Why Startups Fail with Tom Eisenmann
In this episode of Product Thinking, Tom Eisenmann, Professor of Business Administration at Harvard Business School, joins Melissa Perri to dive into "Why Startups Fail," his transformative book. Specifically, they dive into the inspiration of the book, six unique types of startup failures and how to avoid them, as well as strategies for hiring your first product manager.
Episode Transcript
Melissa Perri: Welcome to another episode of the Product Thinking Podcast. Today, we're joined by Tom Eisenman, and we're going to be talking about why startups fail, which is a name of his transformative book. Tom has been teaching at Harvard Business School for over 26 years, and he's actually the driving force behind their product management class, which is what I taught there.
I've learned so much from Tom during my time teaching at HBS, and I'm really, really excited for you guys to learn from him as well. His expertise and his research really bridges the critical worlds of academia and real-world entrepreneurship, and it makes him a wellspring of knowledge for anyone navigating the complexities of product management and startups. But before we bring in Tom, we're going to change course a little bit.
So I've been getting lots of questions from you on our Dear Melissa segment, and they have been fantastic. So I want to be able to answer them more quickly, and we're going to start answering one question at the beginning of each one of our episodes, then followed by a guest. So this week's question is very much in line with what Tom and I are going to be talking about when it comes to entrepreneurship and product management.
So let's head over and see what we had written in. Here's our question. Dear Melissa, I have around three years of experience working as a product manager, always with mature products or at least immature companies. I've joined now to a company where I'm the only product person, and the product is not in the market yet.
And there is no product culture at all. This is kind of a side job. What advice could you give me? What would you focus on? What steps would you follow? So being the first product manager in a company is definitely a challenge, especially at startups. So when you come into this company, you are going to be responsible for creating that product culture.
You're going to be responsible for managing the founders and working with them closely. And then also setting up the systems, because usually there's going to be no systems whatsoever. So understand that this is a big task. The first thing is really figuring out why did they hire you. So talk to the founder.
What did they need? Did they need somebody to actually execute, work with the developers to get things done? Did they need somebody to come in and actually stand up some of those processes? What are they really looking for? Because some of the founders may not want you to act kind of like ahead of product, do all the vision, do all the strategy.
They might need somebody who just can dive in and get work done. And that's totally fine. Other times, they're going to need some help with structure. If you have three years of experience, you can usually pull this in, which will help them a lot. So you might have to start thinking about what kind of things are going to make us go faster.
Do we have a place to actually write out tickets? How are we writing user stories? Do we have a roadmap? If not, can I actually suggest a roadmap with a template? How do people talk about product management? How do we review what we're going to build? Is there a prioritization framework? You're basically starting from scratch.
So I need you to think through what are all the things that worked really well at the last company and then also what's going to be different here. Since you worked for a mature company, all of those things were probably in place when you walked in and you took them for granted. And that's fine. That's what happens with mature companies.
But in the case with the startup, usually it's like everybody's running around like chickens with their heads cut off, not knowing which way to go. And you have to bring some of that standardization in, but not too much. In mature products, we need a lot of overhead because we usually have lots and lots of teams.
You probably don't need everything because you're a small team. It's just you, right? You and the founders. So think about what's going to make our team go faster. How do I get my hands dirty? How do I get things out faster? How do I test stuff? How do I make sure we're doing really good customer research and focus on that core, especially when you're in pre-product market fit and it's not out there yet.
You're going to be using some of those lean startup techniques. You're going to be doing a lot of good customer research, but make sure you're documenting it. Make sure you're recording it. You're probably going to be the one coming up with where those things live. So think about a little bit into the future.
If we scale, how do I make sure that we scale well? If I bring in one or two more people, how do I make sure that we can all collaborate really well together? And then how do I make sure I'm on the same path as a founder, the same track? We're talking, we're collaborating, we're doing things together.
That's really important. So usually at this stage still, your founder is going to be the head of product. You got to figure out what they brought you in for. Was it just execution? Was it somebody to bounce ideas off of? Try to figure out what the purpose is and then work your systems from there and be lightweight about it.
Don't go crazy. Don't make a lot of overhead that is not necessary for a very tiny team and just really think about how do I get to product market fit and what's the best way for us to actually test that. So I really hope that helps and I'm sure we're going to hear a lot more from Tom as well about what people should be thinking about in startups of this size.
Welcome, Tom. Thanks for joining us today.
Tom Eisenmann: Hey, Melissa. Thanks for having me.
Melissa Perri: So you have been teaching at Harvard for over 26 years. You've been working with a lot of the students who want to be entrepreneurs. You've seen lots of startups come out of Harvard. And two years ago, you wrote this book called Why Startups Fail. What led you to write the book?
