YC SUS: Eric Migicovsky hosts founder office hours
All right, hi everyone! Um, my name is Eric Mikowski. I'm the course facilitator here at Startup School. I'm excited to do another round of online office hours with companies in Startup School.
So office hours, um, for those that might be tuning in for the first time, are an opportunity, um, that we take at YC for the partners to actually spend a concentrated, a focused 20 to 30 minutes, um, with each startup, really to dive in and talk about a particular problem or a particular situation the founder is experiencing, um, and help them brainstorm ways to solve that problem or figure out ways that they can work through and, um, and get past that problem.
So we're going to be doing that in an accelerated fashion here. We've got a bunch of Startup School companies that have signed up through the Startup School forums to, um, to do office hours, and we will do them in front of everyone who happens to be watching this.
Um, as I talk to the founders, um, I'm really going to try to keep my advice kind of, um, focused in such a way that even if, even if you're tuning in and you don't have the exact same problem, hopefully you can still get value from, um, from these sessions.
Really in general, like one of the reasons why I like doing office hours is oftentimes we as founders are really like stuck in our own world. We're working from home, we're working on a small team, working on a project, and sometimes we just like spin our wheels trying to solve problems. Um, with office hours, it's a great opportunity to kind of like talk through your problems with someone else and potentially even figure out ways internally that you can solve them just through the act of like saying them out loud and sharing them with someone else.
Um, so as you're watching this, if you're, um, if you have friends that are also working on startups, consider that maybe at some point you should just take the time to talk founder to founder through the problems that you're having. Um, might be a way to solve.
So without any further ado, let me pull in the first person. Um, first up is Boxcore. JC, this is JC. JC, can you hear me?
Hi, good to meet you! Let me just, me okay, here we go. Um, how are you?
Uh, good! How are you?
Good to meet you! This is, uh, Jose from Boxcore. Great to meet you. Why don't, um, why don't we jump right in and start by, uh, what's the name of your company and what are you working on?
Uh, the name of the company's Boxcore. Uh, what I'm working on or what we have is we have a beta that's live. What it is, is a futures gaming platform. So what it really is, is we ask trivia about future, uh, events that are televised. Uh, so mostly sporting events, but we also do entertainment like the Oscars and stuff like that. And people just answer what's gonna happen at the event, you know, who's going to win the match, who's gonna win best picture, all those types of things.
I don't completely understand from what you're doing. Let me ask a few questions. Is it a, is it a betting or a gambling site?
That's where we're going to end up, but we're not monetizing it yet. But yes, um, and you're gonna play it. Instead of playing it against the house, you play it against your friends, you play it in groups against your buddies, you know?
Okay, um, and what stage are you at right now with the company?
We have the game live. Uh, we had 240 plus users last month for the NFL playoffs. Um, we, it's a beta what we have right now and we need to basically build the next version now. Um, we believe we got proof of concept at least to ourselves; we prove it to ourselves like, uh, that it can be a second screen during a live event.
Uh, so for example, for the Oscars, we just had a lot of people reaching out and being like, "Hey, like you're late to scoring this or that." Like it just, we felt the energy, you know?
And like, and so we need to...
And how many users do you have right now?
Um, we have about 20 in the latest competition, active users right now. But the truth is, is I've been doing the whole thing by myself. And so I've brought down the amount of content we have, so basically no trivia is really. Um, and I'm building a team, uh, trying to scale.
What's your website? I'll take a look at the site.
Uh, boxcore.dolive. Um, so yeah, uh, we have the competition, but there's two other legs that are very important that we need to add to this product that aren't there yet. And one is, uh, utility to the user, and that would be in the form of live scores, previews, recaps, all that type of stuff.
Um, why do you think your users need that? Like, those are available on other sporting platforms?
That is correct, but, uh, we think we can do it better than the other sporting platforms because the key is that users are giving us in real time their preferences. Uh, when you, like, if we give you a catalog of like 10 different sporting events that are happening this week, you're only going to pick the ones that you're interested in. And so because of that, we know in real time what you're following. If you go to ESPN.com and you put in your preferences, you're gonna put in who's my favorite team, but your favorite team may not be the most relevant story, uh, for the following weekend.
Does that make sense?
I get it, but I think like what I would be pushing to do, especially in the early days for a company like this, is really to test what is the smallest thing that users actually want that they can't get out of the existing apps. Like right now, um, if you're a sports fan, there's a multitude of different apps that you can use. Um, but presumably if you want trivia about a specific team or a specific sport or specific type of sport, maybe that's missing.
So have you tested like carving off just a very tiny portion of what users want and just delivering that, rather than trying to do every sport, every, um, every feature?
100%! And that's, that's what we did, you know? That's what we have live. We have live, it's just a trivia competition of future events that you play against your friends, you know?
And we've had people that have done betting on the by themselves. For example, the Super Bowl, we had a group of 10 people, one guy brought nine people to play against him.
Um, but like even on your website, you've got, uh, the Grand Prix, you've got the World Cup, you're talking about the Super Bowl. I see some like, uh, other World Cup. Like why not just pick one specific sport to, um, uh, to focus on?
Because that's not how sports fans behave. Sports fans, um, and I say from experience, don't tend to follow just one sport. They follow multiple sports year round, you know?
Um, I don't think what what I'm trying to say is like even if they follow multiple sports, if this is truly a problem that they can't solve anywhere else, wouldn't they want to solve it just for one sport for one team at one time to start with?
Um, like couldn't you make it go viral within just the community of like one particular football team or one particular cricket team and just get everyone who's like interested in that team using it to start?
Yeah, no, it's possible, and let me think about it a little bit more because basically what we're acquiring users, or we're working on the user acquisition is in Mexico. So it could be, uh, soccer and MX. Now, the way I've been thinking about it and, and, and sound like you may not be on, on the same track, but the way I've been thinking about it is that, um, for example, the Mexican soccer fan follows Liga MX, but he also follows Champions League. He also follows some of the European leagues, you know?
And so, um, what we, the way we're approaching it is more like, hey, let us give you, um, all the different things that you follow that—like most, like the way it works is domestic players will give you these type of competitions, but they only focus on domestic or on one particular thing. And so we're trying to kind of break down the barriers and give you the—if there's many apps that are already solving this for local teams, international teams, etc., why do you think there needs to be one app that has all of them? Why can't people just have three apps?
Because, well, one is because you just said three apps is very uncomfortable or not optimal—not uncomfortable, but I agree. But I think it's mostly about, hey, I want a product that gives me—that becomes my second screen for the sports that I follow, not just a subset of the sports that I follow, but the all the things I follow, you see what I'm saying?
So second screen, your second screen is it would be that like you are watching the event and at the same time you're following live how this—how you're faring against your friends or how the stats or how things are, are kind of evolving during the match, right?
And so I'm thinking, the way I'm approaching it is I'm thinking about the users themselves. And so like I know a lot of people or like the people that I've pulled in that like, hey, we want more content of, for example, the Real Madrid Barcelona, but we don't want all the games happening in La Liga this week, that type of thing.
