Live office hours with YC's Kevin Hale and Dalton Caldwell (2017)

Hey, how's it going? We're going to get started. Yeah, I'm Dalton. I'm a partner at Y Combinator. And we're going to be doing office hours today. Yeah. Yeah. A couple things to note.

I know that a couple of live office hours happened beforehand. We're going to try to spend a little bit of time, even after these, to talk with the different companies. 20 minutes for office hours is not a normal time period that we'll do stuff. So we're going to try to do it as fast as possible. Apologies in advance.

We might interrupt you, just because we know how short the time is, to try to get as much information out of you for us to communicate as well. All right, thanks a lot. I think. Yeah. How you doing? Kevin. Kerry. Nice shooting.

Dalton, how's it going? And then I think you'll need to use the mic, right? Yeah, the handheld mic. Yep. So just very quickly, introduce each of yourselves, like your name, your role with the company, and then name of the company, and what you guys are doing. Yeah, I'm Shruti. We're with Hubhouse. And.

Kerry, and I'm the CTO. I'm the CEO. And then what does Hubhouse do? So Hubhouse does co-living. We provide co-living spaces for professionals. So if you're new to the area, you've just, you've just gotten here, and you want to rent a room out in a house with a curated community, and you want it to be set up, you can move into one of our rooms. And then, what does co-living exactly mean?

Co-living means living with a group of people. So basically shared housing. How many people? Our houses are five to ten people. Some co-living spaces can range up to 30, but ours are five to ten. And then, is it, is the characteristic that you're sharing a bathroom with a bunch of other people also? Not necessarily. Some are, some don't.

Gotcha. And then how are things going for you guys so far? Pretty well. So we launched 18 months ago. We are now at 220 tenancies across over a dozen cities in the Bay Area and Los Angeles. What's the occupancy rate for the house? We currently have a occupancy rate of 94%. Most of the vacancy rate comes from our launch vacancy.

So our vacancy rate, not including the first two weeks and the last two weeks, is about 2%. 2%. So when you say 94%, like how many people does that mean, like in 220 properties? So I'm sorry, so 220 tenants total. Tenants. Gotcha. 220, yeah, that includes, so that doesn't, so we have more rooms than 220. We have about 230 rooms.

Yeah. 230 rooms. So in terms of the timeline, when did you first start? Like you said, 18 months ago, but what did that look like, the very first tenant, I guess? Yeah, so first thing is we got a house, and I was actually, we both still live in our hub houses, but I moved into our very first hub house. And then we got a bunch of tenants, and the first few houses was very much so a learning process.

We were trying to figure out what works, what doesn't, and try to create this backbone for our co-living spaces. Does that answer that question? It took about seven, I think it's seven houses in the first three months. Okay, and so you guys take on the lease. What are the terms of the lease that you take out? It's just a standard one year lease, but then it goes month to month? I see.

Yeah, so we do one year to two year leases for each of our houses, and Do you need any special permissions from the actual owner of the property to do this? Yeah, so we are completely upfront with our owners. We let them know exactly what we're doing. We sell it as a benefit. We act as a property advisor to the owners, and we guarantee them rent every month, even if the house is completely vacant.

And then what's the margins that you get currently? 30 to 40%. 40%, so let's just say, let's just take a typical property. You have one house that has how many rooms on average? Six. Six rooms, and then how much are you paying rent on that house? So roughly $6,000. $6,000, okay.

And then you guys will make 40% on top of that, usually through renting out. So if I'm a property owner, why is this a good deal? You said an advisor, I don't really know what that means. Why am I like, wow, please have a bunch of people come into my house, and maybe it's going to depreciate a lot.

Yeah, so most of our property owners have pretty large properties which are hard to rent out, so we're able to quickly rent out a house which maybe rents for like $6,000 to like $8,000, $9,000. And instead of having to deal with a typical roommate situation where you have to deal with five, like six or seven- But the landlord doesn't care about that.

As long as they get paid, this has been going on a while. The landlord is like, I don't care. As long as the rent shows up, whatever you guys are doing, not a problem. Well, so at the end of the day, as a landlord, you want one check. You want to make sure your house is taken care of, and you want professional people living there, and you don't want to have to deal with it.

So we take care of all the hard stuff. If there are any property problems which come up, we advise them on what to do. But at the end of the day, all you're getting is one check, and you don't hear anything from us. So what's the rate of growth? So how are things going right now? You said you're doing pretty well, but is revenue increasing monthly or weekly?

Yeah, so we just raised a round, and after we raised, we've kind of scaled back, and we've been growing out our team. Why would you scale back? We wanted to hire a team. We wanted to make sure our processes were going well. We will. Sorry. We were working about 80 to 100 hour weeks, each of us, and we were kind of reaching the limit of what we could do without a team.

So just to go back, so what is the rate of growth right now? About 8 to 10%. 8 to 10% a month? Yeah, and we're growing, so before that we were doing about 15 to 20%. How fast are you adding on new properties? Is that pretty much your bottleneck, finding supply? Right now, our bottleneck is operations. So we want to make sure that each of our houses is launching smoothly and well.

So that's what- So right now you don't have a playbook for launching, how many properties do you have? We have 40 properties. 40 properties, and you feel like you still don't, why do you feel like you don't have a playbook after 40? I mean, we definitely, we do have a playbook. So we do have a step-by-step process that we go through.

But we've just gotten our team on, and we're building out our tech and training so that we're able to scale sustainably and provide a fantastic experience to each customer. So when you say you want to build out a team to build tech and training, can you just dig into that a little bit? Like, what do you need the team to do? We now have a team.

So before, until December, I was the only full-time person working. And in December, Kerry came on full-time. Kerry was working as a co-founder before that, but he came on full-time. And then we had one employee in December. So the past few months, we hired on a property manager. We hired on a person to help out with partnerships.

And we want a second developer to basically help us build out the tech to automate a lot of these manual processes that we're doing. So just like responding to emails, responding to tenants. With property management inquiries that come in, there's ways to automate a lot of those steps.