Tom Eisenmann: Yeah, so the catalyst for the book goes back 10 years. I was maybe even a little longer than that. We have a required first year entrepreneurship course at Harvard Business School. And it's taught to all 900 students and taught by the case method like almost everything at Harvard Business School. And we would teach 30 sessions and students, we get feedback like good product manager.
You ask the customer, how are you doing? And a consistent feedback we got from the course was, hey, you tell us along the way the two-thirds of startups fail. But in this course, we saw 30 brilliant successes. So these founders come in. We bring the protagonist in very often when we teach the case. And they strut around and show the peacock fellers and so forth.
And so the students want to know, isn't there something we can learn from the failures? And that sounded reasonable to me. So I set out to add a failure case to the course. And that was a failure. I was a failure at teaching failure. Turns out I taught the whole story beginning to end, including the founders post-mortem reflections.
And you will remember that MBAs are really good at looking in the rear view mirror and sort of, oh yeah, it's obvious why that failed. There wasn't demand and they did this wrong and they did that. Yeah, it was all in the case. But look, it's a really smart person who was backed by really smart investors.
So there must be something we can learn. So I learned that you actually have to take the story up to the point where there's still a plausible path for the entrepreneur to rescue the business. And that actually works. So anyway, I headed in that direction. But the real catalyst was 10 years ago, former students of mine launched a startup.
The startup was Quincy Apparel. The concept was these were two very tall women who had trouble finding, they went to work as consultants like so many MBAs after business school, had trouble finding clothing for work that was affordable, stylish and fit them. You can get two of the three. It's very hard to get all three of them.
So they wanted to create a startup that would provide better fitting, stylish, affordable clothing. Quincy Apparel, they raised a million dollars, tried to raise a million and a half. That's part of the failure story. I was an investor in the business. And these entrepreneurs, I had really just started to master and understand lean startup practices and pull them into the MBA curriculum.
And so I pushed the team to do the minimum viable product trials and go out and do customer discovery, textbook perfect. They did everything you would want founders to do. Validated demand. And sure enough, launched the business. The demand was there. It was always strong, good repurchase rates and so forth.
But they couldn't get the operations under control. They didn't have prior experience in apparel. They didn't have enough money. And the investors who backed them wouldn't put more in. So the thing failed after just a year. And I looked at it and like, wow, they did all the stuff I tried to teach them really well.
I can point to a lot of things that went wrong, but I'm not sure I can pinpoint the cause. So again, I'm a failure at explaining failure. And this is really like the, I'm supposed to be an expert on entrepreneurship. This is probably the most important phenomenon in my field. And so that put me on a really a 10-year path to learn everything I could.
Talked to probably a hundred failed founders and lots of investors who backed them. Read everything anybody had ever written on the topic. Did a big survey and all that culminated in two things. The book that you mentioned, Why Startups Fail, but also an MBA course, which is a different story. It was a little scary to launch a course with wall-to-wall failure in it.
I thought I might bum the students out. That turned out to not be the case. So that's the genesis for the work.
Melissa Perri: That's great. So in the book, you start to talk about six different things or six different types of failure and why startups get there. Can you tell us what those are?
Tom Eisenmann: Sure. I'll take them in two chunks. So there's three early stage failure patterns in the book and three late stage. And by early stage, I mean essentially pre-product market fit. And late stage, you've got product market fit, we hope, and you're scaling. You may be scaling without product market fit.
And that's actually one of the failure patterns. And the argument of the book is a lot of startups fall into one of these six patterns. Some have the misfortune of actually hitting more than one of them. And that's surefire prescription for getting into trouble. The early stage patterns are false start, bad bedfellows, and that's the Quincy example I just gave, and false positives.
So I'll explain each one. As I do, since the listeners are really interested in product, there'll be a lot of product themes that pop up in some places where a better product manager or a product manager at all, in a lot of cases in an early stage firm, the founder's really driving product. And that is as it should be.
But at some stage, you're going to bring in a PM or a product leader. And in some cases here, that individual could have made a huge difference. So false start, just like track and field, or swimming, where the athlete literally jumps the gun, hoping to get an edge and gets penalized. Here, what happens is the entrepreneur is so eager to get the thing out in the world, to make it and sell it, that she skips an important chunk of upfront research.
The lean startup folks call it customer discovery research. It's essentially doing everything you need to do to figure out if you've got a real problem, a hair on fire problem worth solving and the right solution for that problem. Defensible, differentiated, differentiated in the sense that nobody's doing it and you're doing things a lot better than other people.