And so tell me, tell me what are the problems that you're struggling with right now?
Uh, the problem that I'm struggling with the most is, uh, I'm not a technical founder, uh, but learned how to code, and I basically launched the beta that got over 200 users.
Uh, congratulations, that's awesome!
Thanks! Yeah, so I learned how to do this. I started in June. Um, and so, but I basically hit a wall at the end of January, and that's what I told you right now, like there's not a lot of many users. I hit a wall, and I realized like, okay, this is as far as I'll go by myself, you know? Or like I'm not gonna be able to do a lot better. Like maybe I can, but I don't have the time, you know?
Um, so I need to scale. I need to get people. And so, uh, so someone had been offering to give me money for a while, so, uh, not that much, but 50k, but like, anyway, I'm going to take that money. Uh, maybe I'll go raise some more, and I started hiring people, and so I went back to my alma mater in Mexico, and I'm getting engineering students for $5 an hour.
Um, and one guy is already working out, there's a couple more that may work out as well. Um, and so my problem is how do I structure the whole thing to be ready?
Uh, to your point of one event, one thing: Focus on that, um, Euro 2020 and the Olympics happening this summer; and so I kind of want to like reposition, rear the whole thing.
My recommendation is, that sounds like it's a long way off. Um, the Olympics aren't until August; I don't know when Euro is, but my recommendation would be don't wait that long. You already have 20 users who already started using it, and I would start talking to them and seeing how could you build features that cause them to—for this to become like an addictive app—something that they use on a regular basis.
And then iterate on a weekly cycle, saying, okay, well, I talked to users; they have these problems that are holding them back from becoming active users. We have the engineering team try to make, uh, try to make batches of work that are less than one week.
So this is like a really important strategy. Oftentimes people think in these really long-term future dates; they're like, okay, I want to have the app ready by—yep! In the problem with that is the feedback cycle for your engineers and for the whole team is really long because it's February; it'll take you a couple months to get the app ready, then you'll spend some time getting the launch ready, and then you'll launch in like July, you know, before the Olympics. That's way too long.
A much faster, a much better cycle would be to do this on a fast basis, where every week you sit down with your engineers, and you say, here are the things that I'm hearing from users; here are the things that I think we should fix; and then here's how we're going to test to see if the fixes that we just built are actually useful!
Yep, you do that every week! Oftentimes I talk to founders that set up like one meeting on Sundays for maybe a couple hours between them and their co-founders, and they actually use that time on Sunday to plan out what they're going to do for the following week and to measure kind of the results from the last week's work.
Yep, no, that makes sense, and that's actually how I have been doing it since the launch of the beta in September until I burned out last month. Um, I've had users that have given me over 500 predictions multiples. I've had and I speak to a lot of them regularly, and I've had about 300-plus signups, you know? And I think half of it was organic, half of it was like a little promo we ran.
But anyway, um, well, Jose, really glad to get the chance to chat. Um, I hope to hear some updates and hopefully not all the way to till, uh, till August.
I agree! I think that makes sense!
Cool, thank you. I've got to move on to the next one, but it was really great meeting you, Jose. Thanks!
Sounds good! Bye-bye!
Bye!
Okay, next up is Core Care. Um, do you see them, John?
Nope, maybe not.
Okay, next up is Mary from Parley.
Oh, that's IQ, IQ, cool.
Uh, I think unmute your audio.
Hi! Um, Mary, I think you need to unmute your audio.
How's that?
I can hear you! Yay!
How are you doing, Eric?
I'm great! It's good to meet you!
Nice to meet you too!
Um, very excited about talking with you.
Yeah, we're actually building a hard tech company; I guess I saw somebody say, hey, there needs to be more hard tech companies.
Um, what we're doing is essentially turning your smartphone into smart binoculars. Okay?
Allowing users to zoom in, view, capture, and live stream events. Imagine if you're up in the nosebleed section and being able to zoom in on the game and certain plays, uh, be able to record, snap photos, share on, on, uh, YouTube or Facebook Live.
Um, we're at a point where we've, um, developed this product; it uses your smartphone.
Okay, okay.
And we have a lot—like the, um, the magical binoculars in Harry Potter?
Yes! Yeah, exactly! You got it!
Yeah, we actually were inspired by Star Wars electrical binoculars; that's what they call it.
We—um, so it's kind of like you use your camera on the phone, and then it magnifies it.
Okay, exactly! Yeah!
And it—it's a new category in consumer tech, which is why I wanted to talk with you, because you brought the first smartwatch to the market. And so, um, would love to hear some of your advice.
What stage are you at right now? I see that you have the prototype; maybe show it off a little bit?
Sure, sure!
Um, so, uh—black? White—pardon—do you have any customers yet?
Yeah, we did a Kickstarter campaign. Um, we had very little funding, but we did raise our goal of, um, uh, 27 thousand. Now we're up to 30—but thank you—$30,000.
Um, and we've raised half of a seed round. Um, we should be closing that by the end of—
Okay, so, when will you start shipping them to customers?
Before for, um, the, or by the 2020 Olympics. We have our tooling already done.
Okay, uh, hang on, just got something.
Um, yeah! We've already done our tooling, our first article, and um, we, one of the obstacles I see right now, we're already to bring these the first batch, the first um, pilot—what we're calling it—the pilot production to ship to our, uh, 200-plus backers, and also have enough to seed the market. Um, we have major leagues interested as well, as—
Okay, so, so, um, as of right now, so you have some early customers that ordered on Kickstarter; you're getting ready to ship them.
Um, do you know where your customers live, or have you talked to any of these customers?
Yes, they're all over the world! Um, we, we opened it up, and so based on our crowdfunding campaign—
Um, why are you waiting till the Olympics to start shipping if your users are all around the world?
Oh, it's just a timeline in July; rather than say, July, we're saying, you know, we'll be there in time for the 2020 Olympics.
But, um, one of the... um, if I understand it correctly, it's a plastic holder for your phone with a lens.
Like so we have a proprietary touchpad on top, yeah, maps to an app, and you're able to control your smartphone camera functions, um, with our touchpad.
Okay, we have an app in the App Store, um, in the iOS app store, uh, called IQ Space Live.
Um, and it's, I like your eyeball, not the letter, I. And so what's the biggest problem that your startup has right now?
Well, what I wanted to ask of you is given that you brought in a new product, a new category as well, if you were bringing that to market right now, um, just what would be some of your advice you would give us?
Considering the coronavirus is causing issues in supply chains, and so forth.
Your question is like around the supply chain?
How do you—?
Um, we're in—we're on in the crossroads right now where we're ready to manufacture, we're talking with factories, we're trying to mitigate risk, um, for manufacturing in China.
Um, but our tools are there, and we certainly could get the first plastics out. Is there a reason why you need to build plastic injection molded parts? Is it possible to do this with 3D printed parts instead?
