YC SUS: Kat Mañalac and Eric Migicovsky discuss Week 2 SUS Lectures

Good morning everyone, and good evening. It could be anywhere actually. I'd like to start somewhere. My name is Eric, and I'm the facilitator, of course, facilitator here at Startup School. It's a pleasure to be joined by Kat.

"Hi everyone, I'm Kat, a partner at YC."

Kat and I are here today to answer questions about the subjects of this week's lectures: how to talk to users, and how to launch again and again and again. We've pulled a bunch of questions together from the thread on the Startup School forums that we'll start answering, but we will also take some questions from the YouTube comments, from Twitter, smoke signals, basically anywhere that you can find us. We will attempt to answer some questions, so let's jump into it.

The first question is from Gustavo from Flexo. Flexo enables you to build a financial model for your growing company in minutes, not days. So I only have one question about this one-liner, and it was, who is this for? Is this for startup founders, or is it for small businesses? I would just add that even though the one-liner is pretty good, just add who it's for.

Your question was, "There are so many channels we can launch on, but many of these I can either do something quick and hit launch, or I can spend days and weeks trying to maximize the chance that each channel is successful. What's your recommendation for balancing between maximizing the number of launches versus maximizing the chances of success at each launch?"

So one of my main goals for the talk was to sort of stop founders from putting too much emphasis on that one perfect launch. I would say, wrap spending— you shouldn't be spending weeks trying to perfect any of these channels. What I would recommend is, you might spend a few hours trying to get the messaging for each of these channels down. You might spend a couple of hours to a couple of days getting feedback on that messaging, but don't spend weeks stressing over each of these channels.

Granted, launching a kind of pre-order Kickstarter campaign or pitching press does take a little longer, and I'm sure Eric has some insights on Kickstarter and pre-orders, but don’t spend too much time stressing over any of these channels.

"Yeah, a launch—one of the things that I've learned is that think about it this way: everyone in the world doesn't know who you are right now. Every time you launch, 0.01 percent of the world finds out about it. So every single launch will be brand new to so many people. That's kind of the motivation to keep launching. It's like an MVP; if it's perfect by the time you launch, then you've waited too long. Yeah, you want to maximize your surface area, and it's so hard to get people to care about you at this stage in your company's life that you want more and more of those chances."

Next up, Matthew from Katara writes, "Katara, we help you find bias in your AI training datasets and then help eliminate it." My question is for Kat: how long is too long for a landing page call to action, like 'Sign up to be notified of our alpha launch' or 'Get on our mailing list to be the first users on our platform'? Is there a general rule of thumb on how long people should wait or stay interested?

"The longer you have people on a waitlist, the harder it is to re-engage them. So I would say get an MVP of the product out as soon as you can. Your question was also like how long, yeah, like how long the call to action— how long should the actual call to action be? It can be as short as possible, like no one—users don't care, like alpha, beta doesn’t mean anything to a user. Just say, like sign up to be notified of our launch or sign up for our newsletter, whatever it is, short."

I think oftentimes, when I'm writing content or copy for a site, I sometimes get a little bit too caught up in my own head. One thing to consider is you're the one who's been thinking about this for ages—like for hours and weeks and days. So it's possible that you may just need to get whatever you've written in front of anybody, and just ask them what they think and get their opinion on it, because oftentimes, we will obsess over the perfect copy when in reality—like, they still don’t know what we’re talking about.

"Yeah, you try to get as much feedback as you can as possible. Schools like a great place, too. You can post your content on the Startup School forum and just like, if you need some eyeballs that aren't your own, just post a thread saying 'Hey, can I get some feedback on this?' I'm sure that some people would jump in and say 'I don't get it.' Try to be—if you see something on the Startup School forum that you don't understand, jump in there, because you need to hear those honest voices of constructive feedback."

Okay, Daniel at Story Low. "Story Low helps brands hire photographers for their digital marketing needs." Okay, I get that. "We have a round of user interviews scheduled next week. We are also heavily engaged in paid customer acquisition through an incentive program. While we are talking to a user, should we pitch them on a new coupon incentive, or just ask for feedback? Is it better not to mix up these two user touch points?"

"Yes, I have some great advice for this. I like to separate my user interviews into two different groups: one is with people who are using the product, and one is with people who are not users, who I just want to learn about what problem they have on a day-to-day basis and what existing solutions they're already using. Once they start using the product, generally, I find that people just want to talk about how to improve it and give me feedback. It's really difficult to get them to context-switch back to a world before they started using your solution. So in your case, I would say separate. If you have user interviews with people who haven't begun using your products, try to treat those as kind of wide open interviews where you're asking questions to find out, what other products do they use? Have they paid money for similar products in the past? How did those products fail before you introduced your product? Start getting feedback on that."

"And then in terms of coupons and incentives, I always try to just ask for people's feedback without any incentive. Because honestly, if this is a real problem that people have on a regular basis, then you're helping them just by asking them about their problem and trying to build something as a solution. That already should be the incentive that they have if this is truly a big problem for them. If you're finding that you're constantly having to pay money to people to answer surveys and stuff like that, it's possible that they don't really have a burning problem. And one thing I would think about, too, is what you're building— is it really better than what's currently already out on the market? Because that might be a challenge, too, converting people who already use similar services."

"Yeah, as Eric was saying, if it's not a real problem that you're solving, or if your product isn't 10x better than what they were using, it might be difficult to convince people."