And it's pretty easy to see, this is probably the number one killer of early stage startups is skipping this step. And it's a bad trade. If you think about typical software business, maybe it takes four months to build the thing, put it out there, see if it's working and then figure out what to do if it's not working.
That's a pretty typical cycle if somebody's agile and moving fast. And the customer discovery week, in a lot of instances, customer discovery work can be done in about a month, four weeks. And so what this entrepreneur has done, they've made a bad trade. They've essentially, in order to save four weeks, they have probably wasted four months.
I mean, sometimes they'll be, they'll nail it. They'll get it right on the first try. But very often, because you've skipped the research to figure out if you've got the right problem and the right solution, the first go is flawed and you've essentially wasted four months. And if you've only got enough money to last for a year or 18 months, I mean, that really raises your odds of failure.
So that's the problem. And it's completely understandable why it happens is, right? What is an entrepreneur if not somebody who's got a bias for action? So what's better action than get out there and make it and sell it? And then you'll appreciate this. I mean, we have in the MBA program a lot of students who are not technical.
They hear correctly that to succeed as an entrepreneur, you need great product. How do you get great product? You can have a great engineering team. How do you get a great engineering team? You use those networking skills that you polish in an MBA program. They're really good at sort of persuading people that those people should join their parade.
And, you know, even with a tech, so you hire the engineers or you bring them on board as co-founders and to keep them busy, what do engineers do? They build. So you give them something to build even though it isn't thought through. It actually, this false start pattern impacts technical founders too, because what do they love to do?
They're engineers. They love to build. So in both sides, you end up prematurely diving into the engineering work.
Melissa Perri: Yeah, that seemed to be a really common one that I would see, you know, and in our PM101 class too. We had, you know, you made the PM101 class. So it starts with trying to get people out there and do the customer discovery. But everybody's like, no, we just want to build. We don't want to go talk to people.
But I think what you said was really interesting too, where a lot of times you can do this in just a month. So do you find that it's, when you were, you know, researching these founders who failed this way, did you find it was a kind of like an arrogance or like a bias saying, oh, I already know what the problem is? Or do you think it was just like a laziness of, I just really want to build it.
That's what we need to do.
Tom Eisenmann: I don't think it's, laziness is the wrong word, eagerness.
Melissa Perri: Eagerness, okay.
Tom Eisenmann: This is a zeal to get going. And it feels so natural for an entrepreneur. You know, it feels like you're studying, you're back to school, you're doing research. And so that's it. There is some arrogance. Particularly if you get a founder who's got domain experience, right? They, I know this field. I know this market.
And sometimes they do. And sometimes you can dive right in. But, you know, even, it's very often the case that if you think you know it, there's still some subtle differences. By definition, an entrepreneur is trying to do something new. So almost by definition, the thing you want to do, even if you think you know the market, the problem and the solution, there's some aspect that you need to think through.
It's often the case that when somebody gets whacked by the first version of the product failing, they will come back and do the work. You know, a little humble. But it's also true that sometimes take two, sort of the second try is just an echo of the first try. Namely, okay, you know, I'm going to try something different and build a new thing and put it out there.
Again, without sort of thinking through whether it's going to work.
Melissa Perri: Yeah, that's definitely the common one that I see, especially in product management. Okay, so we've got the false start one.
Tom Eisenmann: Yeah, and so you can see, just to sort of bring that back to product, and then I'll go on to bad bedfellas. Bad bedfellas is probably the one where the lack of a PM or a product leader is most acute. But you can see on a false start, if you have a well-trained PM in your midst, or if a well-trained PM is the individual who is the founder or the entrepreneur, it really reduces the odds of the false start.
I mean, a good PM is going to know how to basically listen to the customer. And so bad bedfellas is, when we teach about entrepreneurship at Harvard Business School, we look at four different, we call them resource providers. There's the founders themselves, there's the rest of the team, there's outside investors, and then there's often some kind of strategic partner.
It's a channel partner for marketing, somebody who's providing essential technology. And the bad bedfellas pattern, in some ways it's almost the photo negative of a false start. You can have actually a pretty good team, but they're just so far from the right starting point that even if they pivot and pivot and pivot, they may never get there.
With bad bedfellas, and Quincy is an example of this, they actually had a good idea and they had validated both the problem and the solution. They just couldn't deliver the solution, they couldn't execute. And the failure of execution, in Quincy's case and in a bad bedfellas failure, is a failure, dysfunctional resource providers all the way around.
I'm drawing a square because at the business school, we use it, you'll remember the diamond and square framework, diamond as elements, the business model, the square surrounds the diamond, it's the four resource providers. So look at what happened with Quincy. So the founders had never worked in apparel design and manufacturing, and it turns out to be a really complicated process.