We do look at ourselves like a marketplace, but it's not a true marketplace because we're not connecting the landlords with the tenants. And so we do have the supply and demand factors of it. We found that it tends to go back and forth between. That doesn't make any sense to me. I don't understand that. Which part? It's two parts. Yeah.

So, right? And you're connecting them together. So I don't. So the tenants never deal with the landlords specifically. And so we're not connecting landlords with the tenants. We have to make the deal with the landlord themselves. Right, but it's one ecosystem of supply, one ecosystem of demand, and you're bringing them together. So regardless of how the connections happen, it still works that way.

What I'm trying to figure out is, how does this sort of win? And the thing is, do you feel like part of co-living is this movement? People are realizing, I want to work. And it's a lot nicer if I can work with a bunch of other people. The places where I want to work, I take it you mostly are in the Bay Area, right? Yeah, we're in the Bay Area and LA. Then it's really expensive to get places.

And so for the places where I'm trying to get in there really quickly, I might not have enough capital to move in. So co-living is a way to step into that. So what I'm trying to think is, do you have a reliable source of supply for people where this, let's say this is a really big trend, that you're going to be able to onboard.

And I'm trying to figure out is, at what point do you feel like you are ready to add on new properties? And at what rate do you add on properties? So we're picking up approximately four to six houses a month. Four to six a month? Yeah. Okay. And then how long are they vacant before, like how long does it take to get a house up and going?

So the time from when we sign a lease to the time the first person moves in is about two weeks. The time from when we launch the house to when it's 100% full is another two weeks. So about a month before it's a completely up and running hub house. About a month. So we've seen a lot of people work on this idea, and it seems like you can get about as far as you have gotten, which is really great.

But because you actually have to take on the lease, and because you have hard costs, sometimes these can look much more like a real estate business than a tech business. And so I think what I kind of want to get at is, are there ways that you can not hold the leases? Is there a way you can make this much more efficient from a capital basis?

because right now, this is basically a real estate business, right?

And it sounds like there's some tech in there, but I haven't really gotten from you that tech is the differentiating factor that will make, it sounds like it's a lot of operations work to sign leases and then get people in units, which is pretty traditional residential housing type business, so how are you thinking about what lets us get huge, right? Like what are your thoughts along those lines?

So part of that, so there is a good deal of tech. I do agree that it is not the key differentiating factor that we have. Some of it's different, and we automate a lot of the operations that are happening. And one of those- For like, hey, the faucet's broken, can someone fix it? And you open a ticket, is that what you mean? Can you tell me what that means?

So more along the lines of automatically finding houses to go and see using some different methods there, automatically posting. And so we have a constant in stream of probably about 100 tenants reaching out to us a day that's being automated by tech. A lot of the interoperations going between that is operated by tech. None of that's very, some of it's unique, some of it's not so much unique.

I'd say the bigger differentiating factor, which we haven't mentioned yet, is the community fact that we build into it. Which is, a lot of people come because it's cheaper and because it's a place that they need quickly. But we have a very strong community element and create an infrastructure in each house. So people are actually forming these really strong bonds. How do you make that happen?

So part of that is we set up monthly dinners and we give them formats of these are things you should go over. And we throw events every month for the entire community. And so we're actually connecting these people. And for that reason, we've had a continual growth.

While we started with month to month leases, we've moved up since then, our average stay length has been a continual process upwards since the beginning and still going. Quite honestly, I feel like this really comes down to cost. Like people want to move into cities in places where it's highly desirable but property is like not abundant.

And they're trying to figure out how can I get to this in a cost affordable way. So co-living is a way to sort of make it happen. You can advertise all these community stuff. It's stuff that you can tell the people, but it's no different from an apartment or condo complex. It's like, hey, we have these community stuff. We have a pool room. We have these wealthy dinners or barbecues.

It doesn't sound any different, because no one says, we don't have a community here. We don't do anything nice for the people. Make sure there's a bunch of shitty people here living inside of it. You don't sound any differentiated from it. So all I can think is this is pretty much, it's either a land grab, because usually you have more great properties than anyone else.

And then onboarding feels so simple and easy. Basically, it's like someone contacts you on average, how long before they're moved in? Is that a month? No, so it's, I mean, it's as short as a week or two. But I think the difference is that the house is, access to rooms isn't easily accessible right now.

Access to rooms in single family homes with curated communities, the only way you can get access to these is through Craigslist or by setting it up yourself. So we're unlocking a whole new market, which is- The whole thing is you build a brand that gets people to trust you. Are dead houses still a thing? Rob Levitsky, anybody? Dead houses? Anyone live in a dead house? Yeah, okay, cool.

Yeah, okay, so clearly there's people that want to do this. And okay, so it sounds like we have about five minutes left. So what are the things you're thinking about that can make this as constructive as possible? Because just to summarize, sounds good, you're making money, good job. What might be tough is, what's the move to make this huge? Let me give you some options.

Sometimes people that are working on the startup idea decide to actually buy their properties because you can renovate them. And then you actually own an asset, and it becomes almost, again, like a real estate business. And I've seen, that's kind of the WeWork model, right? So I've seen that happen.

On the other hand, you can try to make this as asset light possible, where maybe you're not taking the obligation of the leases, right? And then you can grow really fast and somehow have a special deal with the landlords. Where you can just grow really fast, and then you're a true marketplace, because you're not holding any of the liability that right now you're holding.

Or you can, again, maybe you found the just right compromise between the two. So this is sort of where I would think about from the long-term perspective. Do you have thoughts around where you see this going in a year or two? So the latter is probably the direction we're going to go. The reason why we have it is just because it hasn't really been a problem until now.

But in the future, as it does become more of a problem, we do want to shift more of the liability to the property owners. Got it, and one quick question while we're on that. Is there any insurance implications to this? What kind of insurance do you have to take out?

So we do have a pretty comprehensive insurance policy to cover our properties for tenants to damage anything, up to if the tenants burn down the house or something, they are covered. And then we are also asking our tenants to get renter's insurance on top of that. If your problem right now is operations and you're recruiting, trying to find people, what is your pitch to people?