Next up, we are Latter, a platform that simplifies the process of finding effective mental wellness practices and solutions. "Okay, it's a few months—a few problems with this one. Like I can't tell if it's an app, if it's a website, if it's a service platform—doesn't really tell me. I think I get something to do with mental practices. I think it would be also—it's missing like, who is this for exactly? Like is it to consumers? Are you directly connecting consumers with therapists? Or let's assume, okay, I'm going to assume that it's—so who is it for, and what is the exact problem you're solving?"

Okay, their question is: what are effective strategies for getting influencers, bloggers, and leaders to help us launch, preferably as cheap as possible, if not free? A good question. How do you get people to talk about your product? It is really tough.

"I think one of the things that I had to think through is do your research. When you reach out to bloggers or influencers, are you actually solving a problem that they care about or that their readers care about? I would say that, you know—and frankly, it's a numbers game. So for a lot of the YC companies that I've worked with that have tried to get influencers and bloggers to cover them, they will reach out—you know, if you're doing cold reach and not getting warm introductions, you can reach out to 50 bloggers and only get one person to respond. That quantity—it’s quantity. So do a little bit of research, know who you're targeting, but also just keep reaching out. So if you need more than one person to respond, you're gonna have to send hundreds of these emails, and often people say, 'Oh, we tried to reach ten or twenty or thirty.' It's not enough, so you need—it's a numbers game."

"And I also—one of the things to aim lower. So there's a lot of people who want to help. Basically, the people that have this problem that you're trying to solve, they exist everywhere on the Internet. Some are these like top-level influencers, others are people who run like a Facebook group. We've seen a lot of companies get their first users and people talking about their product by finding like a Facebook group. There's a company, two batches ago, that made a device that helped kids with cerebral palsy learn how to walk. They simply just went on Facebook, found a group of parents and families whose children have cerebral palsy, and got dozens, if not hundreds of signups just through that group."

"So there are subreddits, there’s like half Twitter hashtags, there are Facebook groups—what are some of those? You know, there's Product Hunt, there’s Happy News. If you're reaching out to bloggers and you're emailing, don't think that they're going to respond after the first email. It's just like sales—you have to follow up with a follow-up email. There was a company that I mentioned in the talk called Joy. They were trying to get featured by a wedding website builder, and they were trying to get featured by wedding websites and blogs. So they, you know, they sent out—out of fifty, only one responded. And they had a whole sort of like drip campaign to follow up with those sources."

"Have we definitely follow up? How are we doing on the tech thing? Still streaming? Yeah? Oh, okay, cool! Oh yeah, forgot that the internet here is not perfect. Okay, we're good now. Okay, you can hear us. Okay, next question.”

"I'm the founder of Porch Ship. We provide an affordable pickup for any return shopping returns directly from your porch."

"That's very clear. It's clearly not talking twister with a lot of mentions of porch. Okay, I'm launching in Pittsburgh this week. What can you advise for launching in a specific metro area? In my case, Pittsburgh. I've been to Pittsburgh once; I don't know if my Pittsburgh area knowledge is perfect."

"Okay, so launching in an area with like a geo-targeted service. I guess you have like—you must have runners that go and pick up these porches or something like that. I mean, one story that I love is Lug. When Lug is Uber for movers, they basically will come deliver snow, come up and pick up anything—though don’t pick up any delivery items. So what they did on their first week just to figure out, you know, just to start getting people using it is they stood outside of IKEA. They actually had—one of the founders was driving a truck, and one of the founders would run up to people in the line and say stuff like, 'Hey, do you need help moving this? Getting this home?' And they would have—you know, they're user like click a button, and then the other founder would drive the truck up. So I think in the very beginnings, you're gonna have to do some manual stuff on the ground. Ideas like that."

"So, like, presumably, you might have gone to school; you might have worked at a company in the area. Why don't you get all of your past co-workers, your friends, your colleagues, other classmates using this? It's kind of a universal problem; like people who buy stuff on Amazon need to return it. I'm sure that you can just talk to your friends and see if they have the problem."

"Other alternatives are—we've seen companies have some success with door hangers, where they just go to Kinkos—print off, are those still Kinkos? No, I think it sounds like that’s... okay. So they go and they print off little door hanger things and they go around to houses. Just hang them on doors; it was kind of top of mind."

"I could imagine like, what if they printed off stickers and you've, like, as you saw Amazon packaging. Then people would think that you're trying to steal the Amazon. You could put stickers on other people's Amazon boxes on their porch. One thing that worked for me is I recently started using a startup because they put something up in my building's elevator. So they said, 'Hey, if you want, you know, to have these types of things delivered.' And then I took down the numbers, just like wherever there's a higher concentration of people."

"Yeah, that's right! That's a cool idea."

"But like, did you rip off a little tag or was it like—"

"I didn’t; I just took a photo, I guess."

"Okay, next one from She—she is the founder of YD S Medicine, which provides self-driving molecular design services serving as a GPS for the journey of drug discovery."