Many, many steps that have to fit together well, done by specialists, they figure this out reasonably quickly, thought they could just hire a few people to do all the functions, but that's not how it works. In a big apparel company, you are the pattern cutter, the fabric source or the quality control person. And in different companies, these people work together in different ways.
So Quincy had to hire people to do these jobs and hire them from established, from big apparel companies, people who had never worked in an early stage startup where you have to sort of scramble around and sort of help put out whatever fire is burning hottest. The fabric cutter would say to the quality control person, I don't know how to do your job. I can see you're backlogged.
I'm not going to help. They would sort of sit on their hands waiting for the work to arrive to them. So bad team fit because they didn't adjust to the rhythms of an early stage startup. The founders lacked experience, which meant they had trouble actually finding people. And then once they thought they found people gauging the quality of those people, they made a mistake that I think a lot of MBAs from top schools make, especially from the ones we work with at Harvard Business School.
They just sort of assume if you're going to be an entrepreneur, the only respectable way to raise money for your venture is to go to venture capital firms. And it just never occurred to the Quincy founders that there was other money out there and there might be better sources of funding. And the problem with venture capital, as many of your listeners will have experienced, is the whole model is we're going to make a huge amount of money on a small fraction of the companies in our portfolio that are going to sort of rocket ship to the moon style.
And we're going to push every single company in the portfolio to achieve that kind of result, knowing that most of them are going to fail. But if just a couple, just as a modest percentage work, we will make a lot of money. And that was the expectation when Quincy raised from venture capitalists. They don't run out of inventory.
The stock out is the worst thing you can have if you're trying to grow sales fast. Well, if you've got fashion and you're sort of trying to figure out if it actually meets the needs of the customer, the style preferences, you don't want to be sitting on a lot of inventory that doesn't work. So they raised from and VCs don't usually invest in apparel lines.
So they ended up dressing up the business, which it was as direct to consumer, which goes back 12 years was a new thing then. We take it for granted now, but Bonobos was just getting started. And I don't know, Warby Parker probably had just launched. And so VCs were looking for, I need a DTC company in my portfolio.
And Rent the Runway had just launched and had a lot of success. So here's a pair of female founders from Harvard Business School who look a lot like the Rent the Runway founders. So maybe lightning will strike again. So they managed to raise, but from second tier VCs who didn't have a lot of money and couldn't bridge when the company got in trouble and frankly couldn't offer particularly good advice.
So bad bedfellows, founders lacking domain experience, team wrong cultural fit, investors who really are pushing in the wrong direction and can't add a lot of value. And then strategic partners, they had the apparel, as is often the case for a new apparel company, outsourced the manufacturing to third party factories. And of course, these factories, you got this little peanut an order from Ann Taylor comes in that needs to be expedited.
So the Quincy stuff got pushed to the end of the line by the factories. So all the way around, even though the idea was good and what killed the company was it took a long time to sort out the operations and deliver with quality, the promised good fit. Turns out there's a reason why women's clothing isn't...
The idea for Quincy was we're gonna sell it in sizes like men's clothing, where you've got chest size and sleeve length, and et cetera, et cetera. And I'm sure other people in the history of apparel have thought of that. And it's actually hard to manufacture. They were making progress, but they ran out of money before they could get there.
So bad bedfellows. And here again, you can see that the apparel equivalent of a great PM, somebody who could organize that whole process could have made a difference. Franklin, maybe it's not a classic PM role. It's really just an operations role, but they didn't have that. It sounds like, yeah,
Melissa Perri: It's that industry knowledge. It feels like without that, without knowing somebody who can own those operations end to end on that side, it feels like that piece was lacking. And that's an interesting one because in product management, we go through this debate all the time about whether or not we need a product manager with industry experience or we don't.
And what I come back to tell usually founders or teams you have to look at the whole team and see if there's industry experience in the company. And if there isn't, yes, you're going to need somebody with good industry experience. But if there is, and you've got a PM who can learn, usually you can go back to that.
But with a founder, you've got two people. And if they both don't have industry experience, it sounds like that's a hard one right there. Yeah. So I agree strongly there.
Tom Eisenmann: Two insights came from this story. One is the Quincy founders, they had these jobs at BCG, Boston Consulting Group. And one was married and had a spouse who could help cover the living expenses. The other wasn't and hadn't saved, was sort of sitting on a bunch of student loans and didn't have family money that could cover this.
So the minute she quit her job, they had to go raise money and get started. If they had stayed at BCG, they were working on the venture nights and weekends. If they had just taken more time to learn about the business, they might've been able to avoid some of the problems. So that's one thought. The second thought sort of relates to what you were saying is, do we actually need domain experience?