Well, we've already— we've already done that for the first—
Yeah, but like oftentimes one thing that I—as I talk to founders, people sometimes skip ahead to the plastic injection molding step way too fast. Oftentimes end customers are totally fine with 3D printed, um, products, even if they're buying it off Kickstarter. In fact, we've had several YC companies go through to the point where they were making thousands of units, 3D printed, completely, just by themselves off 3D printers, and actually selling them to customers.
That would be one way of eliminating the need to actually, um, get parts from factories in China because you just have them coming off of your 3D printer.
Yeah, we sold the products at a reasonable price—we're retailing these for $199—and of course with Kickstarter, you always have to do a little discount there to get, you know, the interest going.
So, um, I'm not sure that we could do the 3D printed ones at scale for the price.
There's a—you—like I mean, you, not—it's not just the plastics that we're concerned about. We just—the overall, um, so one thing that I would do if you—you have about 300 orders, right?
Yeah, yeah!
So if you have like something like 300 or less than a thousand, one thing that you can do is just place the orders now for all of the components since you've got the money from Kickstarter; you kind of have an idea. So that you can just like put it out of your mind.
Yes, it will take a couple months for the things to ship, but if you just place the orders and you work with the factory to make sure that they’re like, you know, they—like one thing I like to do with factories is just, you know, let them know every week that I exist.
Like remind them over WeChat that you need units.
Um, but in your case, with 300 units, I don't think you're going to have much problem; I think that'll be all fine.
Well, we have— we have more! I mean that's just what the pre-orders were; we have a lot of interest for—
Yeah, but I would focus on building those pre-orders first, right?
One that I've noticed a lot of other hardware companies, they kind of get ahead of themselves; they think, oh, I've got 300 orders, that means I should really build a thousand units. Um, because it's cheaper to build several thousand units than just 300.
Um, but this is actually like a very potentially dangerous mistake, um, because first of all you want to get your 300 units out to the Kickstarter backers so that they're happy and that they tell their friends. So if you start trying to plan for like thousands of units, it may actually delay the first 300 to those customers.
And so, um, if it's at all possible, try to build just 300 or 400 or 500, some small number of units to start so that you can actually ship those to the end customers as soon as possible.
Yeah, great!
Um, and then, uh, if I can just ask another couple questions, um, what would you say would be the most important hire once you got your funding and so forth?
Um, in the early days, uh, especially before companies have hit product-market fit, some of the best—some of the best companies are run extremely small. Um, usually it's just the founders at that point.
Do you have any co-founders on your team?
Yes, yeah! There's, um, our CTO, uh, who's a mechanical and optical engineer.
That sounds very cool!
Yeah, so he's—it was his invention, and um, we also have some folks in the sports community. We have a software engineer who's building out our AR app, so that you'd be able to see like the yellow line when you're at a football game!
And that sounds great! I mean, sounds like you got the whole team right there!
Um, like oftentimes, like I really try to encourage founders not to hire too many people because you know, the smaller your burn rate the easier it is for you to become break even and even profitable. Um, every additional person that you hire is going to be is going to make it harder and harder for you to kind of break even above that point.
So you're saying, um, what we have is is good for now?
If you can make it happen with the people that you have, then I would highly encourage you to do that.
Yeah! Sometimes it may be necessary to hire particular contractors or people with specific expertise to help get over a particular problem. This is especially true for hardware companies, where you may need an electrical engineer for two weeks, and then you may need—
Right! But it's not necessary to hire them as a full-time person. You just won't have enough work for them.
Right! Right! No, that's good advice! Very good advice!
Um, and, uh, what else can I help with? What's it—speak up to our CTO here too?
Yeah, hello!
Yeah, no, I just wondering, uh, if you might have any advice about bringing in, uh, interest from, uh, developers to, uh, you know, to, to develop on a platform as a category?
Yeah, certainly do! So the qu—yeah, so my recommendation here would be to focus on shipping your units to early customers as—as, like, when in doubt, that is always the right move to do.
Because this is really interesting! Within that group of 300 people that have bought it off Kickstarter, there may be actual developers who are interested in hacking and building on top of the product that they already bought.
So the—the first developers that will be part of your community may actually already be there!
Okay! In your user base! But you won't know until, um, one thing you could do now is you can email the 300 people; I'm sure you get the email list from—from Kickstarter—you can ask them like, hey, is there anyone here who's interested in hacking or building on this product?
Um, that would be the first step. And then the second thing is once you ship it to them, that's when people's mind start thinking about what are the cool things that you can do with it, and they may end up just hacking on it themselves!
Especially since it’s just like their phone fitting into this, right?
Right! Right! In fact, um, just your—the previous developer that you were just talking with would be like an example of a, uh, app that our product—
And we've—I mean, so that's actually one of the uses that we've looked at is real time, you know, so forth.
Well, it's great to meet you both! Um, I wish you good luck with the Kickstarter! I'm pulling for you and hope to see it in a couple months!
What's the name of the product?
IQ!
IQ! Got it! Thanks very much!
Thank you!
Bye!
Okay, next up is Trackport.
Jake! Track—actually, did we get the other one come in? Core Care?
No! Track Point, Joan! Do you see Track—
Jake! I’m missing at the top!
Okay! Jake! Hey, how are you doing?
This is John.
Jake and John, um, great to meet you!
Great to meet you too!
Um, so we're here with Cargo Labs Incorporated. Uh, it's a fintech company for truckers, and we—our first product launched in December; it was—it’s called Trackport. It's a mobile app that tracks how long trucks spend at a ship or receiver while the truck's getting loaded and unloaded.
So it—it tracks the truck while it's like an app that runs on the phone?
Yes!
Okay! And what's the problem? Like, I'm not—I’m not a trucker. I don't have that much experience in this space. Tell me what—like what's the problem that people are trying to solve with your app?
Yeah, the problem is that trucks spend a long time at shippers and receivers, uh, waiting to get loaded and unloaded. And many times, not only is it not clear to the brokers that are paying them this money that the times are not clear, but also they have a difficult time invoicing for those rates because—
Okay, so you just said you just said a mouthful, and I’ll let you know that like as someone who's not really familiar with this, with this space, I'm like, I'm still working to try to understand who this is for; like who the primary user of this app is.
Yeah! What is the problem that they have, and then how does your app solve that problem?
So maybe you could start, like, is this for the trucker? Is this for the owner of the trucks primarily?
It's going to start for the, uh, the trucker and the trucking company to get paid for the time that they wait at that ship or receiver inefficiently because there's times when the loading process and unloading process moves very inefficiently, and they wait there for an excessive period of time.
Uh, our app calculates how long they were there, uh, give also with a given detention rate. This problem's called the detention pay problem because they're stuck in detention when they're at the shipper and receiver; so our app calculates how long they're there with the given detention rate and then sends off an invoice for the amount of money that the truckers owed.
