Jan 24, 2020 · 54m · y-combinator
YC SUS: Michael Seibel and Eric Migicovsky discuss How to Launch an MVP · Y Combinator
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this Y Combinator Startup School live Q&A, host Eric Migicovsky and partner Michael Seibel address real founder questions on designing, launching, and iterating Minimum Viable Products (MVPs). They deliver actionable advice on early pricing, unscalable customer feedback, managing technical constraints, and pitching early-stage startups.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The partners hold 42.6% of the talking time here. How this is scored →
The partners as informed peer 7.0 Guest teaching 0.3 Guest disagreement 0.0 The partners pushing back 1.1
speaking balance: gold is the partners, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 17:45 Dismissing flawed founder assumptions about car servicing
Michael bluntly challenges the founder's business idea by pointing out that numerous YC startups have already failed attempting this exact model without realizing its unit economic flaws.
Hardest push from the partners ▶ 7:18 Pushback on the paper-to-software startup mythMichael refuses the popular assumption that taking a workflow from paper to software automatically creates a viable billion-dollar company.
Biggest teaching moment ▶ 49:46 Eric educates on the misconception of startup patentingEric explains from his hardware startup experience why non-biotech founders excessively focus on IP protection instead of securing real product traction.
The partners hold their own ▶ 51:53 Michael synthesizes 15 years of platform risk dynamicsMichael breaks down the structural shift from open web protocols to closed platform ecosystems with deep historical and economic context.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The partners as informed peer | Guest teaching | Guest disagreement | The partners pushing back | Why |
|---|---|---|---|---|---|---|
| Zephyrus: Building MVPs in Established Markets | 7 | 0 | 0 | 2 | Michael immediately draws on his Justin.tv experience to address whether scrappy MVPs work in established markets, clarifying that users who are in enough pain will accept imperfect software. Eric complements this by suggesting building browser extensions on top of existing tools. Both speakers are completely collaborative. | |
| Flatro: Avoiding Over-Customization for Early Clients | 7 | 0 | 0 | 1 | Michael explains the cost of learning an industry and pushes back on the assumption that simply digitizing paper workflows creates a viable venture business. Eric adds advice on doing unscalable manual work to understand user workflows. The dynamic is purely collaborative advice-giving. | |
| Fridgeworthy: Launching User-Generated Content Platforms | 7 | 0 | 0 | 1 | Michael uses Harvard-era Facebook and TechCrunch/Crunchbase examples to illustrate seeding UGC networks with small communities. Eric builds on the point with examples from 9GAG and Reddit. Both operate as aligned startup advisors. | |
| Swish Technologies: MVPs for Hardtech and Long-Cycle Projects | 6 | 1 | 0 | 1 | Eric takes the lead given his experience advising hardware and biotech startups at YC, citing Embark trucks and off-the-shelf prototyping. Michael supplements with advice on securing pilot LOIs before full vehicle launches. | |
| Volley: Quick Launches vs. Legal Liabilities | 7 | 0 | 0 | 2 | Michael directly questions the founder's premise by noting how many on-demand car maintenance startups have failed and pointing out pricing dynamics with price-sensitive car owners. Eric agrees, emphasizing caution regarding risk and liability. | |
| Bin Stop: Free vs. Paid MVPs for Service Marketplaces | 7 | 0 | 0 | 1 | Michael strongly argues that operations with marginal human labor costs must charge from day one to avoid artificial demand subsidization. Eric reinforces this by urging the founder to price higher than incumbent studios to test genuine pull. | |
| High Five: Differentiating Social and Event Platforms | 8 | 0 | 0 | 1 | Michael cites Peter Reinhardt's YC talk about identifying company skeptics and operating under the assumption that the startup is dying until proven otherwise. Eric expands by discussing Superhuman's PMF engine and testing via service outages. | |
| Local B: Knowing When to Pivot or Pursue Fit | 8 | 0 | 0 | 1 | Michael advises time-bounding iteration cycles to six months rather than relying on subjective feelings about having tried all ideas. Eric provides practical techniques such as whiteboard sprints and landing page experiments based on TaskRabbit's history. | |
| Siteful: High-Bandwidth Feedback Methods After Launch | 7 | 0 | 0 | 1 | Michael warns against premature optimization of automated helpdesks, advocating unscalable methods like direct phone calls and WhatsApp groups. Eric explains that wide-bandwidth communication is necessary when founders don't yet know what specific questions to ask. | |
| Car Nuba: Pitching Current MVPs vs. Vision | 7 | 0 | 0 | 1 | Michael outlines a four-step pitch framework starting with current reality and traction before introducing long-term marketplace vision. Eric agrees, pointing out that vision-only pitches offer no concrete metrics for evaluating founder competence. | |
| Live Q&A: White-Label Services vs. Simple Tools | 6 | 0 | 0 | 1 | Michael explains that white-label platforms or Google Sheets are sufficient for initial demand testing, but proprietary software is essential if building a venture-backed tech company. Eric notes regional context variations in India. | |