And what I learned from this research is the answer to that with a founder is it depends. And what it depends on is the complexity of operations in this instance. I'll give an example. In food and beverage, it turns out there are dozens and dozens of decisions that a lot of MBAs are attracted to launch a new food or beverage product.
And the decisions, should we pay for end of aisle displays? Can I trust this co-packer to actually deliver quality? What should the labeling on the product look like? Will this wholesaler, wholesalers are notorious for abusing new brands and sort of not paying them the money that actually is coming from the retailers.
Just up and down the line, so many mistakes you can make as a founder if you're doing food and beverage that having domain experience. So apparel, but then you contrast it to something like think of Instagram. The founders of Instagram did not need to have worked at Kodak or Polaroid for 20 years in order to have insight on photo sharing.
So sometimes it's much more important than other times.
Melissa Perri: Do you find like it's something to do with, you mentioned it's a complicated operations piece, but do you think with the Instagram stuff it's because we would use it or they would be consumers of the same app? I feel like there's a tricky thing there where people go, oh, we are the users and then they don't want to do the customer research and they don't want to go out and do that first part we were talking about, but it does probably make it easier to launch a business like that from a domain experience side. Yeah, I haven't studied it closely.
Tom Eisenmann: I suspect there were two things going on. I suspect that the founders were millennials who were online native, right? And sort of had a lot of personal experience and could sort of look over their shoulder and see a lot of people photo sharing. So that was probably more relevant than 20 years at Kodak when people were printing out, film was expensive, developing was expensive, you got photo albums and that whole experience.
I'm sure there's some valuable lessons for translating it online, but going mobile was such a sea change that I think whatever domain experience related to the real world management of photos maybe was less crucial to the success of these early mobile first applications.
Melissa Perri: Yeah, I think that one makes sense. So I was actually at Kodak right before Instagram launched and we were part of this innovation team that they brought in because they wanted to understand more like what were the younger generations doing with photo sharing and it was a couple years before. It was right after the iPhone launched, so like definitely before Instagram.
But something else I saw there was people understood conceptually that everybody was starting to get away from digital cameras and start using their phones and didn't want to carry it around. But because of the inertia of the big company, they couldn't do anything to actually act on it. So like we came up with all these concepts and we said it'd probably be better for you to take your software for photo editing and embed it on something else instead of like building a phone out or building these things out.
Instead of making more digital cameras, let's say it that way, just like go for the software, go for like put your camera lenses on phone, something like that. And they just couldn't act, right? Like they had this innovation team built in but we just couldn't act. Like there was no budget for us. There was no way for us to actually get the ideas out there.
And everybody was kind of like, no, this is the way it's always been. This is how we do things here. That's not how we do things.
Tom Eisenmann: Yeah, there's such strong embedding of just a way of thinking about the business and what's important to the business and the business model of selling film, which was so profitable, huge profit margin for so long. There are echoes of that at Kodak all the way back to the creation of digital cameras. They had the technology for digital first and it was very good.
And they watched as the Asian manufacturers sort of pushed the stuff out so much faster and at such a lower price point they fell behind. In another life as a management consultant, I had a chance to work for a big phone company that was exploring the use of two-way cable, the cable television system could carry signals in both directions and we take it for granted now we're almost all using cable modems. That technology was new and they recognized it could be used to do video phone.
And we spent years developing plans for launching cable delivered video phone and they could just never push the thing over the finish line. And 20 years later here we are essentially doing what we were looking at 20 years ago. So sometimes it's too far out. Sometimes it's too big a change in business model or too big a cultural change and lots of inertia as you say.
So it's just a kind of like good technology
Melissa Perri: At the wrong time, right? Too early for the market and that could be something there too. Some of the listeners will be old enough
Tom Eisenmann: To remember Webvan, which was the first go at online grocery delivery blew through a billion dollars. It's a later stage failure of a type I look at in the book. We take online grocery delivery for granted now. Sometimes things are just ahead of their time. Yeah, like Vine and TikTok
Melissa Perri: That's like a very relevant one these days.
Tom Eisenmann: Or Starlink. Speaking of Elon Musk, Iridium was the first type sort of circle the planet with low orbit satellites that can reach provide connectivity anywhere was a $10 billion failure.
Melissa Perri: Yeah, that's a big one. Okay, so that sounds like definitely a failure thing too just being too early. What else do we have after?
Tom Eisenmann: Yeah, so the third early stage pattern is false positive and everybody will be unfortunately too familiar with that. Turns out that from COVID turns out entrepreneurs are subject to false negatives and false positives as well. False negative is heartbreaking. I mean, it's dangerous obviously in healthcare if you've got a problem and it's not diagnosed early enough you can get your body into a lot of trouble.