"Okay, I work with a lot of bio companies; bio founders have amazingly interesting products and ideas— they generally have to dumb it down for the rest of us. So this might be—this one needs a little workshop, maybe. The question is how are we to launch biotech or pharma tech companies in terms of press and media? We've found that Fierce Biotech and Stat are great, and they are often interested in early-stage bio and life sciences. And what I do is I, you know, look at what reporters have covered companies like mine, or like the company I'm going to pitch before. Is what I'm building in the wheelhouse of things that they're interested in? And then I make sort of like a list of who I might reach out to, and I prioritize the first. And I, you know, say I'm going to Fierce Biotech, I'd find the right reporter there, I’d reach out, offer them an exclusive on the fact that we're, you know, we've launched, and then go down the list from there. If they don't respond or pass, then go to the next, you know, reporter or publication."

"I think that I always love launching when I have some reason to launch, like some sort of event. So oftentimes scientists are going to conferences, they're publishing papers, they're publishing posters—which are a thing. So if you're doing that, try to consider reaching out to reporters or bloggers or people who are in this space and pitching them your idea timed around that. It's a good reason to say like, 'Hey, pay attention to me, there’s something happening like this week.'"

"Yeah, I think that's great. I think especially if you're from the life sciences, you might be publishing a paper, and oftentimes, like, the university you might be publishing that with will help reach out to you. They sometimes have groups that specialize in trying to get like the work of students and researchers out, so that could be another help. Mmm-hmm. Cool."

"Next up, we are Robin and Rover, an online platform for people to trade their purchasing data directly with companies."

"Okay, I think I know what you do, but I'm guessing there's something a little bit missing, like purchasing data? Like, my Amazon shopping history? Okay, so this is like one of those companies that is trying to get from ads—like monetizing based ads, like a clear transparent giving consumers control of their data. Cool. Yeah, okay."

"Regarding the Mom Test, you say to talk about specifics, not hypotheticals. I need to test which functionalities to develop to evolve our current MVP. To do so, I've gathered some insights from users on what they would like to do. I want to test them with others, but should I ask hypothetical future features? How would you recommend to test them?"

"Okay, so I always like to—when I'm testing an MVP with users, I always like to build something—whatever is the actual MVP, like the smallest set of work that I have to do in order to solve their problem and then put it in their hands and get feedback. Talking about a hypothetical product is hard. Talking about a hypothetical—sorry, talking about a hypothetical product is hard. It's much easier to talk about a problem that they're having and then build a solution, give them the solution to the problem they told you they actually have, and then validate your hypothesis. Do they actually have this problem? Sometimes people will talk about a problem that they think they have or that they wish they had. Like, you know, anyways, so yeah. So as you're gathering these insights, try to gather the insights about the problem that people have, and then in your back of mind, think about how you could solve them and then give people the solution so they can try it."

"Okay, next one—okay."

"Okay, aq. Okay, Keep Calm, teach kids AI using video games. Cool! Does it make sense to save the hard launch—the full version 1.0 public release for a big event such as demo day?"

"Well, if you were going through YC, we would say no! Don't save that big launch for demo day! Launch as quickly as you can so that by demo day, you have some amount of traction or you've been able to get real user feedback and you know, iterate the product based on what your actual users are saying."

"Yeah, I think like just hearing some of the words that you're using about this, saying like hard launch, full version, 1.0 public release—I think you may already be kind of getting too complicated with this. In reality, what you're making sounds interesting, like a game that helps kids learn AI—like just get that out there! I'm sure that either you have kids yourself or you know some people who have kids. Don't treat your launch as like, oh, it has to be perfect, sacred thing. Just like push the button, make it go live, make the website live, and tell your friends. That's a public launch, and that's something—like Kat was saying—that's what the first point on which you can iterate afterwards."

"Like I said before, at the beginning, every time you launch, there will be a multitude of people who never heard of what you put out—your app is going to be so they'll be happy to find out about it then. So yeah, don't overthink things."

"Okay, Blake from Live Simulator. Live Simulator is VR training for first responders. Cool, got it! I let users run the call by placing them in VR and having them command virtual crews to resolve an emergency, communicating using an in-game radio, and trying not to get anyone virtually killed. It's created by a firefighter for firefighters. Oh, this is—oh, it matters. Okay, it's a long one-liner. I think your one-liner, 'VR training for first responders,' already does. Great. I've built three modules over the years, but the two failed launches convinced me that the bar is much higher for my MVP. Does YC have good examples of solo founders who've built very big projects while working full time?"

"I mean, we probably do have some great examples, yeah. I mean, we certainly have founders who built their MVP while working full-time, but I think after the MVP, you know, that was enough to convince most of them that they needed to go full-time to really build it into a big company."

"I agree! I think like working a full-time job and splitting it is a good way to get started. And really, like the first test that you want to run, if you're working at a big company and you're working on an idea, is do you even care about this idea? Like, you're the first person that you have to convince. If you're not convinced that this is gonna be exciting and interesting, there's no way you're gonna convince other people. So get yourself convinced, and it's possible that while you're working part-time on this, you'll be convinced that at least there's a problem. So while you may have discovered that your MVP doesn't solve that problem yet, I hope that over the two years or over the time that you've been working on it, you've kind of talked to enough firefighters, just leaders like the captains of that kind of thing, to figure out if people actually have this problem."

"In terms of like—maybe to add, I guess," Eric continues, "we'd have to have a bit more information to understand why you think that launches have failed. But I guess if no one is constantly using your MVP, that's tough. I would say that like if you find that if your MVP is designed to be used regularly, and people aren't using it regularly, then that's a good sign that like you don't have the features yet, or you're not solving the specific problem that they have."