To get people on board to say, come and do this startup. Because it's not many people's dream to be like, I'm going to work on basically the new form of apartment complexes or dorms, basically get dorms for millennials. So I think the biggest piece of our company is our community. So each of our houses is almost like a family-like environment, and people can really buy into that.

I mean, I think a lot of people out here don't really have access to that, and around universities, sure, some people have access to things like dead houses. But most people move out here and are lonely and don't really know people. So being able to offer the sense of community and the friendships that form from within our community is probably the biggest pitch that we make.

And so you think that's the long-term defensibility thing is ultimately I'm just trying to make sure I understand. So you're saying, look, we're going to have the best community and the best brand for community, and that's what will attract people. because again, with this kind of business, it's very easy for new entrants to come in.

So if we use the example of Airbnb, anyone can make an Airbnb clone. But what they actually had was network effects, where just like eBay, it's hard to communicate with eBay because they have the buyers and sellers. What's the equivalent here? It is the community? Is that what you're saying? So it is branding, a lot of that.

And we do have a network effect, because it's not just community in the houses. We connect all the houses to each other as well. So we do have a bit of a network effect there. I do think branding is the defensibility part. The other key part that Shruti's mentioned to me in the past is, most of the competitors can do one or two things well.

So Campus was a previous company who did something very similar, but they had very low margin on it, and they did the community aspect really well. People loved staying in them, but the operations was awful. And then there's other companies that do the operations very, very well. But, and many people have come to us and said, everybody talks about community, but you're the only ones with it.

And so we feel that what we have is different, but we are also good at operations. And so that's the part that we are trying to scale and trying to hire for tech to turn it more and more into a tech business. And to take all the lengthy processes that are taking operations, like we're hiring for more engineers right now.

And making that as automated as possible, so we can do less of the actual, the signing the leases and the direct things there. I think all the community stuff are good. It's stuff for you to focus on. It'll prevent churn from happening. Before the conversion, my feeling is your bottom line is probably still supply.

So whatever it is you need to do to relieve yourself on what allows us to say, we want to take on more properties. Because really, what you've set up is your liability is that first month's rent that you need to take on, which is pretty decent, sort of capital liability. So if you can figure out, how can I turn around where I sign a house, right?

And I can turn around, I think you also have to put a safety deposit, like a security deposit also. So it's almost two months worth of rent that you're taking on before you can have it covered back. And how long before you have that covered? So it's about a month. About a month, so- It's going to be about two, no, it's going to be about four months before you cover two months worth of rent.

Well, we collect back security deposits from our tenants. Great. And then I think that's all it comes down to. You need to figure out how to do all the community stuff up ahead of time. But everything that allows you to say, I want to take as many properties as possible and figure out how to jump start them. And get people into the houses or homes very, very quickly without losing quality.

And so the whole thing is, what is that exact playbook like? And you want to almost time yourself. It's like, okay, I've identified a property. How long does it take? What is our checklist that's very, very reliable for us to determine if it's viable or not? Go in there, figure it out. And that's where part of the tech is going to be, seeing how much can we do online.

And then the second part of it. it's going to be like, we send someone down there, they do a very quick assessment, and you have a team that quickly turns around the place, takes the pictures, puts the listing up there, and gets it out to people. So that's like the first, supply.

And then your demand is just like being very proactive on getting the demand so that a lot of the people, tenants, like you get testimonials, you send it out to whatever universities or wherever people are going to be looking for housing. And so make sure that they all think like, HubHouse is the first place I go to consider moving to a new city.

It's like, I want HubHouse to be there because I know that's a good, even if it's transit or it's like, I want to go there, be there for maybe one to two months, and then I'll, maybe I'll transition to something better. Or some percentage of them say like, I just want to stay here because this is the kind of community I want to be with. Imagine most people are going to eventually graduate.

Nobody wants to live in a co-living forever. Some people might, but that's actually, I think, the thing I most worry about. Co-living, kind of like co-working spaces, the thing I always get nervous about in terms of brand erosion, is that it's not an easy way to like churn people out.

And the thing is like, in places where you go where great people will graduate, what ends up being is like, it slowly starts rotting with people who like can't leave. And it's like, I don't have anyone else that loves me enough or wants to live with me. Or, and it, it looks like online forums, it looks like co-working spaces, and it's just like people don't want to be in those spaces anymore.

Cuz it's like, oh, it's all filled up with people who like never left, and they start getting grumpy, etc. , etc. And it's just like, where are those places? But, what you want to really pay attention is like, how do I figure that out? Like, how do I make sure that this stays fresh? How do I make sure that like, people who might be creating rot will not like, like poison the rest of the house?

Yeah, to paraphrase, these kinds of businesses, churn kills you. And the founders never realized that they have high churn until, like, you may be like, oh, we have really great churn. Everyone always says that. But like, and you don't, you don't have enough data yet. But over the long run, sometimes you can, things can creep up on you.

And if all of a sudden you start to have high churn, and then you're on the hook for all these leases, like, these can get tricky. This is, a lot of the food delivery type startups, like there's a whole class of things that don't work. And they're like, we have positive margins, and we have low churn, and we have great, like they all say the same things.

But if you really dig into them, the, the secret hiding in their numbers that the founders don't want to admit to themselves is that actually, churn can get you. If you have hard costs like warehouses, not for you, but for other people. If you have all these hard costs, and then you have like some down months, like you have to pay your rent every month, you can't get out of it.

Like if community is everything. Yeah. Then the tools you want to think about also are like, the equivalent of what you might see in like moderation tools. Like what, like, we have to build all this stuff to work, work on communities at scale. How does that work for like, within homes and houses?

Like how, who do we have to empower within our community to make them sort of work and carry things onward? Lots of great real life communities that have grown to be really big, they have, they actually be really careful of like how they sort of set that up.

So there's going to be this delicate balance between, that you're actually going through right now between like, I need to achieve growth and revenue, but also need to make sure the community is like taken care of. And so like the moment, I think you choose this way above this, then then there's the sacrifice.

But the thing is you have to figure out like how do I get these to work in tandem in a way that doesn't break everything. And hopefully this trend, and this is the other thing I always worry about, is that this is a trend that just might break over time. Like just people start, cities and governments are building enough housing that basically people are able to find cheap housing on their own.

