Yes, gaming applications. Really, so these 100,000 users, what they do on your site is that they play games?
Well, we call them games. But because everyone talks about ramifying e-commerce and so actually does a typical user do? A user shows up at your site and what do they see on the screen, and what do they do?
A user shows up on our site, and they see a variety of ways to interact with shopping, all the way from discovering, trends.
What's the most common thing they do?
The most common thing they do right now is, bid on items and peek at prices, which they don't So you have an auction site?
We have an auction component. But in our model, we think of an auction as a single game on our site, a single model.
Is this a penny auction?
We do have a penny auction component, but we have other components of our site that we've built on our platform.
What's the biggest driver of traffic?
The biggest driver of traffic are the products that we offer on our site.
No, wait. What, is it penny auctions, is that the biggest driver of traffic, like people come to the website and they see, here's something you can buy for such and such, and it's going to stop him, like 30 seconds or so?
That is a big trouble driver for our site, but let me clarify this, what we have done is We have built an entire platform, and we have 9 experiences on our site right now. We have social profiles Yeah We have something that we call lucky look, which is fashion trend, game that you can rate other members, do you essentially break even on the penny auctions? We essentially break even. Ok Well, what we've done is we have bootstrapped a member base that's very targeted.
This is sort of a grim way of, I know you're trying to describe your start-up in a nice nicest way possible, right, but if this were actually office hours, we would be talking very candidly here, and I would be saying essentially what you have is a penny auction site with 100,000 users.
No, that's not true.
But in a nice general architecture with a platform underneath.
No, actually not, because I think the concepts how you think about your business is incredibly important, right, whether you are a telegraph company or a communications company Yeah And I think that what we have built is a platform. And we have a lot of ways for people to play that are not just oriented around paying. Like we have whole point system.
So far this is the one, that they actually use is the penny auction.
So that is how we started, yes, but, but just because you start in a certain area doesn't mean.
Yeah.
doesn't mean that's what [xx]
You kinda want to grow into other areas right? You started out as this popular penny auction. Can you break even? Can you grow users and break even, or does it cost money net?
We can break even but the issue is that of my ambition, as an entrepreneur, has always been to build company with true skills and true scale.
Yeah. And is it just too hard to get these fist hundred thousand? I mean, a hundred thousand users is a lot. How, what do you count as a hundred thousand users? Is this monthly?
Registered. Registered users.
On a given day how many people use your site?
On a given month it is about 7 to 10%. We have about 7 to 10% active users.
So, 7,000 in a given month.
7 to 10,000 in a month, yes.
Do they return, or is it just?
We have a certain amount of churn, but part of our message has been that by introducing additional experiences, shopping experiences and social discovery experiences, that the users will go from one experience to another.
What's the growth rate like?
The growth rate is basically flat to say, about ten percent to twenty percent growth. But that said, we have spent a lot of time on building out our platform. So you are in what we know, what we technically call the trough of sorrow. There is like spike of interest when you launch and then this thing where you are sort of bumping along the bottom flat and then things start to turn up.
Yes, the challenge is definitely creating that lever that is going to get us that exponential model.
At least you're not starting from nothing.
No.
The advantage, so far, you haven't got the product that's going to make you a star, but you've at least got something that gets you 7,000 people a month, right?
We have data, we have data, we have behavior, and we have profiles of people.
So you know, here's an encouraging example. You know who else was in this very same situation? Groupon, back when they were The Point, they had the very same kind of thing, this thing that, they had some number of users but it was flat and when it was originally. The point was this thing where people would get together to do some sort of group action like a protest or something like that and after a year of bumping around along the bottom they decided they would start doing group buying and it just took off, and Groupon had been one of the things they'd been seeing people organize through The Point.
So what The Point "What got them in retrospect was that they got a base of users so that when they did find the right thing, they had some people to launch it on." so, it got them some experience in looking at what people did. Because I'm sure it wasn't coincidental that Andrew's next thing with "It's like these group buying things, because he had seen people organizing group buying on Groupon.
So you should figure out, what are you seeing people trying to do?" you're wanting to do among your users. There's 7,000 people, right, statistically significant size of sample. What do you see them trying to do that you're not really suited for yet? Or that no one really gives them? Is there anything?
Well, We know that they're coming to the site for various purposes. We know that they're coming to save, and we know that They want to save money.
They want to save money, and we know that.
We about to be done?
We're done.
OK. We are done, the time's out.
I'm sorry. I have another question for you.
Yes, how often do you tell entrepreneurs to scrap their idea and start something else?
You mean what percentage of start ups we fund? Do I tell them that?
No, no. When your doing Office Hours refurb in the ideation stage.
Yes. She told me to forget it, should do something else I think this worked really well.
May be 5%, they are more likely to realize.
We are done now.
Should change what they are doing now. going to give away the motorcycle Are you going to give away a motorcycle?
OK, guys, we're done.