A question from Kyle at Phase. "We're Phase, a white-label payments SDK that enables our enterprise partners to issue digital rewards in exchange for the end user's value that's stored on their platforms."

"Okay, I kind of got that. Yeah, yeah, a white-label payments SDK—it's like Stripe, okay? Digital rewards, which I'm not sure what those are, in exchange for end-user value, which also is vague. I'm a bit lost."

"Well, let me get back to Kyle. The question that Kyle asks is once our white-label payments SDK is integrated, we only get paid when the enterprise partner's end users of our service—like end users of our service—we pay X dollars per transaction, and our enterprise partners upsell our services to their end users. In this flow of funds, it doesn't enable us to A/B test different pricing for end users. How do we start thinking about experiments around pricing?"

"There's a lot of content that YC has published around pricing. I always love pricing for enterprise partners because you kind of don't have to publish it on your website. You can have one price for one customer, one price for a different customer, and they probably will never find out about it."

"So especially in the early days, you actually can run a lot of A/B testing just by not having a public version of your pricing. Have you seen what else is doing?"

"What else? Doing trade? Any interesting enterprise—"

"No, I think that's right. One thing that I know surprises a lot of founders who are building enterprise companies for the first time is that you are probably charging too little. That's the most common thing."

"Hmm, I've seen is that you're not asking for enough. I mean, one way to think about it is for any enterprise software—if you're building something that solves a problem that the enterprise has, the alternative is probably them using their own engineering team to solve—to like build a solution to the problem. That's upfront engineering time and labor, and then the maintenance cost of maintaining that software over time for any engineer to pick up a pen. Whatever engineers you used to do things these days like it's gonna be hundreds, if not thousands, of dollars. So yeah, it's Kat's point: like don't be afraid to ask for more. They'll push back if, you know, like again, you think they really truly have this burning problem, like they may say no to the price but they still want it, and there will be a price at which they'll pay the trade-off."

"Okay, Zyre at Boom Loaf. We're in love—we help users to separate most of their inorganic trash by scanning the barcode of empty packages. Okay, good help! With whether it's for companies or businesses or people's homes when your user is different than the paying customer, how do you optimize between the priorities of your users and those of your customers? Great question!"

"This happens a lot less often on the consumer side and more often on the B2B enterprise side where the buyer might be the founder or the manager, and the user might be someone else. So this is a really good example of needing to do a variety of user interviews. I generally will start attacking a problem that has the same person be the user and the payer. But if that's the case, then they’re split. You generally have to figure out how the flow of money works—like who's the person who decides whether to pay to solve this problem? Who manages the budget?"

"I talked about this a little bit in my video. One of the important questions that you need to resolve during user interviews is: are they already paying for a solution to this problem, or do they already have the budget to pay for a solution to this problem? If they don't, if they don't have budget, it's possible that they may not be the best first users for your product. We're selling our services to municipal governments. I'm sure there's—just where the head is. Are there any hacks that accelerate their decision-making to pilot test your product?"

"It's useful! Governments are a slow process; extremely tough! Yeah, I guess some things that you can consider doing are find out when the budgeting process starts. Understand if you actually—if they have a budget or if they have line items in the budget to solve this. I mean, it's kind of a non-starter if they don't; that would be like six to twelve months of getting back on their radar."

"I'm gonna read the whole question."

"Okay, yeah. [Laughter]"

"Okay, Mandor at Contact Pointers—Contact Pointers is a service for automatically updating other people's address books when you change your contact information, and letting you control/manage what info they get."

"We are hoping—okay, gonna pause there—that idea I have heard before, yeah, we have funded before, yeah, we definitely have like—yeah, it’s something that everyone wants—it is a problem. So here's one of those tricks: like it is a problem that everyone has had, it’s not necessarily a problem that people are willing to pay money for, right? There are a lot of dead bodies in the past. I would recommend going to look at them—what happened to them? Try to do a deep dive there."

"Okay, we’re hoping to do a friends/family launch in the next few days, and we’re wondering about what our feature targets should be for an online launch. Because of the nature of our service, it has a strong network effect, but we have features planned to make it useful and sharing contact information with people who don't have the service. My question is how vital do you think the quality of these features is to an online launch? I'm worried about potential early adopters' disinterest if it's not useful because there's a large network, and then it being harder to get them back and for a launch later."

"This is a legitimate worry! No, I don't think so! Because there are a lot of people in the world. And so if you find that, you know, the first few users try it, and they don't like it, they leave, it'll come back—not the end of the world; because your product, from what we can tell, is designed for a lot of people. If there were only like a hundred people in the world who might use this—I talked to a product yesterday who is making something for sailing coaches—there are probably only a thousand people in the world. So you can ask 100 of them; that’s not a good thing. But in your case, like yeah, you should not be worried about this, especially your friends and family; like they will use it as many times as you tell them to use it. So, yeah, you can launch again once you have an improved product if you can launch to the same people."

"I've launched the same people over and over, and they, if anything, like as it gets better, they'll be like, 'Wow, this is way better than last time!' The problem that I could see a little bit with yours is it sounds like it has a network effect where it needs to have a network effect. I would try to build it so it doesn't need a network effect—so just when you use it by yourself, it's more useful."