They just live by themselves. From your perspective, what do you think the biggest risk is? Like what do you, is it, is it, you need to hire operations people? You know what I'm saying? Like what do you think is the, like if you had, in the future, let's say it didn't work, and you're like, man, it was this reason. You had to guess. What do you think that is?

We definitely do run a risk that if the market crashes, rents could just crash. And we are locked into one to two year leases, which is why we try to restrict it to one to two year leases. Yeah. I think that's, that's a big one. Can you think about it? Yeah, I think, I would agree that that's the number one risk.

The other part is not being able to out, be able to continue to scale because operations cost are not able to be automated in a valid way. Yeah, and like for your own edification, I think a good thing to do would be to amortize all of your costs for operations and your time when you're calculating your margins.

This is a, this is a classic thing people do is like, we do food delivery and we have positive margins and it's, it's just that they sell, you know, a burrito for ten bucks and it costs them seven bucks to make it. But if you include delivery costs, if you include support costs, if you include returns, if you, like, if you actually bake in all this stuff, they're definitely not positive margins.

And so I think in your case, just for your own edification, try to bake in some of these costs you have from an operations basis. That you have your 80 hours. Yeah, you gotta put those into the, your margin calculation, you know, and that'll just help you know that you're actually building a great business, versus you're just like, leaving out some of the costs in your margin calculation. Yeah.

All right, we are out of time. Thank you so much, I had a fly spot. Thank you, appreciate it, thank you. Thank you. Thanks. Hi. Hi, how's it going? Good, how are you guys?

Good. Okay, so what's your name? My name is Osh. Hi. Hi, Osh. Shea. Hi, how's it going? Hi, Shea.

Good, nice to meet you. So we are Lambda Intelligence. I'm CEO. And we've built an API that allows developers to extract product information out of digitized receipts, so image, receipt images. And turns out it's, there's huge volumes of data that's locked up in these receipts. Okay, before you go into your back story, can you be really clear, so who's the customer?

So we're actually selling to mobile app developers. So we've got a mobile app developer who's, who we just signed a contract with. They're our first customer, but they're paying us $2,000 a month right now for processing, basically processing our data. For processing, basically product information out of receipt images.

Okay, so you guys offer an API to be able to take receipts and just give them structured information. Exactly. And then can you give me one more use case? Why does this developer need this? So the core piece here is not just the OCR, but it's also mapping the shortened product descriptions and the prices to the actual product, the full product name, description, brand.

All of that type of structured information. So, but what is their app to? So like, why would I want to download their app? What do I get to do with it? So, they're a loyalty car vendor. Yeah, so they're using this to better their app and better target themselves to their customers. So that's why they need this product information.

So, I always get nervous with API companies because basically APIs is like the bottom rung of infrastructure for software projects. And so, they tend to want to be commodities ultimately in the end. And so, the great API companies are going to be like, okay, if we're going to figure out how to do this as cheaply as possible, because this is going to eventually become a commodity.

Then, which ones are going after markets that are super, super huge? So that's why I'm very interested in who the customers are. You're like, mobile app developers that need to read receipts. How many of those are there in the ecosystem? That's actually a good question.

There are only 200 mobile app developers, but the second side of the business that we are focusing on is that brands are really interested, deeply interested in this type of data. So, we're in talks with one of the top brands, with Kraft, and they've expressed interest in getting this type of data from us. So, okay, so your real customer is CPG company. Exactly, yeah.

But the thing you need to, okay, so you have like a multi-step process. You're like, are you giving the API away for free? Why not just give it for free?

If the real customer is these giant CPG companies, why not just be like mobile developers who have a really great OCR receipt reading thing, have at it, and then because you have the best data collecting all that information, then you're able to actually close these enterprises. That's kind of our plan going forward.

Right now, we're able to get some revenue off of just providing this OCR API to our mobile app vendors. And so, that's kind of our way of bootstrapping. I think it'll just slow you down. The thing is, how bootstrapped are you? Are you going to run out of money? We are completely bootstrapped right now. Yeah, and right now, they're paying us $2,000 per month for processing their receipts.

This is on track to scale up to $15,000. Sorry, just like, you process the receipts. What are you pulling out of the product data? Why are they giving you, that's a lot of money for an app developer to give you. What is the thing that they can't build themselves that you can build, that's super useful to them?

So, the thing that we got is that we built a bunch of scrapers, and we've got some partnerships in work. So, are you scraping their email, or what are you scraping exactly? We're scraping receipt images, so people take photos. Okay, the user photographs a physical receipt, and then the app developer calls your API and passes it like a GIF.

And then you OCR it, and you're like, okay, this is what this thing says. Yep. Exactly. Okay. We structure it for them. Okay. So that it's not just the receipts, but the discounts. And they need that to do loyalty.

Okay, so I'm just making sure I really understand exactly what your API does. Okay. Yes. You get it? Yeah, I do get it. Just makes me nervous, because if your real customer is a CPG, it's like what's your real business model. How much data do you need to collect for craft to say we'll even look at this? So.

I know it's got to be a lot. They want a million receipts a month. Gotcha. So all the mobile app developers that potentially need this, how much? Is that much? because I feel like the people who mostly are going to track receipts are going to be people who are tracking expenses. It's the only app I use that I take a picture of my receipt. Nothing consumer related do I do that.

And there's loyalty stuff. But going along this line of thinking, so if your real customer is a CPG, and this is sort of, everyone can extrapolate from this in the audience. Maybe you should just be building the receipt tracking app that does loyalty. It seems like you're trying to use this API thing as a shortcut, and it's going to be hard.

But you're like, well, we don't really know how to get a million receipts, so let's just build an API, and other people will get receipts. And that kind of sounds like wishful thinking, unless. And again, it's great that you have this first customer, so congrats on that. But to get a million receipts, are they going to have a million receipts?