For an entrepreneur, the equivalent is you had a great idea something in the universe signaled you that you were off track. It wasn't a good idea but it turned out to be a false negative 18 months later you read TechCrunch and somebody's doing exactly with $50 million of backing from Sequoia is doing exactly what you thought of. False positive is can also be deadly and where it happens with an entrepreneur is it's often the case that the early adopters who you absolutely need on board as an entrepreneur, right?
There's no business until people start using the product. So you better understand them. You better pay attention to them but it's often the case that their needs are different than the needs of the mainstream users. Mainstream users if you're going to build a business of scale are really important too.
So sometimes they're the same needs and you just basically you just flow from one to the other but think of a product like Dropbox, right? The early adopters for Drew Houston the founder of Dropbox were software engineers basically with incredibly sophisticated requirements for file management, multiple devices, mobile tunneling through firewalls collaborating with a lot of people synchronizing and a lot of knowledge about how to actually do that. But Drew had a vision for the product which is he wanted in his Y Combinator application I want something that's so easy to use my mother can use it to store her recipes.
That was the product vision and so he's got his early adopters demanding all these sophisticated features and he's got a vision for the mainstream customer that's very different. He had the discipline to build something that was for the mainstream and maybe on faith maybe through good discovery work he had enough confidence that even that was so much better than what the early adopters were using that they would adopt it which was true. Not always the case.
Your book is so strong on this that what's really needed here is this product strategy and the strategy can go in three different directions. You can build for the early adopters and migrate toward the mainstream. Maybe you hide some features along the way you can build for the mainstream which is what Drew did and assume it'll be good enough and then just sort of chug along.
You could build two versions of the product and sort of a basic and pro version and there's no right answer. It's going to be situationally dependent but the key is to do the customer discovery work again to understand if there are differences in needs between early and mainstream and again it just points to the need to do all of that upfront work.
Melissa Perri: I see that one like a lot with product management because, especially when it comes out to design, I feel like people will start to build products that are meant for experts to use because they are the expert themselves or they're talking to the experts who are going to be the ones who need it, and then those things become extremely complicated and you get into the mainstream or the rest of the people who are not experts but you made it so hard to use that they can't actually pick it up and start adopting it and a lot of times they have to. If it's an existing company and they have money and they can do this and they're not running out of time and they're not strapped for cash like a startup, they can overhaul the design and do like a simple mode versus a complicated mode and hide those in advanced features. But when you're a startup and you don't have the kind of luxury of time and money and a bunch of people to put on that, I can see that being extremely dangerous.
Yeah, exactly.
Tom Eisenmann: So the failure mode there was a company called Beru. I love the name Beru is the motion a dog makes when it turns its head trying to understand what you're saying. That's called a Beru.
Melissa Perri: I had no idea.
Tom Eisenmann: Yeah. Anyway, it was a dog walking service with a pretty interesting twist. The idea was the entrepreneur was going to go to luxury apartment buildings and basically sign them up so that when you came in as a new renter and asked the concierge in the building what am I going to do with my dog? Concierge would say, well we've got exactly the solution.
So otherwise, it looked a lot like Wagga Rovers applications that some people will be familiar with, but a very different go to market approach and the early adopter, the first customer there is in Boston and South End you may remember the old headquarters of the Boston Herald newspaper. Oh, yeah. Called the ink block and newspapers fail and you turn the facility into 400 luxury apartment buildings or apartment units, all of which opened on the same day back in, whenever this was 2013, something like that, and this company actually launched in early 2015.
It was a. I hope I'm not scaring any future students away from Boston, but that was a winter where we had eight feet of snow in 30 days, and so if you were a dog walking service, you were popular, if you could actually come and help. And then the third thing that happened was this building had a whole bunch of units filled with a Hollywood film crew.
They were they'd come east to film on location. They were there for months, they brought their pets and they had per diem so they could afford the dog walker and they meet, you know, as they were working hard. So all of these things that the Baruch just got off to, this stunning start, huge demand, and they just assumed incorrectly, I mean, by the way, the other luxury buildings in the neighborhood heard about this fancy concierge dog walking service and said: please come and help us too, which they did, you know.
And then the original plan for the entrepreneur was, to her credit, not raise venture capital. She thought she would just do it with a little bit of angel money, get each market profitable before she'd moved to the next city. But then when she saw how fast the thing was growing, she threw that plan out and said: let's go get VC.
She moved to Chicago within a year because the buildings a lot of luxury buildings, apartment buildings are owned by national companies that are in many different cities. So they were saying: oh, come to Chicago because we got buildings there, we love your service. And she just got out over her skis, overextended and the big killer that she didn't really anticipate was ink block filled up all on the same day.