Thank you.
No, but Paul, I need you to stay on stage for this.
OK. You guys, that was pretty cool, right? That was awesome.
That was very cool. Thank you so much. Stay on stage.
You're going to be a special judge. Don't worry its not going to, only one
Today we tried something new at TechCrunch Disrupt: a special, on-stage office hour with Y Combinator cofounder
. The goal was to reproduce the sessions that Graham and other YC partners hold with each of the startups who participate in Y Combinator — except the startups at Disrupt were getting sage advice in front of a few thousand people. And boy, was it awesome.
Six companies were chosen at random from the TechCrunch Disrupt Startup Battlefield. Then, for six sessions of less than ten minutes each, Graham spoke with each startup founder to flesh out their idea, asking probing questions as he tried to figure out what they were setting out to do, and what they might need to change.
Graham has a knack for being insightful and critical on his feet, and he doesn’t require much background information to hone in on some of the pain points and weaknesses in a startup’s idea (it’s a skill that likely comes from practice, as he’s held office hours with hundreds of YC companies). At the same time, he comes across as being curious and empathetic rather than overly negative — even when he’s telling a founder that their company is dead in the water.
If I were to recommend one video from TechCrunch Disrupt (and we’ve had many), I’d choose this one. Founders should consider watching the video above and asking themselves how they would answer each of Graham’s questions. It’s really fantastic.
Below are my notes from each startup, but it’s really best to watch the video, as it’s hard to convey the train of thought that Graham exhibited on stage.
The company is offering a graph database as a service.
Last words: “Find out the users who use you first are, figure out the kinds of problem they are solving. Then when you’re talking to investors tell them the problem you’re helping solve.”
Apply for jobs by video. Job search is about asking rich questions, not work history: “What’s a good reason to be late today?”, “Tell me about yourself.” “Is it ever appropriate to lie to a friend?” Character-related questions.
PG: “What’s hard about what you’re doing is that you’re trying to create a marketplace. Once you’re established then you’re set.”
A trading site for individual investors that generates queries on the fly for a company, sort of like a Google finance but with other data instead of stock tickers.
PG: I worry that this won’t have enough gravity to stay on top of people’s heads. Something people would do occasionally and then never come back.
PG: What queries would people run if they came back over and over?
PG: I wonder how many investors there are that are that interested in statistics.
PG: Doing these queries isn’t something people are going to want to do over and over again.. I worry.. You’re a startup, people will seek out a startup if it has something they desperately need it. But if they don’t desperately need it, they won’t hear about you.
PG: I asked about day traders because they do need to do queries all the time. Long term investors only think about this occasionally.
PG:Journalists aren’t a very good market. There aren’t many of them. You can’t start a company where it’s just journalists that are users.
PG: If this were real office hours we would keep talking longer, because I feel like we’re at an impasse here. I’m worried. It’s not going to have a big enough attraction to enough people.
Setting out to make every television interactive. Sitting on couch, point at TV with iPhone, one second recognize what you’re watching.
PG: Where is that going to be most used? FIll in the blank. The biggest single case of people use this is for ___?
A: As an infobot. If I’m watching news, going to look for other content related to news.
PG: Right now if I’m watching TV and I’m curious I look on Google or IMDB. So you have to replace a Google search. You’re not going to decide what to search — I decide what I want to look for. If I see a commercial and do a Google search, the search probably isn’t for the product, it might be the actor or where it was shot. So unless you give people the same thing they’re searching for… they’re going to look at this and say it’s not right. What if it’s wrong? I would test this on a lot of people… see if you can get them to a point where they won’t watch TV without using it. If it won’t stick on them, it isn’t going to stick on other people. Make it stick to friends and family, and if you can’t you have to change something. You could try making the advertisers drive the adoption. Or try to use it to game Nielsen.
PG: If you get a show to run a contest, and people have to install your app to participate, that could get you millions of installs, and then you have a user base. Somehow you have to encourage people to keep using it. But if it did actually work to cause people to check into shows, a lot of shows would use it.
Introduce people at airports. You’re traveling at an airport tell people they’re in the same airport.
PG: You need a small subset of users much more driven than average user. Frequent fliers can’t be that subset because they’re essentially random groups of people. Could be a sales dude, could be an engineer.
PG: I don’t feel like airports are the secret. I feel like airports collect people at random.
PG:You don’t know where people are going unless you expect them to tell you all the time. You have to find a subset of people who super-want to talk to one another. Enough to seek out and sign up for something they aren’t using. If I were you i’d stat over. You’ve spent 60 days on this. Plenty of companies start over after 60 days. The very best startups solve problems that founders themselves have. What is the worst problem in my life.
PG: The question is do the people you want to schmooze with want to schmooze with you.
A social shopping network for women.
100,000 registered users. Build games.
PG: “What’s the most common thing they do.” Bid on items, peek at prices. Mostly penny auctions.