Okay! So it's—it's an app for the trucker; they run it on their own phone and then they set a—do you have to—
Like, if I understand, it's like the trucker arrives at a Walmart Distribution Center; they wait there for like half an hour or an hour or multiple hours to get loaded, and right now, they're not getting paid for that time?
Yes, absolutely!
It's—they have a hard time getting paid for it; it's not always clear because there’s not a good system. Is it part— is it part of their existing contract?
Uh, it’s usually outside of the contract, typically, because it can't be predicted. So it's—it’s a secondary contract, uh, typically. But this is all done right now verbally; this is all a gentleman's agreement through the phone, and, um—
Cool! So what's the next—what's, uh, what stage are you at? Have you already launched the app and do you have people using it?
Yeah, so we have seven or eight truckers using it so far. We launched in December, um, and we just got, uh, I think Summerfield Trucking—they put us on their website last night. He also runs an accounting firm for 25 other trucking companies.
So tell me about the seven truckers that you're working with.
Um, do they use it for every job that they do today?
They use it typically for areas that they know they're going to have problems at. So not all places that they pick up product have problems, but many places do.
How many times a week do those truckers have problems?
So it's typically a really—a trucker that encounters this a lot will have it two or three times a week, actually.
I'm not asking hypothetically, like for those seven truckers, you probably know them by name, right?
Yes!
Like how many times did they—last week?
Yeah, so like two to three times a week typically in a five day—
So you had 14 stops tested?
How many—how many, um, I guess, like track?
Five to seven orders—not all of them have been using it consistently; we've been working out the bugs for the last couple weeks.
Work—do you have like the phone numbers for all of those seven truckers?
Yeah, or phone or email; we have contact info.
You—you like have you tried phoning them all up?
Yeah, we give them calls. So one guy's name is RL; he's an owner operator, so he runs his own trucking company and drives for it.
He's been our most helpful—
So cool! That's many times!
Yeah, yeah! Great!
So what's the big problem? Like, what are you—what are you guys working on right now?
Um, so we're working on actually having a cryptocurrency processing system where we have a dollar-backed cryptocurrency that Pro, that, uh, makes the system work under the hood. So we can move the money from the broker to the trucker instantly with a smart contract that releases it as soon as the trucker finishes his order, if that makes sense.
To pretty much automate the payment system for between the people buying the trucks and the trucks themselves.
And the reason why that's important is that sometimes it can take about a month for that payment to process, so if we can get that processing time down to a second by working with both the broker and the trucker, uh, we think there's a lot of value there.
Is that a critical component to this?
Like, you managed to explain your whole startup; I understood what you were doing. And it sounds like there's a problem here that truckers—and you introduce this whole other concept of—like let me try and bridge that a little bit better.
So right now, my major question is like why don't people just use US dollars?
That's what we want to do, but we’re gonna have a digital way to interact with those dollars.
We're not sure, but, like, correct me if I'm wrong—these truckers are already getting paid for the truckloads that they move, right, in 30 days?
Yes, yeah!
Like, they're already getting paid; that's like a very known, you know, method for how they get paid or whatever.
Um, I thought the problem that you're trying to solve is they're currently getting paid this much, but they're missing out on like a couple hours' worth of extra pay because of the time of just being at the dock, right?
Yes, and that's the primary thing that you're trying to fix; like they're getting, you know, $100, you're trying to get them an extra $10.
Um, yeah, okay, so why don't they just pay the $10 in the same—like why don't you just add that to the invoice? They already have the $100 invoice; why don't you just add the $10 to the invoice?
That's—that's the—the difficult part is the calculation of this extra rate that comes on.
Sure, but you're calculating that with the GPS and on the app, right?
You don't need a cryptocurrency to add $10 onto the—
We—I—I say cryptocurrency. We could do it like a processing system similar to PayPal or Venmo, where they process internally.
It doesn't have to be cryptocurrency. We just know that—
I would—I would highly—if you have the chance to adjust any part of your plans, I would highly recommend keeping it as simple as possible and potentially just reconciling it in cash or Venmo or PayPal or even the same contract that they’re paying the truckers on.
The reason being is I like—I really encourage founders to focus on the pain point that their users are having, and the minimalist simplest solution to that, because then it tests your hypothesis.
Your hypothesis is that truckers want to get paid for the two or three hours that they’re spending at the dock, um, right now.
And it's a good hypothesis; it makes sense, um, and you—you so you want to test that. But adding all this extra complexity with like smart contracts and stuff makes it harder for you to test your core hypothesis and makes it—like it’s going to take longer for you to test that hypothesis.
So at what point do you think we could comfortably say that the hypothesis has been validated or invalidated?
Like, this is—this is like generally called this the search for product-market fit, and you're right in the thick of it. So your mission right now is to try to figure out what is the core problem that your users have, build the smallest app, and then start giving it to the customers.
And when you find that the customers are telling their friends about it because I’m sure they know other people who have trucks when they're using it for every load, um, when they're phoning you up with features and bugs and requests and it's just spreading like people are pulling it, like they're downloading it.
Even with you, you like getting in touch with them, that's when you know that you're on to something.
And so the mission in the early days is to get to that point before you run out of money.
Yeah, um, well, we don't have too much; we're bootstrapping so far, so—
Exactly! Yeah! So like that's perfect! And I would just like encourage you, so you have seven users now. What is your goal for the next week?
Um, well, we want to sign up—this payment processing system is a crucial component to our overall business plan.
We do have—we talked to our first broker yesterday; we're trying to get partnerships with broker, just as well, so we can—
Why do you need a partnership with a broker? Why don't they just use cash or PayPal to pay?
Because we—we're gonna take a processing fee on the movement of the funds from PayPal.
PayPal has a functionality where you can act as a marketplace and take a small percentage fee.
Sure, we could—we could use PayPal too, but either way, we want to work with the broker and have a system in place where they can instantly release detention money to the trucker without having to—again, like they're already doing this, right?
Because the trucker has to get paid for the—the $100, the main part of what they're doing reliably right now—a lot of the time they'll get denied—
Reestablish—ah, but now now it sounds like you're introducing a new problem.
It sounds like the first problem that you were trying to tell me about was truckers aren't getting paid for the hour or two that they're at the dock and now you're telling me about this new problem, which is truckers aren't getting paid reliably at all.
Well, the reason for the—the not reliable payment seems to be that they can't accurately track how long they were there.
Typically, they pick up the phone and call, and say—
Don't people just make a deal? They say, like, I have this much, this load, I want it delivered here by this date; I'm willing to pay this much. They don't get reliably paid detention?
Yes, the—the regular contract they get paid within 30 days because it's written and they're told from point A to point B, no.
The reliability comes in is originally what we were talking about was that last portion is not relied because they can't verify.
And all the reasons—
Cool! So like if you're doing that, I would definitely say ignore all the cryptocurrency stuff and just like try to focus—like just get them to send a check or get them to send a PayPal or get them to send a Venmo.