Typical luxury building will turn over about a quarter of its tenants per year and if you are in the building already you probably have a good dog walking service. You're not eager to let a new stranger into your home or introduce a new stranger to your pet. So they were signing people up, but essentially the addressable market was a quarter of the size that they thought it was based on the results with the early adopter.
So false positive, yeah okay.
Melissa Perri: That's definitely like a big one so this sounds too did she end up raising dc money with that to keep going
Tom Eisenmann: Or she yeah yeah she did but not enough and then they wouldn't follow on okay yeah
Melissa Perri: She actually managed to launch
Tom Eisenmann: In Washington DC and New York City before the money ran out
Melissa Perri: Wow yeah pretty impressive cool okay so that one is we've got number three and then we've got a whole different section so this is about scaling is the other one so this is so those three are about you know early days getting stuff off the ground finding your product market fit this is how you can fail like short term
Tom Eisenmann: Then there's a whole failure that comes from scaling and trying a lot of. A lot of pms will be in an organization that that will sound suspiciously similar to what i'm about to talk about. I won't go into as much depth here because you know i can go on and on about these, but the three patterns here are speed trap that's the number one killer of late stage startups and, by the way, the failure rate for late stage.
If, if the metric for failure is your investors didn't make and never will make, money didn't or never will, then about one in three late stage startups failed to deliver a positive return for for their investors, which is still shockingly high. Sometimes that's because the investors overpaid for their equity sort of tripping over each other to put money into something that looks like it's doing great, but you know, ultimately there was some kind of stumble and that's on the entrepreneur. So speed trap, and that's a strong analogy to the false start pattern.
The second one is help wanted, and i'll explain there. Helped in the form of either some senior executive position isn't properly filled and something in the company is out of control because we have the wrong person, or and it could be both of these the capital markets meltdown. I mean, we're in the middle of a really brutal capital market right now for ventures, and sometimes it hits specific sectors like clean tech or e-commerce.
Sometimes it's the whole startup economy. And then the last pattern is cascading miracles, where somebody's doing something audaciously, big and bold. Think tesla, think spacex, again ilan, think of federal express 50 years ago, which was a crazy idea to ship packages from buffalo to cleveland by sending them down to nashville and putting them on a second airplane, is the biggest venture capital startup in history.
And so sometimes these things work, but often with a cascading miracle. So many things have to happen correctly and if any one of them goes wrong, you need all the things to go right, and it's a miracle if they all go right. So big behavioral change in terms of customer behavior, lots of money, complex execution, cooperation from incumbent companies, often regulatory change, and if any one of those things is off the rails, then it's like multiplying a bunch of things together in a mathematical equation: if any one thing is zero, the whole expression goes to zero.
So those are the patterns and 4 p.m.'s out there who are living in through one or more of these. I feel for you, speed trap is the big one. It's basically cuz VCs are looking for hyper growth. It's often the case that a startup will get momentum early, often from virality, just word of mouth, referrals and so forth.
VCs pile in, pay a high price for their equity, expecting further growth. Entrepreneur doesn't need to have her arm twisted right. Entrepreneurs love to grow too. But then a couple of things happen: is you lose product market fit sometimes, or sometimes you still got it, but it's just harder to get the next wave of customers right there, intrinsic almost by definition, less interested than the early adopters, and you have to lower your price, you have to market harder and then your very growth will often attract competitors.
New new startups or incumbents who sort of see somebody new coming into their space. And that puts pressure on price, that puts pressure on marketing costs. So you get into a squeeze between LTV and CAC. The lifetime value of a customer, customer acquisition costs, LTV coming down, retention rates are lower, repurchase rates are lower and the cost of acquiring customers goes up.
And you still got some momentum so maybe you can raise more money and keep going. But eventually the VCs will look at that pattern and say not too sure. And in the meantime other bad things can happen. If you've got any kind of business that requires humans to do things answer telephones in a call center, pack boxes in a warehouse it's hard to hire those people, hard to train, and you're doing it in a startup that is growing fast so probably doesn't have the internal management system, some processes in place to basically coordinate all the activity.
Whether it be product processes, inventory management processes, budgeting processes, sort of processes for checking the effectiveness of a marketing campaign and then the last thing that can happen is you can get when the company starts to grow, you can get real cultural conflict. Old guard, new guard, that engineer three cubicles over was here early. He's got stock options that are worth five million dollars.
I'm doing exactly the same job and my options don't look like they're going to be in the money for a long time. And then cultural conflict that comes between the functions as the functions start to get really well defined and have specialists, sales complains about the quality of leads for marketing, marketing complaints that engineering is too slow to deliver the new features that marketing wants, et cetera, et cetera. So all sorts of problems.