So they're on track, so they want us to do about half a million receipts right now. So they're getting us close, and so we'll need to work with a couple others. They're actually processing that many? They're getting that many. And so that's one of the reasons. Okay, along this line of thinking, and they're cool with us, right?

because sometimes what's tough with an API company is if you're doing something, and that it's really straightforward how you're monetizing it, you're kind of a middle man, and eventually people like to squeeze out middle men, right? And so maybe they're like, hey, this is really valuable. Maybe the app that you're selling the API to also wants to sell us the CPGs. Do you see the tension there?

Yeah, yeah. Yeah, we see that. And actually, we've done research, and CPGs are buying lots of data from third party companies like Oracle, DataLogix, Experian, and they're spending millions of dollars to get this type of data. And what's interesting is that this data is more aggregated in nature, that it's mostly transactional.

And what we are getting at is product level data that we are focusing on, which we know brands really want. So that makes sense, but I think just from a scale perspective. My understanding of what data brokers want for ad targeting is tens of millions or hundreds. The scale that you need to build a data business is much larger.

Have you seen anyone, even with millions, maybe for millions they give you a small. So the closest ones I've seen the people doing this is they're offering companies for free. So grocery stores, electronic stores, retailers, basically for free e-receipts. They say, we offer this technology, e-receipts for free. And then what they do is they take all the data.

And what they have is it's good for those customers, it's good for retailers because they're getting information they wouldn't have otherwise about their customer base. And then they resell the data as a data play, the CPG, and even bigger retailers. And so the whole thing is a race to the data. So if it's like, I closed this company, they're doing half a million receipts.

Well, okay, maybe this is easier than I think. Then what's the barrier for you closing more people? Right, and I think that's something that we're actively trying to go after more of these kind of data sources. And so the more data sources that we can get, obviously that improves the value of our- Yeah. Yeah, so how's that going? You said there's 200, so have you emailed all 200?

So we've emailed about 100 of them so far. And what have you learned? We've learned, so our response rate has been about 25% so far. And we've learned that the current mobile app vendor that's providing us this data, like them, others are also interested and they are willing to share their data because they are experiencing similar problems. They have all of these- You had 25 people respond to you.

Why are they not customers? That's actually the question I'm interested in. So we actually only launched our private data with this mobile app vendor about a week and a half ago. So we're still super early on this. And so we've got a list of meetings like next week.

And so one of the reasons why they're not customers yet is because our sales cycle is a little bit, I think, probably at least a month or two. It's still pretty good. Yeah. But I mean, these aren't really enterprise prices that you're charging. Right. I think for API, the biggest thing is, for any API, is you have such a long cycle because you've got to meet with someone on the CTO.

You have to have someone actually try the thing. Are people going to be invested? Is it meeting at the right time, right? So it's at the time that someone's developing the app and need to work on the feature is at the time that they need your product or service.

So it's really a matter of you need all cylinders going up once to have as many of those people try, because you don't know at what point, and all of a sudden, there's a proper nexus for the sale to actually close, or usage to actually happen. When you talk to this craft, how much were they willing to pay? So they were looking at $0. 20 a receipt. $0. 20 a receipt for a million receipts? Yeah.

Yeah. Okay. So, fascinating. So, one thing, this is just like a general purpose piece of advice. It's like, when you're dealing with large corporates, they employ lots of people in the business development role, and their job is to meet with people and to find innovative stuff all the time.

And for every deal that they actually do, and for every, you know, contract they execute, they probably meet with 100 startups. And so, we meet with a lot of startups all the time that have similar things where from their side of the table, you just ask for something crazy. Like, go get a million receipts, and then we'll look at it, and maybe we'll pay.

But, you know, I doubt they're saying that in writing, right? Did they send you a contract? No. Okay, well, so you can give people these impossible tasks just for fun, and see if they come back with it. But even if they do come back with it, that doesn't necessarily mean they're going to do the deal with you.

And so, we're always talking to startups that they believe the best case scenario, and it's so good to be optimistic, so I get why. But they believe the best case scenario that, you know, hey, this huge Fortune 500 company is going to do a deal with my two-person startup, and they're going to give us a million dollars.

And man, does that rarely, like usually, to pattern match here, usually you have to stair step up, and you start with smaller customers that say yes more quickly. And it's for smaller dollar amounts. And you'd have to do that for quite a while before you're able to get one of these deals. You can, sometimes people call them silver bullet deals.

You know, it's like the deal that makes your startup where you go from zero revenue to a million. Somehow, the universe makes those very, those things rarely happen. And so, I think in your case, one negotiation tactic here, maybe, is to be like, hey, we don't have a million receipts, but why don't we start with a pilot? We do have whatever. Half a million. How about that?

Let's start with that, and let's just get this going, and let's do our first data drop, and let's get this on paper.

And what you can do when you're negotiating with these kinds of people, is to try to, if they're saying, come back with this really hard thing, and then maybe I'll say yes, get, move them down to something you can do, something that's more minor, and then just try to get it in writing. And once you get something in writing, it's so much more likely they'll do the deal, because talk is cheap, man.

Like, anyone can say, oh, we'll give you a million dollars if you do whatever. But to get someone that works at a corporation to put something in paper, they're either coming up with an excuse why they can't, or they're actually serious. And think about how much better that would be for you as founders if you had something in writing.

Because then you know this is real, and that can give you so much more confidence that you're going down the right direction. So that's always something I would suggest. This goes for everybody, that when you're dealing with these big organizations, is try to get anything in writing. You have five minutes. The thing I don't like about this kind of code is inefficient.

Like, one, you're working on commodity level, data processing, it's being set up, and then. You have to do one thing before you can do the other thing, all right? And it's a two-step process. And so there's so many additional risks. So to me, it's always like, all right, if I'm trying to get this other thing, how do I make that go really, really fast to prove out this sort of nature?

So that's why my first instinct is like, you should just make this free. It's like, if your mobile developers really want this and everything comes down to you having as much data as possible, then you make sure that you get as much data as possible, as quickly as possible. And that means doing it and doing it at a scale that other people won't be able to sort of compete.

because I firmly believe down the line, this technology for OCR processing for receipts and stuff, that's going to be fairly common. Maybe this is a dumb question, but is there a reason why Expensify and all the receipt tracking things aren't just doing this? Why can't the CPGs get access to all these other receipt tracking data things?