Like, there's got to be a simple way to do this without having to make a deal with the distribution set.
Like you want to—again, you want to test the hypothesis, make it as simple as possible!
Cool! So what's—what's the goal for next week?
So I feel like our goal for next week would be to—it'd be nice to like double our user size, um, or at least double our people that we—we're talking to and have them download, you know, let’s say majority of the loads—over 50% of the loads they move.
For the seven people who used it in the last week, did they use your app to get paid or did they use their app to tell the trucker I was waiting, to tell the, um, the distribution distributor that you were waiting for this many hours or something?
So, so far, all—it's been able to do is accurately log those two times.
We have an auto-invoicing feature set up, but no one's turned that on yet because they're not comfortable sending that invoice off of the broker yet.
We're still working out the—
Well, you can also—you can also manually just send an invoice—like you can—we have the app too, so you can tap a button and send the invoice.
Like, sure! But I mean, you too, you personally can do it as well; you don't have to let the app do it on the back, yeah!
Cool! So the goal over the next week is to get 14 users total!
Sure! Yep! Sounds great! Well, it's good to meet you both, and uh, good luck on the goal! Send me a tweet if you, uh, in a week with how it's going!
Sounds great!
Thanks!
Okay, next up, uh, Nate from Meme Party. Nate, can you hear me?
Nate, are you there?
I can see you!
Oh, hey! Can you hear me?
Yeah! Awesome! Thanks for holding these office hours, by the way!
No, I’m happy to do it, and it’s, uh, good to meet you!
Um, why don't you, uh—oops, why don't you tell me what—what your company’s doing?
Yeah, so my name is Nate, and I'm with Meme Party. So Meme Party is a mobile party game where people, um, can create memes competitively.
Okay! Um, what does that mean? That sounds, uh, um—that sounds that big! I'm not sure exactly.
So we currently have an app release on iOS and Android where, um, every round there—so people can join into a game with friends—like up to eight people can play at a time, and there's like a series of rounds—every round—up, um, each person's presented the same meme, they all edit it to create like the best meme they possibly can.
Okay! And everyone votes to see who made the best meme?
Cool! So it's like Instagram for memes, sort of?
Yeah, it's kind of like Apples to Apples, uh, those kind of games.
You say meme; you mean like the, uh, kind of like image macro with text?
What's your website? I'm gonna check it out.
Uh, so it's memepartygame.com; it's primarily just a landing page for our mobile apps right now.
Okay, cool! And how many users do you have right now?
Yeah, so last week we had 56 active users!
Cool! That's awesome! Some of your friends are using it?
Um, a little bit. So we actually had an event last week that we hosted where we got about 20 active users from.
Okay!
And then we also—
An in-person event?
Yep!
So we hosted an event at a brewery where we advertised and brought in some friends and some strangers, and we played it with them!
Great! So, um, tell me like what—why are people, like—why are people using the app? There’s many apps where people can use um, can kind of create memes; why are they using Meme Party?
Well, I think Meme Party is the only app that I know of where you can actually, um, compete to make memes where it’s actually like that’s what it’s designed to do. There’s lots of apps where you can like post memes or look at memes, but it’s not really designed to be like a party game or like competitive.
No, so there is—like, what do you mean, which is a kind of a card game which is similar?
Yeah! Kind of like what it’s called, Cards Against Humanity or apples to apples or that kind of thing.
Yeah! Yeah, but there’s some issues with that game more like—
Okay, so what’s the main—so are you the programmer? You—you wrote the app yourself?
Yeah, I'm the main software developer.
Nice! And what's the, um—what's the main KPI or what's the main metric that you’re tracking right now?
We’re just tracking, uh, weekly users and a weekly—like, is your goal to make people daily users of the app or weekly users?
Um, weekly! Because this is a game meant to be played with friends!
So it’s the way I think of it, it’s—it’d be primarily used, like, during weekends or when you get together with friends.
So it might not become like a—
Oh, sorry, I misunderstood! So this is actually like an in-person game! So everyone opens the game on their phone in person?
Oh, cool!
Yeah! Party! I used to play Space Team; that was like a fun in-person video game!
Yeah, cool!
Um, awesome! So what are the—what are the big things that you’re working on? What are the—what are the problems that you’re facing right now?
Yeah, so we released our iOS app about three months ago, and then about a week ago, we released our Android app. So I’ve been spending most of my time just doing development, but since we've got both the apps released now, we’re kind of transitioning more into like marketing.
So what me and Lucy, who’s another member of the team, I’ve been doing is like several nights per week we’ve been going out to breweries and just kind of asking people if they want to play with us, and that's how we started growing our initial user base!
So great idea!
So it’s kind of—you're kind of thinking, how do we make this a trivia night game at the bar?
Because we started doing like social media marketing and putting up flyers and stuff, but we found that didn't really have any effect; no one is actually using it, and part of the reason why is because since it's a party game, you can't really use it by yourself.
Yeah, need your friends to be there anyway!
Yep!
So we found—so I think like that’s—that’s a really good learning that other companies in Startup School could benefit from.
That sometimes you just can't market to people over the internet at the beginning; you need to be there in person like talking to them. Tell me, what did you learn from talking directly to people like what was your conversion rate when you were actually like talking one-on-one with someone at one of these bars?
I would say our conversion rate was like almost 100%!
Really?
So Lucy, um, she’s a very good people person, and she just looks around the brewery for like a group of people that looks welcoming, goes over there and asks, like, “Hey, would you be willing to help us test out our app and give us some feedback?” And almost everyone says yes, and they, you know, spend like 10 or 15 minutes.
Here’s the important next question: has anyone who downloaded it at one of these brewery events used it again later on when they were back home with friends or otherwise?
Uh, so I actually don’t know! So we actually don’t have a method right now for tracking users over time, but we’re—we’re adding—
Yeah, that's a pretty—that's a pretty important feature, you know? There’s many ways that you can add that quite easily to your app. One would be say to add Mixpanel. Have you considered doing that?
We've actually already added it, but um, we don’t have data from last week—
You added it after that?
Okay, yes! We added it like this—
Good! Well, I'm glad you added it! It's never too late!
Yeah, the right time to add analytics is now for yesterday!
Okay, so are you going to try that again this weekend to go to some breweries and bars and do that?
Um, yeah! So we're planning on doing it a couple times this week!
Okay, cool! I think like my hunch—my first impression is that it’s a great strategy for getting your initial users because people don't want to say no in person.
Like they’re always—they're more than happy to have a conversation!
Like they're at the bar for a reason!
Um, the problem will be getting people back after. Uh, that will be diff-difficult.
And in order to do that, I hope that you have things like in-app messaging where you can push notifications to people, say for example—like on Friday you meet them in the bar and on Saturday night, you send another notification to say, “Hey, if you’re hanging out with friends, why not try this app again?”
Okay, it’s a good idea!