| Live Q&A: Charging vs. Free Features for Education Apps | 7 | 0 | 0 | 1 | Michael and Eric advise on monetization choices, warning founders against asking non-customer startup mentors for product advice. Michael also provides a structured approach to resolve co-founder conflict through experimental milestones. | |
| Live Q&A: Pitching Beta Products to Early B2B Clients | 8 | 0 | 0 | 2 | Michael debunks the myth of early Uber discounting, clarifying that Uber started as an expensive luxury solution solving an acute shortage in San Francisco. He and Eric stress that early beta access is a high-touch selling point, not a weakness to hide. | |
| Live Q&A: Patenting Technology Before Launching MVPs | 5 | 3 | 0 | 0 | Michael defers patent expertise to Eric, who has extensive hardware background. Eric explains why early patent focus is usually a distraction from the real existential problems of finding users, product, and revenue. | |
| Live Q&A: Managing Platform Dependency and Risk | 8 | 0 | 0 | 1 | Michael delivers an analysis of platform risk, detailing how the internet transitioned from open protocols (SMS, email, early Google) to closed walled gardens that actively eliminate or acquire threatening platform dependencies. Eric provides Periscope and Vine as historical examples. |
Statements from this episode (34)
Seibel: Needing a polished MVP means incumbents are doing a good job
“If you can find a small population of people that have enough of a problem that the other folks aren't solving, that they're willing to use something that's pretty bad, that's a very good sign. If the only way you can find any group of users, even 10, even a h…”
Migicovsky: Startups should test hypotheses by layering MVPs on existing products
“One of the hacks that I like to use at this stage is to actually layer the MVP onto the existing product. So for example, if there's already some software that your users, not your users, like the people in the community are already using to track their compet…”
Seibel: Founders without domain expertise must build bespoke features to learn
“Just learning the business means you're gonna be building a lot of features that probably are not generally applicable, and that's the cost of learning about this business.”
Migicovsky: Target problems customers are already actively attempting to solve
“Those are usually the problems that I like to start with because Like the person's already trying to find a solution. So they're primed when you come to talk to them, like to be excited and to try it versus you having to like pitch them hard and like changing …”
Seibel: Digitizing paper workflows alone does not guarantee a billion-dollar company
“If something's being undone on paper, There is a business in putting it into the computer, and like, by nature of the fact that it's being done on paper, you could build a billion dollar company and bring it to the computer, and that's actually not true. Like,…”
Migicovsky: 9GAG bootstrapped early content by scraping memes from other sites
“A lot of the best user-generated content networks, like, Ninegag, which is now this, like, massive Meme, meme website was started by, like, scraping memes from other sites.”
Seibel: TechCrunch originally seeded Crunchbase profiles using media reports
“Before people cared about their Crunchbase profile, TechCrunch just built the profiles on Crunchbase. With media reports and so on and so forth, and then after a while, people cared about their profile, and so they tried to keep it updated.”
Migicovsky: Top hardware startups begin by assembling off-the-shelf components
“Some of the best hardware companies that I've worked with, even though they now make their own hardware and they build a bunch of custom stuff. They started using, they started by using off-the-shelf systems like Raspberry Pis, cameras, just existing technolog…”
Migicovsky: Autonomous trucking startup Embark began with a camera-rigged golf cart
“So, we've had companies that are now off building, like Embark, which builds, ah, self-driving trucks. Like, literally, they're running on the streets and the highways in Nevada. They started with a club car. I think, like, at demo day, they had a golf cart th…”
Seibel: Securing non-priced LOIs proves demand and sales capability to investors
“Generating letters of intent or agreements to pilot shouldn't be very hard if there's a lot of demand for this sort of thing. So being able just to get your customer to sign anything, even if it doesn't have a price on it, is evidence that you can sell to an i…”
Migicovsky: Startup one-liners should explain current product rather than all-encompassing visions
“Here's actually a problem that a lot of founders run into, is they think that their one-liner or their short description has to encompass, like, everything that they'll ever do under the ends of the earth, because they think that, like, investors only will get…”
Seibel: On-demand car repair startups struggle because frequent break-down customers are price-sensitive
“The population that has the most problem with their cars tends to also be the population that is extremely price sensitive about fixing their cars. And so make sure that you have some insight on how you can be offering the cheapest option. Cause if you're the …”
Seibel: Startups with real marginal costs must charge from day one
“In any situation where you actually have real cost, you should be charging from day one.”