"Hi, hi! That's Sunny from Start Stream Today Industries. We tell our bank that we make silicon-case antennas for responsive sensor frameworks, but we could—we actually make AI sex toys that double—the length of a woman's orgasm. I have a question for Kat about relaunch narratives. Long story short, I successfully launched X story five years ago as CMO after initially going viral, a dispute that led me to believe, and the company to pivot into medical devices, and then being asked to return to refocus the business to our original idea. Recently, here's my question: If you were a potential investor looking at a rebirth company like this, what indicators—what kind of indicators would you be looking for to prove that a team can continue to make new moments in the spotlight?"

"Uh, a little complicated question, okay? But yeah, it’s really—eliciting the question. Yeah! So, she, he, or she, left and came back—now they’re relaunching this company and trying to look for investors. Yeah, they started out as a sex toy, pivoted to medical devices, and now pivoting back. And I think I’d want to understand, like, what you and the rest of the team, you know—what convinced you that this was the right direction? What have you learned? What have you learned from all of that that has just happened?"

"Anyone who works on something for five years has learned a tremendous amount about the industry and the problems that—yeah! So you should hopefully know a lot about—what are some major insights there? Exactly, yeah! And then, you know, also have you—you know, I don’t know whether you’d raised a ton of money and burned through all of it now—or whether this is the first time you're raising that because that can make things complicated. But yeah, I think the main thing is that clearly your team has expertise in this space, and that's exciting. But yes, what is it that you've learned, and what do you know that others don't? Yeah, good luck! It'll be a fun ride."

"Yeah, I’m Rafi, founder of Aventis Media, publisher of The Bridge Magazine. It's a digital magazine aimed to make news, finance, and career-related information relevant, engaging, and useful to young adults aged 13 to 18."

"That’s a little bit long in terms of what led on, but I get it! This is a subscription-based service. My users are the kids, and customers—the kids and the customers are the parents who pay for the subscription. I'm currently in the process of customer discovery. My research reveals that the kids are not looking to actively solve the problem, but the parents are. I've been asking questions similar to that in the video, but I feel like I'm missing an angle. I appreciate your take."

"Okay, so in general, I think your users and customers are kind of both the parent. Hmm. If you’re like, these topics sound pretty long-term focused on kind of the little, like, what happens after the kid grows up? I mean, not a 13 or 18-year-old, but unless you're super ambitious—yeah!"

"Actually, maybe that's kind of the point. If you specifically are trying to like, talk to those users, maybe focus your product down even more to make it like just for like IB students, or just for Model UN students, or something like that, or—be a very specific niche—a very tiny niche! Because then you won't have to have different users and different customers—they'll all be the same person, which is generally easier. Because, and maybe price it accordingly. If someone’s spending their allowance on it, it's maybe different than if they’re thinking about how much better this is going to be for their kids’ education and get them into Harvard and all that kind of stuff."

"So think about your audience, and maybe play with the positioning of your product, huh? Interesting! Next up, Alexander at InstaCents. InstaCents is one-click micro-donations to support digital content. Cool! My initial hunch was to validate the idea by talking to content creators, and we thought would love to unlock new sources of revenue. However, most of them aren't engaging. I find a prototype would be easier to show exactly how it worked."

"Okay, I'm just going to stop there. There are many solutions to this problem, so the first thing that I went to is kind of the same as the answer to the other question. I would look at who is trying to do this in the past and figure out what are the problems. As you’re talking to content creators, I would only talk to those who have experimented with the other solutions, because if they haven't experimented with any solution in the space, they probably don't care. Yeah, and I would really focus on how what you're building is, you know, better than stuff like Patreon or things that every rock—yeah, a lot of things in this space—yeah, there's a multitude of these."

"But yeah, if you're technical and you know it's going to be fast to build a quick MVP, I think that's always—I've talked about this a little bit in the last video last week. But one thing you can actually do is you can take an existing solution that's already on the market and kind of hack it or white-label it and make it yours so that you can experiment very quickly by taking something that's already built so you don't have to build an MVP."

"Bruno from Glee Board: we board matches ambitious CEOs with curated executives and founders into ad hoc advisory boards to help them collectively make better decisions. Sounds useful for certain school companies. You've posted that on the forums. My question is: when you are exploring different user niches, do you prioritize the users that you talk to? Do we go to several of them at the same time and then zoom in where you feel it’s more relevant, or do you go your one group after another?"

"I personally like to do things more in a sequential path. I do it kind of like this: I sit down and I whiteboard out all the potential user groups that I could talk to, and then make a guess as to which ones I think are gonna be most valuable, kind of based on those three overlapping Venn diagrams I talked about in the video. Which ones have the most acute problem, who have the most frequent problems, and who are more likely to spend money to solve those problems? I take a guess at like which groups will be the best, and then I go and start talking to them. If they don't prove to be good, I'll move on to the next one."

"Cool. Next, Yoda from SSS Spot. Spot, we allow—we connect drivers in Latin America who are looking for a public parking lot with public and private parking lot owners. Cool! An app that helps you find parking spots in Latin America! I have a question related to user interviews— we've launched our prototype six months ago, and we have 15 percent month-over-month growth. That's good! Congrats! That's very impressive! Now we're looking for proper market fit. Is it okay if I make the interview with users who are discovering the product, or do we have to make it with our active users? We're testing the feature parking bookings in sports games and concerts."