I mean, what we've seen so far is that a lot of the accuracy levels in receipts is still pretty low. Interesting. And especially because you can get at the direct OCR, that's pretty good at this point. But it's going from that OCR kind of piece to an NLP piece that a lot of companies either are really good at. Do you actually need the SKU, is that what you're saying? Yeah, yeah.

Is that they can get pricing data, but they don't have the actual SKU to map it back? Exactly, that's exactly what it is. And it's one of those things, this is a problem that I actually solved when I was a co-founder at a healthcare company. We had lab reports that we actually had to match to individual SNOMED lab codes, and so I did that there. And so we were able to do it for 40 million patients.

That's great, so here's something to think about when you're at this stage. It's like, you think about, what are my strengths? What are the things that we can do really well that no one else can do? And what are the areas where we're on the same level playing field with everybody else? And so I feel like what I'm getting from this conversation is that a lot of your strength is in this technology.

And it is in the fact you can build this amazing OCR. And it sounds like you've been trying to find what the best possible market is. And so by doing a bunch of customer conversations, you ended up saying, hey, you know what, CPG's got money, and we can use this technology we already know we're good at. And so I would think of your core strength as being the technology.

On the CPG side, until you have your first thing in writing, that's still in the question mark. It's not not a strength, but you don't, it's not like in the bag in the way that, it sounds like you have this technology and it works today, right? That's in the bag, like you can take that to the bank.

And so I would think about, thinking about what your strengths are, and then maybe being open minded to how you monetize it. because it could be, until you get this right, until you lock in this risk, that this may not be the best possible market for this technology. I call it being promiscuous. Yeah, be open minded to ways you can use this.

Like other people may be like, hey, you don't need a million receipts. We'll give you money today for this other, like semi-related. Like it could be really close to what you're doing, or it could be a little farther away, we don't know. But you have something really special, and you have. How long ago did you talk to Kraft? So we talked to Kraft about three weeks ago.

Great, so you, and then is the deal closed with the half a million? No, we are in relatively early conversations, but they've. With Kraft. With Kraft, and that's with Kraft. No, no, no, I got you. The deal closed with the app developer. Yes, the deal is closed with the app developer. Did you actually have data for half a million?

Yes, we have. That's awesome. Okay, you should go immediately back to Kraft. It's like, we currently have. Let's do a pilot. Half a million. Right. Would you be interested in doing a pilot, sign a letter of intent to start looking at the data we provided to you.

You guys can check to see if, is it what you're looking for? because that will help you know that you're delivering something on time. And then have them sign a letter of intent. A letter of intent, all it's going to say is just like, okay, and then if you're going to initially look at this data, we'll work back and forth to see, is this what you're looking for?

This is going to be the format you want. And then when we have a million receipts, we'll start talking about pricing. because again, just put numbers on this. I would say, when a startup says we're in discussions with a customer, the close rate on that is like ten percent. I discount to zero. Let's just say, between zero and ten percent, it's very low.

Versus if you get an LOI, then I'd discount that to like 50%? because maybe even 60. What's that, 25? All right, I guess I'm more optimistic than Kevin here. But, and then once you have a signed contract, then obviously it's 100%. But like, we, from our side of the table, we've seen so many people that are in discussions with all these great signing deals.

And we're like, well, can you get into writing? And they're like, then they go try to get into writing, and they're like, nope. So, that's such a good test at this stage, and it doesn't cost you anything. Absolutely. Yeah.

The other piece of advice is, everything Don talked about in terms of like, what you have in the bank in terms of what you're good at, another way of thinking about it is that like, that is the default for your business. The default is the technology, the OCR, it has to work, and it has to work better than anything else.

You have to do this other thing, usually at a startup, that is going to actually determine whether you'll grow and be really successful. And that is going to be sales. And so, that is the part you have to work 10x harder on. So, this thing that you're not very good at, you actually have to spend, ironically, a lot more time on it to make your company better.

And so, yes, for the sales, you're going to talk to all the rest of those mobile developers, try to get them all in discussion, and close them to get the deal. But also, you've got craft going on. But also, you should talk to as many other CPG companies also. Well, let's see, it's like, is there a market beyond these guys? Is it similar sort of pricing? What do I need to understand?

What needs to be delivered? Because I'm actually basing everything that I'm going to give to CPG companies on what craft has said. So, it's really like, by the end of the next two months, it should be like, you've talked to, and figured out how to talk to all the other mobile developers, and you've talked to 50 other CPG companies at the very least. And that's almost all your time.

It's almost no programming or coding. It's just developing the language for, do I understand, what is the process for closing these deals, which is actually going to make me money? We're out of time. Guys, thank you so much. Thank you so much. Thanks, appreciate it. Thank you. Thank you.

Hey, how's it going? I'm Dalton. Hi, Kevin. Hi. Hi. Cool. So, we are working on Commaful, a place where people share and read short stories. We found a new format that's a little bit like a picture book.

So there's a little bit of text on each page, then there's a background image that relates to the text. And you tap through to flip through the pages of the story. People share anything from poetry, fiction, as well as random blog posts about what they're doing in their daily lives. Currently, we get about 100,000 monthly uniques. And mostly teenagers currently using this site.

How many stories are being created? 20,000 have been created total, about 1,500 a month. So let's say I'm one of your active users. What format do I consume this in? Is this a mobile app? Is this a website? Right now, it's all mobile web. So, okay, 100% mobile web.

Yes. Okay. And what's the breakdown of what devices they're on? It's about half and half on iOS and Android. Okay. And then what's the geolocation of those folks? What countries are they in? It's like, it's really lean towards iOS in America and the rest of the world in Android, so it's still 60-40.

Okay, so 60% iOS, US. Yes. Cool. And then what are the countries that are non-US? Philippines is a pretty popular one. Okay. UK, yeah. UK, India.

And then what would be the breakdown, say, of the Philippines on that one? Like 10, 20%? Like 15%. 15, okay. Who's your most popular storyteller? So there's a girl in LA, and she basically writes kind of fourth wall fiction stories. And she already had an audience before joining Commaful, and she brought a good chunk of her audience over. Gotcha, and so what does popular mean?