Seibel: Subsidizing unit economics creates artificial demand that vanishes when prices rise
“If you're into a situation where you're providing a dollar worth of value and you're only getting 75% cents back, you're gonna get a lot of demand. But you're gonna lose a lot of money, and when you actually equalize the demand and the price, you'll lose a lot…”
Seibel: Founders should charge uncomfortable prices to find desperate customers
“Charge more than you might feel comfortable so you can identify the people who are most desperate for it.”
Seibel: Founders in crowded spaces must know why previous products failed
“Make sure that you're studying those products and you have a strong opinion about why they suck. It's going to help you two reasons. One, it'll help you figure out what to make as a good product. And two, A very common investor question, if you ever want to ra…”
Migicovsky: Build MVPs around the smallest action with outsized impact
“The way that I would start building an MVP is kind of like what we mentioned at the beginning. It's taking what you already do on a regular basis, like maybe you go to meetups, maybe you, like, are on Reddit for the subreddit of that city, and, like, think cle…”
Migicovsky: Founders should iterate on target market before adding features
“Instead of iterating on the product, like building more features, they actually iterate on the market.”
Seibel: Time-bounding startup attempts to at least six months clarifies pivot decisions
“I think the best strategy for this the one that's easiest to execute is to time bound it. So what I would say is like, Hey, we're going to try everything as hard as possible to make this work for this period of time, probably no less than six months. And if we…”
Seibel: Founders should establish demand benchmarks that convince company skeptics before continuing
“Identify a critic in the company. Identify a sign that would either, that would even convince the critic that there's something here. If you can't hit that sign within a period of time, you should move. So, like, don't trust The faith of the optimist, like, tr…”
Seibel: Early Airbnb customer service routed to Joe Gebbia's cell phone
“Another famous one was Airbnb where customer service was one of the co-founders Joe's cell phone number.”
Migicovsky: Early startups need unscalable feedback because founders do not know what questions to ask
“The reason in this case is actually you don't know the questions to ask yet, and so you want a really high bandwidth A really wide, broad way of collecting feedback, because you don't know the questions that you want to ask yet”
Seibel: Pitch current product and traction before long-term vision
“So here's how I would probably structure the pitch. Step one, what you're working on now. Step two, the traction. Step three the business model, and then step four, the vision. It's a lot easier to sell me if I know what the hell's going on, and those first th…”
Seibel: Scaling and raising capital is easier when building proprietary software
“If you have seen demand, and you're thinking to yourself, okay, I want to make this a company, if you want to make it a software company that can scale, that can raise money, it's a lot easier if you're building your own software.”
Seibel: Founders building a lifestyle business should start charging immediately
“If you're looking to build a lifestyle business, then like I would definitely charge right now.”
Seibel: Non-customer startup advisors cannot provide real customer insights
“Be very careful about asking any startup person who's not your customer for customer insight, because they won't have them.”
Seibel: Resolve co-founder strategy conflicts with time-bound empirical tests
“Ask him, okay, what's the test that would prove that the direction we're going in right now is a good one, and what is the time period that we can all agree with to go after that, and then bam, we're running an experiment now. If we don't make it, we'll switch…”
Migicovsky: Developers prematurely shoot down ideas before evaluating potential upside
“I think too often, especially, like, developers, we're already internalizing the, like, How difficult things are going to be to build when we might not have understood, like, the potential benefit from building this, and it's always a trade-off. So, like, go b…”
Seibel: Early B2B customers should value founder access over polished software
“You're looking for An early customer who is excited by the opportunity to work really closely with the founder, to be able to give feedback quickly, to be able to get changes quickly, and is not intimidated by something that's going to be broken.”
Seibel: Asking investors for clarification shows sophistication rather than ignorance
“It actually shows a level of sophistication to ask the question when you don't know, as opposed to just kind of, like, nod as if you do know and not, and then leave without any insights.”
Seibel: Early Uber avoided rider subsidies due to immense natural demand
“So in the extreme beginning of Uber, what was so interesting was that it launched in San Francisco, and it launched as Uber Black, which is a really expensive product, actually. It wasn't well subsidized. And the real value is that it was impossible to get a t…”
Seibel: Uber would have failed if it initially launched in NYC
“If Uber had launched in New York City, it wouldn't have worked. Because, like, the pain did not, yeah, did not exist in New York City. The pain was so great, great in San Francisco that everyone was willing to use Uber Black.”
Migicovsky: Patents are not a critical initial step for startups
“This is, like, one of those things that people who are not in the startup world think is a critical piece of starting a business, because they may have heard of, like, someone else who got screwed out of their idea or something like that. But if you're startin…”
Seibel: Dependent startups face existential risk from their host platforms
“And so the problem is, is that you're not only competing with other players in your market, you're competing against the platform you're on. If this is a significant revenue opportunity for the platform you're on, they have every incentive to either buy you or…”