"I kind of touched on this for another question, but generally I try to segment these user interviews into two different groups: one with people who have never talked to who aren't using a product, and one with people who are using the product. The questions are slightly different for each of those two groups. It sounds like you're on the right path towards product-market fit. That growth is indicating that there is a problem, hopefully that's revenue. Yeah, I think that's revenue growth. If you're testing a feature, I would recommend testing it with your current users, but it sounds like this might be like a new group of users—like parking for sports games and concerts. Do you think you’ll find for that, reaching a new audience? Yeah."

"So I'm definitely trying—I would definitely also try to talk to new users, because I could imagine, yeah, people who are using your app to like park for work aren't the same people who are going to Gweek or something like that."

"Okay, moving into moving into Twitter. John Dougherty asks, 'What's the fastest way to identify the beachhead when you have paying customers for a mix of industries or sectors?' Kind of—kind of again to that last question. Like you probably—if you're building a product, hopefully, you are a user in the past, like you're building this to solve your own problem, then you should have a pretty good idea of who the potential users are, because they should look a lot like you. If you're building a product for an industry that you're not already a user of, then kind of do what I just said—generate a list of hypotheses—look through who are your most active users or who you know, or you know, who's the easiest sale—like who, then, you know, you're actually solving a big problem for these people, they're willing to pay, and maybe it’s worth exploring that particular sector."

"Well, they have paying customers, yeah!"

"Yeah, sorry, I misunderstood. So if you already have those users, yeah, talk to them. Like within those users, there's probably larger demographics, and you can like double-click and say there's two users in this one group. Maybe we can find a hundred other users that are in a similar space."

"Okay, we are Assist Force, the most reliable freelance platform for tech companies that replaces outsourcing agencies. How do we approach Silicon Valley tech companies if we are a no-name European company?"

"Product Hunt and Hacker News are the first sort of communities that I highly recommend. I think those hit a lot of the, you know, people at the companies that you want to sell to, and those are—if you haven't done those yet, I highly recommend it! Yeah, Product Hunt especially with yours. I think it's possible that there are a variety of different platforms already that people use, so like to double-click on what Kat said, you really got to differentiate like why are you better than Upwork, Toptal, and all these other—"

"Okay, Sandro Franco asks, 'What's the downside of referencing a product or business that already exists?' For instance, our company is 'I do an interactive TikTok/Instagram for language learning.'"

"So yeah, so I—a lot of people ask me, should we use the X for Y construction, which is Uber for X, or Airbnb for X? I think sometimes the answer is you should use that construction if it's the fastest way to paint a picture in someone's head of what you're building. So you know, for example, if you know TikTok/Instagram, like the X has to be a well-known household name—it can’t be something that only a few people have heard of—and then it has to be fairly clear why—why would they be interested in X? So an interactive TikTok or Instagram for language learning? That paints a pretty fast, yeah, it’s a good picture! Yeah, so it’s a good—like the best one-liners always generate questions. Like, it makes me interested to want to learn more, and I have questions about what you do. So that's a good start! I would definitely use that."

"When wouldn’t you want to reference a product? So when it’s—oh, I’ve seen a lot of people describe themselves as like, 'We’re Work for X.' Yeah, when it's like—if there's a company that's in crisis, you don't want to—like, WeWork is something you might not want to—I've seen—I saw a lot of WeWorks for X in actually applying to this batch of YC, and so I thought about that hard. I was like, 'Wow, I personally wouldn’t use that.' I wouldn’t use it if the X is something nice—an alum once called themselves 'Buffer for Snapchat,' which like—I don’t think enough people know what Buffer is to make that—make sense!”

"At that point, just describe the problems they're solving or what you're building in general!"

"Yeah, Kelly Romina asks, 'Would a simple site like our example Magic, which is just like a single-page landing page, be sufficient and be sufficient MVP to begin raising money, or does the MVP need to show extensive working features?'"

"Magic, I think, raised a lot of money on the back of that MVP. Yes, so the answer is—it depends! If the MVP actually solves a problem that people have, then by all means—a web page that says 'Hi, my name is Eric, here's my phone number—I will solve your problem if you have this!' is a great MVP! Like we have founders in this batch of YC where that literally is their whole product; like their product is their brain can solve this problem for you, and other people have this problem. Like, talk to that person and solve the problems!"

"Absolutely! Like, if you can solve someone's problem and put it up on a webpage, put it out there, friendly—that's all they needed to get—they have tens of thousands of people sign up for their waitlist, so that’s all Magic needed! But of course, not all companies are gonna have it that easy essentially. Okay, should I answer this part? Which question? Yes!"

"We are Inspir, we're a smartwatch company. Our main feature is real-time hydration monitoring. Since you're on a watch, our sensor's up and running. The development of other parts is not. Organic growth by revenue by selling one piece of time to take forever."

"Okay, this is a common mistake that people make—organic revenue by selling is the only way that your company can be successful, and it starts with one. You make one product, you sell it to one person, and they're happy. Then you sell it to one, and you…so forth, then you select, and it grows. So oftentimes people get way too ahead of themselves, where they think of success as only if I sell a thousand of something, when in reality, like you absolutely have to make one of them and get someone to use it, and they have to love it. And if you can't do that, then it's probably not worth working on."

"People in hardware always think like, that's not the way to do it, but that actually is the way to do it. So ideally, if you can use—"

"Your problem is that you've made the sensor but not the smartwatch. Why not try attaching your sensor to someone else's smartwatch? Some other smartwatches have Bluetooth interfaces—just add it onto the strap! Like there's a lot of ways that you can have this together without building it. Good luck! Here's one! That got—Naga Sri from Tori asks, 'We connect global shoppers to international travelers. My launch is delayed for four months, and there's a huge drop in the users. How do I get them back?'"