Yeah, because they won't have uninstalled—hopefully they won't have uninstalled the app, and they're again again out with friends or at a point where they want to try it.
What are some other—you’ve had on how to find users, because this is probably going to be your biggest problem in the early days?
Yeah, so we also have a social media account, so when we started out this wasn’t effective, but we just tried posting like just memes that were just kind of generic just to kind of create like a meme account, but we found more useful is posting memes that people have actually created through Meme Party.
Have you recorded a really cool kind of like short video of how to get started, like a couple groups, a couple people getting started really quickly with Meme Party?
That’s a good idea; we haven’t done that.
That could be something just brainstorming kind of off the top of my head. Um, where I live, there's a lot of card shops where people hang out on Friday and Saturday nights to play card games.
Um, you know, magic and Warhammer and that kind of stuff.
Is it possible that you could go to one of those and see if people who are already playing games might be interested in doing that as well?
Yeah, that would be a good venue to try marketing, yeah!
Um, yeah, another thing would be some pubs host trivia nights on Tuesdays.
Um, that could be like—you could even go to a bar and say—I guess you already tried this—but like go to a bar and say, “Hey, if everyone downloads Meme Party, um, we can like run—we could run a trivia night on the app!”
Um, cool!
Uh, the other thing that I would do is like really think—really as your users are using it for the first time, look over their shoulder and really see how do they use the app; what are the problems that they have setting it up and connecting it with all the other people?
Oftentimes, like my favorite apps are the ones that just go out of their way to make it so damn easy to use, and I think like there are so many things that are happening in our lives; it’s so easy to like have a little bug and then move on to the next app.
So I really think carefully about that first like 30 seconds of usage from downloading the app to getting it set up to getting their friends on the same game.
Like I would really—I would really work hard to kind of make that.
Yeah, I think that’s really critical because we’re still kind of early phase. There's been a few issues with our app, and I think that’s part of the reason why our retention hasn’t been very good is because there’s a few like annoying bugs.
But the good news is, since you're in person, you don't have to wait for people to report the bugs; you can literally look over their shoulder and see like are they just poking the wrong button!
Oftentimes—like as developers, like we have it in our head how a thing works perfectly, but it's possible that, you know, the user who's using it may actually see it in a completely different way.
Yeah, that’s one of—that's been one of the most helpful things about actually playing it with people in person is like looking and seeing what issues they have and what's not working.
Um, what's your goal for the next week?
Uh, so our goal would be to get up to 100 weekly active users.
What are you at right now?
Uh, we got 56 last week, and, but you don't know if people were coming back again, right?
Uh, I couldn't measure last week.
Okay, cool! So week one, 56, week two, 100. It’s pretty good growth!
So I do have a question for you. So right now we have no revenue generation features, so we have no ads primarily just because we want to like make the user experience as good as possible, um, and we don’t have any other—it’s a free app.
Yep!
We have a few ideas for revenue generating—honestly, it’s it’s very odd for me to say this because I always encourage companies to really focus on revenue.
I think that's okay if you're in the gaming world. There are some pretty common models that are just very well known; you can sell in-app purchases, you can sell upgrades, you can add ads; it's pretty common.
Like, you can Google around and see what the CPM apps are paying for um or receiving for ads.
Right now, I wouldn't worry too much about that. Um, I would focus mainly on like, like you said, getting to 100 users, figuring out your retention, figuring out your churn; that’s like the biggest things to worry about on these apps is like do people use it again?
Because you need to get to the point where you understand, um, like eventually you may be buying ads; you may, you know, may be spending money to acquire users; you need to make sure you have a complete understanding of for every user that comes in the funnel on one side, how many active users do you get on the other side.
Okay, interesting because we had, um, back to the revenue for a second, I agree with you that I think, um, focusing on user growth is important. That's why we haven’t added any revenue methods right now, but one idea I wanted to get your feedback on was having like tournaments.
So I was looking at, um, other YC companies that have been founded, uh, funded in the past. There’s one company called Players Lounge where people like compete and like they kind of bet on themselves and if they win, they get money from the other players.
Yeah, um, we’re thinking like, I wonder if something that would work like that for Meme Party possibly.
I mean the good news about your app is you will—like you don’t have to come up—you don’t really have to innovate business model side; you can if you want, but the main innovation is around the game, right?
Like Meme Party kind of like, you know, that—that itself is the innovation.
Um, I would really focus on making sure that that is in tune with your audience and your users and that people are like super excited about picking up Meme Party to use it.
Um, later on, you can experiment with monetization, and if the normal ones don't work, like in-app purchases or upgrades or ads, then I'm sure you could, you know, work on figuring something else out.
But like you won't even get to the fun part of figuring out revenue if you don't crack the first part which is do people like want to use it?
So I would just like spend my time and energy focusing there!
Okay! Makes sense! Thank you!
Yeah, cool! Well, it’s exciting! I’m gonna have to try it with some friends! Meme Party! I'm glad you have Android; I'm an Android user, and I always feel—
Oh nice! That makes me feel good!
Um, well, it’s great to meet—meet you and, uh, good luck!
Nice to meet you too, Eric! Thank you!
Cool! Thank you, Nate!
[Music]
Um, next up is WP Clover, Cheffy! Cheffy, okay.
Cheffy, are you there?
I'm here!
Yes, okay. I just need to—I don't think your video is coming through.
Oh, there we go! I am!
Hi! Um, let me just—cool!
Hi!
Hi! Um, thank you for taking the time to talk with me today.
Um, my product is called WP Clover. Um, I'm a software engineer, and I run a lifestyle company, and we're kind of transitioning that into a platform and service.
Um, and we're focusing on WordPress development, hosting, and maintenance for small businesses and startups looking to build their MVP.
Okay, um, so if I could—so you shared a mouthful; now I'm like trying to unpack each of those components.
So one thing that I find is helpful is for me to try to like repeat it back to you. What I—what I heard was, uh, software engineer building WordPress sites for individual companies, and now you're transitioning to build a hosting platform for small businesses where they can start a website really easily on WordPress.
Correct?
So the small tweak that I would recommend is instead of making WordPress the focus of this, make the focus of this, you are helping small businesses launch their first website or maybe their second website—their first good website.
Um, and it happens to be on the WordPress platform.
Okay, fair enough!
Yeah, cool! Okay, so what stage—so am I correct in saying that you were kind of personally building sites for people and now you're moving to this hosting platform, or was I correct?
Well, it's actually a little of both. We've always hosted our own clients’ sites, but the way we did it previously was not scalable.
Okay, so, um, we're moving into a space where we automate most of the processes so that we don't have to touch the sites as much, but still get them the value they need.
So how a customer onboards with us with their site is they send us a ticket; "I need this P fixed, I need this repaired, can you add this content?" and then we give them an hour a day, um, for to fulfill their tasks.
Cool! So they don't have to be technical whatsoever; they don't even know that it's WordPress. All they're saying is like, "Hey, I need to change the menu on this restaurant," or "I need to like add a holiday to this calendar," right?