Does she have, you guys have a thing like subscribers? Yeah, so she has about 700 followers. So we have people who have more followers than her. However, she gets the most likes on her stories. She gets the most engagement. How often does she publish? Probably a couple times a month. Couple times a month, gotcha.

And sorry, just I want to really picture it. So, and then when she publishes, what is her, you said there's a few things people can make. What is the thing that she makes? Fiction stories. How long are they usually? They take about three minutes to read. They're about 1,000 words. Cool.

Would you say there's any similarity between this and some of the apps that are taking off right now that are collaborative fiction? What's the name of one of them? There's like three. Yeah, yeah, yeah. I've seen those. Is there, is this, if you squint at it at all, does it look like any of those? It's a very different model. Those are built as games.

So there's a subscription model built into them. However, we have very similar users. So many of the target users use both, or if they're conscious about the money, they use Commaful. So that's kind of the similar audience, but different play. So do you have in-app purchases? Do you have anything analogous to that? Currently, no. Got it.

I don't have a mobile app yet, so that's. Okay, sorry. I'm just trying to think of how those companies work. Do you see this as, like how do you see, what's the big version of this? Like, what's the metaphor that you think of? Metaphor wise, we see this as kind of how these teens interact with the written word.

So you see Instagram and Snapchat as these very, very visual creatures, very short form types of content. However, the written word really hasn't seen much change. And as a result, a lot of the stories that you can only tell through the written word that you can't tell through a Snapchat or an Instagram kind of get lost.

And for us, we want to be the bridge for this new generation to the written word and those types of stories. These stories are like permanent, so they exist in a space. Are there some stories from like a year ago that are still like super popular as per the permanence question? Yeah. They still get hits. I wouldn't say they blow up or they get us consistent huge amounts of traffic.

But it shows up on search sometimes. Sometimes it still resurfaces in sharing and Twitter. Do you have a sense of what percentage of your users come from SEO? Very few. We get like 700 clicks, I think, per month. What are the most consistent sources of traffic? Like when you said it's the authors promote it, it's that they post it to Twitter and Facebook, and then that drives traffic?

It's pretty much all of them. This is what success looks like to me in this sort of space. It really comes down to is you build a platform that allows people to get famous. That's pretty much it. Yeah, it is. Because I think there's like three drivers, man, three drivers of desires by most people. And it's like money, then sex, and then power or influence, right? Watch, it's only three.

Right, three. And so that influence is the biggest thing. So does your platform allow people to get really popular? Do they generate stuff that lots of people want to see? And until you hit that, the thing's dead in the water. So this really, it's like, what do I do to bring and on board the very best storytellers, the very best people?

And ideally, it's like, you guys are constantly generating these stories, and they have to be really, really great. But also, it's like, hopefully, your community is generating stuff. And then, are you able to promote them better than they can even promote themselves, like something has to happen there, and that's like your only goal. Like, you can try to monetize, etc.

, but I think it really comes down to is like, is this a platform where it's like, if I use this, it's like, or I want to tell stories on here because people actually read or experience them. It's like, similar to what you saw for Medium, the whole thing is like, what do you do next? I don't know, that becomes a really hard part.

Sometimes you end up raising a bunch of money and it doesn't work out. But the whole thing is you have to at least get to the point where it's a platform that gets people really popular. So everything has to do with looking at what is somewhat popular and saying, how do I multiply that, how do I replicate that?

Is there a viral portion of that where people see those and they start making things for themselves? So it's like, I would go and just figure out, how do I get, let me see here. Best story I heard from within the YC community was Twitch. Basically, they launched a streaming platform, and what they tried to do was, some people were already streaming gaming on YouTube.

They contacted all the top streamers on there and said, what do we have to do to get you to come to our platform? And they did whatever they wanted, because they knew that it's a power law curve for content, right? So only a handful of people are going to be the most popular. And so that gives you the best head start. And so they built whatever features, they super nice them, etc.

And so where are the very best storytellers for your sort of format? And I would think it's creative people who have been really good at Snapchat or Instagram or whatever. And say, look, you want more than this, we can offer this to you. And you have to figure out, what do they most want? And then build those sort of features. Have you started doing that?

Yes, so the last while we've been, we know a lot of top creators and that that are able. What is a top creator? So someone who has 10,000, 50,000, 100,000 followers. No, I want you to find people that are the top ones. I see. You have to talk to them and figure out what would it actually take. So like the best selling authors, the million followers plus.

Could you start pulling people off Tumblr, because Tumblr's. That's actually most of our traffic. So when we say social, it's the bulk of our traffic. 80% is from Twitter. Because Tumblr, Yahoo's not really taken care of. And I imagine you can pull the top Tumblr people, right? Yep, so we actually are friends with almost all the top Tumblr creators. What we've been thinking about is critical mass.

And we think that there's a certain point, like we want to spike everything with all of the users coming together at once. Where you hit this critical mass point, and theoretically, supply meets demand. I would not even think of it like that. That's just too complicated. Mostly, it's just like, all the popular kids, and I'm just going to systematically talk to all of them.

So just land, grab, get as many popular people. As many as you can, because what you're looking for is inspiration, and you don't know what it's going to be. You don't know who is going to be the one that makes the platform take off. You don't know what is the piece that they're going to make. So you need just a lot of lightning striking.

So hopefully, people go like, holy crap, what's this new platform? I agree with that. I mean, look, I think that founders that are working on consumer internet stuff, they undercount how many social things get to your scale, and then top out. Like, they think everyone knows the story of Snapchat and Facebook, and they know about things like that.

But man, even back in the day, there were a lot of Facebook competitors. Like, hundreds. And a lot of them had millions of users, right? Tens of millions of users. And- It's not meaningful. And the thing that's tricky about consumer internet, if you're not going to monetize and charging your users, is the only way you can ever build an ad business, is you need hundreds of millions.