"I think there are a lot of ways to re-engage customers and users, whether that's reaching out—I mean, I think if there are just a few, you can reach out and call or try to contact every single one of them and get on the phone with them. You can send emails. I think there are a lot of ways to catch, and it depends on what—you know, we’d love to hear an episode of what you've tried already."

"Yeah, one thing I like to do is to reach out personally as the CEO. Yes! Like, YC's all about doing things that don't scale, and this is one thing you can do as the founder! Like think about it this way: No big company in the world— the CEO is not individually emailing people to ask them to use the product! Like, I’m talking a real email from your Gmail account, not some drip email campaign. That oftentimes is enough, like not for everybody, from the public to re-engage some people. Another way to ask: We use methods like surveys to determine our product market fit in place of spending a lot of money to build a product. And then we realize we need to pivot or change features."

"Now the question—that’s fine! Okay, it sounds like a good idea! It's really good! Again, like as we were saying before for that other question—like not building an MVP is sometimes a good method for determining if people have the problem or not. If they have the problem, then by all means, go and create the MVP and get it out there!"

"Here's one: We're Metro Push. We're building a push notifications app for public service providers—electricity, water—who have a bad reputation for keeping their users informed. How do we avoid being perceived as a failure before we even get the chance to serve our users?"

"I think that this is one of those things that you can do even before you become—a even before you sell to one of these properties—public service providers—like this is something that you can run just by yourselves! Like, you can be the alert function for regular users of these public services! That’s how you could earn a good reputation. Like imagine if you started a Twitter account that would talk about the public state of the utilities in your city! People would begin to rely on that, and you'd become like a trusted name in place—then you would go to public services and say, 'We’ve already built this trusted service! You need to give us the good data so that we can, like, continue providing this.' Even on a higher level."

"Okay, here's one! Jemelle from Nuques Vision asks: 'Oh, the extrusion is the Google Analytics for retail stores. My question is, we're having difficulties to find our first users since we're targeting big live retailers like Walmart. Are we too small for them to consider us? How do we approach big companies with this type of an innovative product?' It's tough!"

"Oftentimes at YC, we have companies that want to immediately go after big companies like Walmart. We call this whale hunting because one big contract could result in hundreds of millions of dollars. So honestly, the best strategy here is to not go whale hunting at the beginning. Hopefully, your product is useful for other customers indifferent as well as kind of big retail stores. So I would try to go after customers who have the problem but are also small enough where you could get it immediately to someone who could make a buying decision after you show them a demo, or after you integrated into your product. You just want to get a paying customer who's giving you feedback fast! Going to big companies is just a recipe for disaster!"

"The reality is, unless you have some connection to Walmart already, it's unlikely that they're going to be your first customer. [Music] Michael from Biz asks: 'We're busy— busy two! We're busy. We show people's businesses around them and help those businesses receive payments.' Question is, is it okay to stop, rebuild, and relaunch if I'm not exactly pivoting?"

"Hmmm. So honestly when I hear questions like this from founders, I usually wonder if they have discovered something about their business that says that it’s not gonna work. That's kind of okay! Like oftentimes, as we're talking to early customers, we have this hypothesis about what we think the product market is going to be, and after a month or two of talking to customers, we realize that like no one actually has this problem or no one is—the problem is not as painful as we thought it would be. And at that point, like yeah, it’s fine, you have to consider other options, because like I don’t know who's with the quote, but if you keep—the definition of insanity is doing the same thing over and over again and thinking that it’s going to work. So yeah, like people can pivot; people can change. But you cannot—like you also have permission to say like, 'What I'm working on, it's just not the right thing.' And I don't know if that’s the case for yours—I might be extrapolating too much—but it’s possible that you may have actually learned about a better problem in the course of your doing user interviews."

"Over the last couple months, you may have discovered that like there's a better thing! There's a better problem that your users have to solve! Here’s one! I've run Gifted app—Gifted is a platform for users to manage their wish lists from multiple sites to shop on, and shop for gifts for friends. I'm seeing a lot of traction from high schoolers who use it to share things they love, but the audience won't spend a lot of money, isn't adding a lot of revenue. Should I keep doubling down on that demographic anyways?"

"I think you answered your own question! If your users are using it but aren't actually spending a lot of money, they probably don't sound like the people that you should be going after. I’d have changed the product so that it’s something that high schoolers can use, but then there’s also a way to generate revenue from it. Or maybe that's not the right demographic to work on."

"Okay, Cristina from List Alya asks, 'Is— that a why? Yeah, it is! Oh am I? Yeah! One question that I feel we're still trying to figure out in user interviews is focusing on what features they want when the scope of the interview is understanding the behavior rather than what we can build.'"

"Again, like that is—that's an incredibly important kind of thing to consider! As you're doing user interviews, I love to consider user interviews as like 15 minutes of time where I don't talk, and the users just tell me what their life is like. Because oftentimes, like, everyone has different lives, and we're in our own heads pretty often thinking about like what we do on a regular basis and how we would use the product if we were using it. But you really need to do with these user interviews is just let kind of the user's whole life flow over you. Where do they live? What do they do for a living? How often do they have this problem? Like, you just got to get all that context!"