Absolutely!
Okay! Sounds awesome! Um, how many customers do you have now?
Currently, we have, uh, three.
Okay!
And we’ll have three onboarding within the next week.
Nice! You're gonna double in the next week; that’s pretty—
Okay! Yeah! So, um, it was kind of an easy transition because we had some customers on a previous version of this, and we converted a couple of those, and then we brought on a couple more.
So, um, great! The biggest challenge has been finding our target market because in the lifestyle company, my target market was women, and most of our customers were women.
But now that I'm doing actual market research and we're putting out ads and connecting, we see that our audience is largely men.
Hmm, interesting!
So, you know, we changed our website to kind of fit the, uh, women feel and look, but now we're finding that we're actually converting more men instead.
What is your website? I’ll put it: wpclover.com.
Good! Good domain!
Um, okay, we get—be great with WordPress hosting, development, maintenance on one service.
Okay, so again, so it sounds like I—I might have misunderstood it initially—like this is for people who understand WordPress?
No, not necessarily!
Um, initially we were targeting people who probably already have a WordPress website but have no clue how to really work with it.
Yeah.
So that's where we were kind of pulling those folks in.
We do have people who are starting brand new, never had a website, and then we have people who are moving their WordPress site to us because they don’t want to have to manage it.
Got it! So one thing to consider is that you may need different landing pages for those different customers.
So this first landing page that I'm that I'm on right now, it's very good for people who already have a WordPress site and they're like, "Man, this is annoying; I just want someone to take care of it," like you totally nail that!
But if you're going to a new company or a new customer who doesn’t even know about WordPress, but they know that they want a site and someone to handle it, it may be useful to have like a separate landing page.
We have something similar; we have what's called office hours where they can, uh, come and talk to us about their WordPress site or if they're considering it, and then we have one that's just web development.
Um, it's kind of leaned more towards our customer discovery.
There’s—there's like solutions for freelancers, for startups, for businesses, um, but if I were a business clicking on this page, there's just not much for me here.
It says, "Visitors and customers get more leads through your local business."
Like you need to, like, like try to inhabit the mind of this small business.
Like what is the problem that they're struggling with?
They want to get listed on Google; they want to get, you know, be on Google Maps.
Like put yourself in the mind—
Like hopefully through the conversations that you have with these folks, like they'll share their innermost pain or their innermost problems and then you can like put that on the site and say, "You know, here—we're here to solve your problems!"
Okay, that's fair!
Yeah, because I think again, like as you transition away from being just a WordPress site to maybe being more of a Solutions provider for other businesses, you need to maybe think a little bit broader beyond just the WordPress.
Okay!
What else—what else are you—what else what are the other problems that you're working on? How can I help?
Um, marketing.
Uh, one thing when we first started—because we started this back in October—um, uh, I kind of got the realm that our information was highly technical, so we weren't closing as many people because, because of course, I'm transitioning from an Enterprise solution because I do open source for Enterprise to a more B2B with a B2C mindset.
Yeah.
Um, so we started doing simpler blog posts—like one blog post was simply "How to get a yes," and that's probably been one of our most popular, um, as well as, uh, some of our YouTube videos that are really simple little snippets of how to do little things.
Um, I guess in that realm, it’s like how to be—it—how can I be more non-technical while still being technical at the same time?
I've got an idea!
So one of the things that I like to do is talk to my friends who I—who are in this—the specific—who have this problem, um, and really just like, because I trust my friends, uh, and they trust me.
So I would—if I think that I'm overthinking a problem or if I'm like thinking on a much higher level because like I'm spending all day working on this, sometimes I just like to go to them and say—like maybe in your case if you're like, "Tell me about some of the types of customers that you're working with."
Are they restaurants and small businesses like physical small businesses, or are they more like internet small businesses?
More like internet consulting freelancers.
We do have a couple of startups in the pipeline because we do have a startup program.
So it’s like I—I have a consulting business and I want to have a web presence so people can find out more about me.
Okay, cool! So I would say I would go to like maybe your friends who are running these businesses and they already have websites, and just talk to them like, like my conversation to them might be, "Okay, think back to when you were setting up your website the first time; how did you know what you wanted? Who did you go to?"
Did you search on the internet to find out like where how you would set up your site?
Like really try to just walk through all of these like all the steps that led them up to making the website.
And then after they made the website, what are the problems that they've had since setting up, you know, WordPress or Wix or Squarespace or any of the sites that they have, and really just enumerate like very specifically the pain points and the problems that they had.
Um, this I think is really helpful because you and I, like as we as we run our businesses, like this is all we do every day of the week, so it’s very easy to like get caught up in the week of like the specific problem that one customer is having.
Sometimes I need a break out of that and really like talk to a bunch of other people with kind of the similar problem!
Okay, that makes sense!
The reason why I like doing this with friends versus say completely new customers, I think they're both valuable, but with friends, I can really like I could really work them.
I’ll ask them a ton of questions; I’ll like force them to sit in front of the computer with me!
Maybe like in your case, you could actually ask them to open up their WordPress or their Squarespace or their Wix dashboard, and you can see like how they run the problem.
Like what buttons do they click? You can see like what are the changes that they’ve made in the past!
I don't know! I'm just like I think you could totally hack this together!
Yeah, definitely good advice! I'm definitely going to play with that and then after I run a few of these, you know, you could see, well, you could test the hypothesis: do people actually want this? Yes or no?
And then after they—after you like find out that people truly want this, you could see like what are the minimum features that they need in an app to make this work, and then you can try building it.
Like there are a lot of, um, see—you're a developer; are you going to code this yourself?
Not a developer!
Okay! Um, well, there are a lot of ways that you can actually build apps without being a developer these days, like no code apps.
Um, you could use—uh, uh, the name is not on the top of my head!
I think there's some good threads on Startup School about like no code apps, um, but I could totally see you like hacking this together with Google spreadsheets, Venmo, and really providing value.
Understood!
Cool! Um, what’s the goal for the next month?
What do you want to accomplish in the next month?
Um, well, giving your new advice, I'm going to try to set one up and do it manually first to see how that works out over the next month or so.
Sounds good!
Yeah! I do have a question for you, though!
Let’s hear it!
All right! So ultimately, this space, it’s a fintech space that we’re in, and it’s heavily regulated.
Um, because we could, we tither around being a money service business, and we may need a money transmitter license.
So you probably already answered it, but what’s the best way to launch an MVP if we chose to go through the app way?
The best way that I would do it is because you’re gonna take like a small cut or would you take no cut?
We take—we take the cut, a platform cut.
So there are some services that you could use that do this. I know that PayPal has a Marketplace feature where you won't actually be the money transmitter; PayPal will be the money transmitter and you could just say like take 3% and put it in the, um, Wiu account from every transaction that happens on the platform.
Um, Stripe also has a similar Marketplace feature.
What you—the key word that you want to look for here is you want to be like kind of a marketplace.