And so, I think in your, like, think about Tumblr itself. They had trouble with their ad business. But remember, they were the winner. For every Tumblr, there were like ten other little blog sites, which it sounds like you guys are experts. So you remember them, right? There were other wannabe Tumblrs that had decent, you know, even LiveJournal itself.

LiveJournal was the original Tumblr, but like that, and so the point I'm making is I agree with Kevin. If you're going to do this kind of business, you usually want it to grow as fast as you can, and get those economies of scale. Otherwise, you can get stuck in the 100,000 to million user range, and then like top out, and then what?

Like those are, I know the founders of a lot of those types of ones, and those are rough, right? I feel like you want both a bunch of users, and then a bunch of different stories. Like, I think, again, be permissive. It's like, what are the top things that people have done on Tumblr, recreating it on your platform? Like top tweet stories, right?

That's like series of small pieces of text, and then tell them on your platform. And as per his Twitch thing, just ask them, what would it take for you to move everything over from Tumblr? Like, what features do we need to build for you to take everything from your Tumblr over here, and then put up a Adios Tumblr blog post, and then like point to your site?

And then if you just did that systematically, that was the Twitch playbook. I think that would work. I don't think, I think now is the time to raid the Tumblr users, yeah. Yeah. One of the most popular things we have is that it's a two directional thing. So people take content from Tumblr and then post it on Commaful, and we have an export function.

Every social site has a different preferred format. So on Tumblr we export as a photo set. So if you share on Tumblr, it automatically generates a photo set that's pretty and fits on Tumblr. Facebook is different, Twitter is different. You should also look at all the people making the little stories on Medium, too. Exactly. Still all of them as well. We also have edible stories, too.

So, the whole thing is, figure out, talk to the top storytellers, talk to the top people who are on all those different platforms, figure out what it's going to take to bring them over. Yeah. Or start moving some of that stuff over automatically and just showing it off to your community. Right. And seeing what sticks, what people really like, what goes around. Because, yeah, consumer is different.

I think a lot of times people out in the audience world, they hear these consumer things, which is just get big, get big, get big. And that is what we're telling you. But just for everyone else in the audience, that's not the case if you're doing B2B. This is not one size fits all advice, but if you're going to, hey, you're launched, you're doing consumer internet, you have 100,000 MAU.

Like, okay, you made it this far, you've got to power through here. Or else, if you end up stalling out. Yeah, you're dead in the water. Then what? You don't have enough users to really monetize. And we've just seen a lot of people stall out at this stage. So, in your case, yeah, you've got to grow. Yeah, we are in the problem solving position here.

Yeah. Well, you need to generate whatever it takes to get people, the word of mouth about your thing is like, if I'm a really. I'm a good creator, I tell the stories on this platform, I can get really big. And one other piece of tactical advice, I would look at SEO. Usually, like the way you get really huge, I built a consumer site. I got it up to 25 million MAU, and SEO was a big part of that.

And the way you do something like Quora, where you get hundreds of millions of MAU, is SEO. So I would try to start tracking that. We've heard that embedding helps us get backlinks. Okay. And help us rank some of those stories that normally wouldn't rank. We're approaching SEO currently right now, yeah. Okay, good.

Because that's the sort of thing, it doesn't cost you anything, and that'll just get the flywheel spinning for you to hit that escape velocity. Five minutes. Any questions for us? I think the biggest challenge for us, obviously, is getting from our 100,000 to a million, right? Like, that's the next big phase. You've already kind of gone over different strategies.

I'm curious what other companies that you've seen through going through YC, who have successfully gone from that 100,000 to a million. It's still parallel. It's always parallel for those kind of communities. So the top ones bring almost all the traffic. Which is why I'm saying, everything you have to do is, how do I bring on celebrities onto the platform? Or I turn people into celebrities.

Well, and along those lines, here's the pitfall. Don't pay them. Right. Don't pay them. Sometimes, I've seen people try that. That is not, like, paying for traffic when you don't have revenue is like. You need to love it. And once you pay them, they also expect to be paid to make something.

That is exactly right. Like, right? Like, as far as, my understanding is, you know, Instagram never paid anyone. Right. Snapchat never paid anyone. All their clones sure did, right? Like, but, that's, you know what I'm saying? Like, like, YouTube wasn't really paying for traffic.

But all the wannabe YouTubes that were trying to raise money were paying for traffic. So like, whatever you try, don't just take what he's saying and be like, yeah, we're just going to pay some celebrities and, you know. Right, right. It's not a cost effective way of growing. Yeah, yep. I think, if you have the dev cycles, I don't know if you do, but I would consider a mobile app.

One advantage from a retention perspective is that you can do push notifications. Right. And that way, if someone is subscribing to something and you do a push notification to their followers, then you're ensuring a much more, a tighter feedback loop than maybe you have now on mobile web. Definitely. So how's your email, what are you doing with email, Joe?

Because they're teens, email has been pretty tricky in terms of getting the response rates. We tried messenger bots, so we get some decent response rate there. Still nothing huge or significant. I would keep explaining with those. Again, you're asking what have people done. You have to reach out and touch people for them to not forget your thing exists.

Like think about it, there's a million apps out there. There's a million, everything is competing for our attention. And if you think about tools like Facebook, it's basically a weaponized attention grabber, right? They have thousands of people working on making sure that you're constantly looking back at it, right? Very smart people, that's all they do all day, is try to steal your attention.

And you're competing with them for attention. So I think just doing some really basic stuff, like continuing to iterate on email, if chat is the right thing to do. Make sure people remember you exist, and that they are a member of your community, and that they should go check out some new stories. Otherwise, it's going to be tricky.

One experiment you could do, I've heard someone give this talk before, is go create a Facebook account with a fake email, and don't add any friends, and look at how many emails it sends you. I think it sends you like 50, I'm exaggerating. But the point is, they realize that if you don't add any friends, that you're going to have a bad time, and that you're useless to them.

And so they're very aggressive about emailing people. Otherwise, they know they've churned you. So anyway, if you have users that you know are going to churn, or on the way past churning, well, just remind them that you exist. That's best practices, too. Great. Yeah. Guys, thank you so much. Appreciate it.

Best of luck. Thanks. Okay, cool. Well, I think that's it for us.