"I treat those as kind of like wide open user interviews. So if they keep trying to pitch you solutions, you should try to refocus! Yeah! Like ask them why they're asking for that solution. Try to understand under the ask what’s driving them to have that problem. And then you can definitely guide the conversation towards problems that they have and solutions they think would be good for that problem. And then you build it when you give it back to them."

"Here’s a question: Elliot from Unify asks, 'How would you recommend successfully launching a big update for a mobile app? We're planning a large update that hopefully will convert more free users to paying users.'"

"I think you go through the entire launch process again; you launch in all those channels. And so, you know, you'll do Product Hunt. You can do—you know, Unified 2.0. You'll do Hacker News. You’ll always recommend companies do a blog post on their own as well, and they write about the update and what it means for their users and why they built it this way. But I would say go through and treat it like you know it’s your first launch and hit all the same channels that you would have hit the first time around."

"Do you have anything to add to that?"

"I would test your update with existing users! Treat some early launches as a beta launch where if you test it with like a hundred or a thousand free users, and if they’re not upgrading to paid, then you need to figure out what are the words that you can use to cause them to upgrade and use that! Like use that as almost like an experiment of a marketing experiment that you can then use when you launch publicly, when you launch the big update. Like, I always love to treat marketing like I train engineering—that it’s an iterative improvement process with experiments and hypotheses that I would then use to tune and tweak the marketing. Sometimes like the best marketing messages for launches actually come from your users! Like I write down actual sentences that users say, and then I just repeat them to other users and see if it clicks because oftentimes, like again, they are the people who have the problem or who use the app or whatever, and when they like—when they start using it, they get excited, and they may actually share words that you could then use to get other people excited as well!"

"Well, in asks, we empower yoga teachers to do events at your local home. I remember that from last week! We've quickly grown, forcing us to waitlist new instructors. Should we hire more to expand our current city, expand to a new city with a new MVP, or talk to investors to enable us to build an MVP into a platform?"

"So it’s great; congratulations on getting a backlog of new instructors! We've seen this a number of times with kind of these local startups! So let's identify the three camps. The first is to grow in your existing city, the second is to expand to a new city, and the third is to get investor dollars to build—to improve your platform. I think the answer is, it sounds like you already have a backlog, presumably in that first city that you're launching in. I would continue to work in that city because you—it's much easier to build an MVP with one location than to build an MVP that's live in all these different locations for a couple things. One is like usually you perfect it in one location, and then copy-paste it into another location. It's very hard to improve the product if you have two different locations or more alive at the same time."

"Adora has a lot of things—yes! It is! Started a Folkjoy. Another third one, talking to investors. Like if you can help it, try to get money from your actual customers to fund your business rather than trying to rely on investor dollars! I know that sounds weird as an investor; it’ll make fundraising so much easier in the future. That one, I mean, the trick to fundraising is to not need money from investors, then I just love to invest at that point."

"Okay, I think we can take two more questions. We are a serious platform for preoperative teleconsultations to prepare patients for surgery from home. My question is: how to launch fast in the healthcare vertical when you have some legal or ethical discussions for telemedicine in your region?"

"Okay, so launching fast! Breaking things is usually a good sign! Probably not in medicine. Yes, not in healthcare! We definitely don’t recommend—not if there's legal or ethical implications! Especially! Yeah, what I would consider doing here is thinking about your product as an evolution or an extension of the other products that are already on the market. So in your market, if there are telemedicine or teleconsultation products already there, then maybe—maybe one thing to do is to approach those apps and say, 'Hey, I would like to add on a module that is specifically around pre-op teleconsultations. Could we use your app to do that?' In the US, there are a few HIPAA-compliant apps that are kind of like platforms that you can use."

"I think—like you could probably try going to them and seeing if you could like use their entire text back and just copy and paste it into your application if you're worried about HIPAA and that kind of stuff, which they should be! Again, kind of tied to the last point: focus on one region here so that you’re not like balancing the ethical and legal ramifications in a bunch of different regions! But I think that sounds like what you—"

"Mhmm, already! Okay, this one is an audio-to-text platform that transcribes speaker sessions at conferences in real-time to make them more accessible. Cool! Question: I mentioned Eric mentioned that pursuing large customers might be the best—I don't know if I mean—so the question is: like, pursuing large customers— is that good for startups? Actually, it's probably not always the best for startups! Here’s another product question, however, if a person pursues larger customers, they have a lower probability of converting and longer sales cycle— isn’t there an opportunity cost to not try and spend time to convert smaller customers?"

"Yes, the answer is—that's exactly the problem! I'm not sure where I said that pursuing large customers is great, but I might have misspoke! Honestly, as we mentioned before—for the answer to another question, trying to get smaller customers who can make decisions quicker to buy or use your product is generally better, especially as you're iterating the product-market fit in the early days! Yeah, waiting for those big customers might kill you!"

"Cool! On that note, we are gonna end this week's live Q&A! Thank you very much for all those questions! Keep them coming on Twitter and on the forums. We'll have another session next week! Actually, next week’s session is going to be live design reviews with Aaron. If you want to have an amazing designer review your landing page or website, look out for the forum thread on Monday. Just post your website in the forum, and we will do a live review where we will like look at your webpage and talk about what we see as we do this next week. It's gonna be a long session—two hours! So we will have, I think, more than enough time to cover a lot of sites. I'll see you next week! Thanks!"

"Thanks!"