Now, the tough part is—I'm not an expert in this—the tough part is you're actually gonna be—this might be harder, but yeah actually now that I think about it, because you're not actually selling a service; you’re just redistributing money!
Yeah, almost like a marketplace, for lack of a better word because people are coming on the platform to with the same intention of rotating money around.
Yeah, I don’t have an answer here; actually, now that I’m uh, now that I’m thinking about it more.
Um, there’s a website called WePay that I think does this, um, so you can check that out. It’s a YC company; they originally were started for like pooling money, kind of in a similar way—like pooling money to host a party or something like that.
Everyone chips in five bucks to host the party, so it’s a similar idea. Maybe they have a service that you could use—that's wepay.com!
Yeah, cool!
Well, it’s good to meet you, youf! And send me—send me a tweet when you when you host your first circle!
Will do!
Cool! Thanks! Good to meet you!
Bye!
Okay, next up, now we're on F Labs.
Okay, Armen! I hope you're still here, and you're here!
Hey, Armen! Can you hear me?
I can hear you! I think!
Hello, hello!
Hi! Hi!
So what is F Labs?
Uh, we're building software; sort of a moonshot project to predict, um, genetic evolution.
Um, so for example, to help design a better flu vaccine!
That sounds particularly relevant to, uh, what's happening in the world today!
Thank you! Yes, yes!
Give me a little bit—you’re building better flu vaccines! Give me a little bit better—give me some more information; like tell me a little—like the next sentence.
If I were to say like, "Okay, tell me more."
So the problem is that like we don't actually have the statistics yet for this year's flu vaccine, but historically the flu vaccine has got less than 50% efficacy, and the problem is that scientists have actually deciding next year's flu vaccine right now.
So they basically are making an educated guess as to what its DNA will look like.
So the idea is, can we use computer software, machine learning to make a much better prediction based?
And how are you—like presumably, you need to collect a lot of data in order to make this happen, right? That sounds like the hard part.
How are you going to get your data?
Yeah, actually the data is not the hard part; there are tens of thousands of influenza sequences collected every year going back, you know, the last decade at least.
The problem—the bottleneck is that we can't analyze all of this data!
Um, you mean it’s been sequenced and they're sitting on disks somewhere, or they haven't been sequenced and they need to?
Public—publicly available on government databases; anyone can go and download this data.
Okay!
The bottleneck is computational analysis!
What do you mean by that?
What I mean is if we just set aside the problem of prediction, suppose we take these DNA data; can we even just answer the question what happened?
What did the last epidemic look like?
And no, we are very bad at that.
What that means is that scientists put in that they can take like maybe at most 500 samples, wait one month plus for a computer program to run, and then get out the results they want!
Okay, I think I understand!
Um, what stage are you at with the company?
Um, so the stage I'm at is I'm—I'm sort of—I'm prototyping!
I have a prototype that I'm testing on small example data sets!
Real—real data? Yes!
So you're—you’re a programmer?
I—I’m a mix! I'm—I’m a programmer, I'm a computational biologist and I’m just a statistician!
Yeah, you combo for this startup; that’s—we call that like founder-founder startup fit. It’s good, good!
Okay, so, uh, you're working on this, um, and are you—are you like backtesting it right now based on like, uh, flu vaccines that like you have the sequences I guess of the vaccines as well and whether it worked or didn't?
So, so this is what I wanted to ask you about was milestones!
Because predicting flu vaccines is like this big milestone, and so the smaller milestone I've set first is can we analyze more than 500 sequences in less than one month—in like one day or one hour?
What do you mean by analyze 500 sequences?
I mean— I mean so I’m saying even though the big goal is we want to predict what flu will look like next year, the problem is we don't even know how flu evolved last year!
We're not even able to really understand what last year's epidemic looked like because we can't analyze all the data, and we can't do it in a reasonable time frame!
So that’s—
You're—you're kind of losing me a little bit because I lost between, uh, what you mean by kind of analyzing it in time?
Um, this—like because I—I just—I guess I just don't really understand what you mean by that.
So I mean I guess so in—if we want to be able to predict the mutations that are going to happen, the flip side is which mut happened in the last year of flu evolution?
And we can't even answer that question very well!
We also can't answer very well how the flu spread and...
Is that because—but I thought you said we do have or there are public data sets of the DNA of past years' clues, right?
There are the data sets, but we can’t put that data together into a picture of what—how the epidemic actually happened.
Like the way the example is like this—suppose I gave you a bunch of pictures from my family and I said I want you to put together my family tree, but I just gave you the pictures; would you be able to do that?
Okay, so do you have frequency as well as like you have, say, several hundred samples of flu DNA?
Do you know how many people were infected with each of those um strands?
Uh, yeah! So each—we—we have DNA sequences, and each represents one person who’s theoretically been randomly sampled.
And that's—so you do get frequency, and that you have several hundred samples, and they may have the same—
Yeah, we—we have—I mean we have like tens of thousands of samples, and I mean some of them are going to be duplicates; but a lot of them are different because blue evolves quickly!
Okay! And so what do you need to—like you're talking about what the first milestones is, right?
Do you have some ideas on what like maybe you could bounce them off me what you think the first milestones could be?
So so the first big milestone I've been aiming for is to have a scientist bring their 10,000 DNA sequences or 1,000 DNA sequences, let's say even, and for me to build them a picture of that epidemic in one hour!
So DNA comes—comes in, and I build them a diagram that shows how influenza mutated and how it’s spread, and I do this in a timeframe of one hour of computer analysis.
Why do you think that the hour is so important?
Like is—is someone else able to provide this answer, but they're doing it a much longer period of time?
Yeah! So the software that currently provides this answer, it takes at least one month to run!
And why don't you just start by making it less than a month?
Sure! I mean, I two weeks—that works too, right?
Can you do that?
Um, a good milestone? Yes, so, um, no, I can't do that; it’s strange, but it's easier to sort of aim for one hour than—
Why is that?
Because like presumably if I understand correctly, scientists are currently using the—the method that takes one month, so it kind of follows logically that if you were to have that, or even drop it to three weeks that would be good, right?
And people were switching over from the other system?
Absolutely! So the difference between current software and my software is current software doesn’t scale with big computers!
I get it!
Yeah, but like why don't you make the first milestone something—and presumably it's easier to make it have from four weeks to two weeks?
Sure, sure, I mean I—
Yeah, two weeks; that works too, right?
Yeah!
Can you do that?
Okay, so my first milestone is to be faster than current software at two times or three times—that works!
Right!
And, actually, I guess my hypothesis is that once I hit that, then actually I can easily coast to being ten times or twenty times...
That's great because you then you saying you could scale—that makes sense!
Okay!
Yeah, sorry—yes, you're right! My goal is just to be phased by the reasonable fact!
And have you talked to some flu vaccine scientists that could, um, kind of, um, switch to using your software if you do?
Okay!
Yeah!