And doesn't have to kill a startup speech rap. Sometimes you can recalibrate. Think of public companies like Groupon or Blue Apron which are selling for share prices a lot lower than their peak. But they eventually get things under control, but very often it's fatal.
Melissa Perri: Yeah, that definitely sounds like just we'll cut this out but I just want to do time check do you have a couple minutes to go over or do you?
Tom Eisenmann: Yeah, although I got a bunch of stuff to do at school so I can wrap the other ones up quickly if you want or just skip them
Melissa Perri: No, that's fine I think it would be good to explain them just so we don't go out but if you've got like five more minutes after the hour I'll pace to do them fast Okay, cool and then I just wanted to ask what would your be like my question about what would your advice be to founders who find themselves in these scaling issues and how can they mitigate it?
Tom Eisenmann: Yeah, cool, Okay Okay, so the second late stage pattern is I call it help wanted, and the help comes in two forms. It comes in the form of a missing senior key executive and it comes in the form of money from the capital markets, and in both cases that's missing. The executive gap, if it's in a mission critical function, can be really bad.
So the book the example in the book is an online furniture, home furnishings retailer and shipping this stuff across the country is really complicated. Shipping a couch you don't want it to arrive, or if your Amazon books come two days early, you're kind of delighted. If your couch comes two days early and you weren't planning to be home from work, or it comes two days late, you got a real problem and so this company actually managed to generate demand.
The help wanted pattern again: you preserve product market fit as you scale, but you just it's an echo of the bed bed fellows example with Quincy, where they couldn't get the operations under control. Here the operations were out of control. The founder didn't have an ops background decided to hire somebody that was a VP of operations who was a generalist.
He thought could graduate into the COO role which the company would need, but the person had no experience with logistics and shipping, this kind of stuff. And, the worst thing is, picked the wrong ERP system, enterprise resource planning system, the system that any company would use to manage inventory and orders, and so not only were the operations out of control, but they didn't actually understand what they had an inventory or what the status of orders was, and so it really fired that person. Hired somebody who had done operations, but this individual had done it at Netflix and sort of if we remember that the DVDs used to come in these red envelopes sort of shipping hundreds of thousands or millions of those is very different than shipping lamps and couches, and this was a big.
By the time this person had been in Netflix. It was a big, big company and the company the individual brought in big company processes and attitudes and the company was the startup sort of rejected it like antibodies around a virus. That's the really big thing.
Melissa Perri: I see too with startups. I had the same situation where a CTO was brought in from like a really large company to a startup I was working at once and the engineers rejected him immediately because he was like, of course he wanted to use Jira. They didn't want to use Jira, but he also wanted to. They're like using proprietary languages.
He's like: we're going to do everything in C-sharp, we want to do this stuff, and I feel like that's like the wrong somebody who's got like not the right cultural fit, or just from a company that's been so big. You know, going back to a startup is very different and I don't think people really appreciate that. Yeah, in this case,
Tom Eisenmann: The. The individual sort of the version here was was having somebody on the team run Excel models for him as opposed to sort of firing it up. You know, god forbid he would have to do a SQL query so that eventually that person got a lot done but not enough, and they nailed it on the third try but and eventually got their operating margins under control. But just as they did and went out to raise a series C, the company found that the capital markets for e-commerce had completely shut down. Even good, healthy businesses were having trouble raising money, so ran out of money and and went bankrupt. Had to shut the thing down.
Melissa Perri: Wow heartbreaking yeah
Tom Eisenmann: This is also the last pattern. This is: this is rare, because it's rare that an entrepreneur does something so audacious. You know, when these fail, they leave giant steaming craters in the landscape. You, you hear about it. We think Theranos, think Webvan, think people remember Iridium Segway. These kinds of businesses where many, many things have to go right and the chance that they all will, they're often pushed by what I would describe as a monomaniacal founder.
You know somebody who is laser focused on on their vision and often really good at selling the vision. So I mean, the only way you're going to raise a billion or 10 billion dollars is if you are just dazzlingly persuasive and you know there can be a line between visionary and cult leader and and sometimes you're on the wrong side of that line. So not a lot that I can think of.
To correct that. It sort of takes a good board of directors and it's not always the case that somebody that fits this monomaniacal profile is eager to surround themselves with a strong board.
Melissa Perri: Yeah that's a good one too so with the scaling one I think that particularly a lot of these issues I think affect product managers one of the ones that we talk about in scaling too is when do you hire your first product manager what what do you suggest and what have you seen teams do with that yeah that's some
Tom Eisenmann: I've thought a lot about that and I've done some research with a with a product leader out there turned a venture capitalist. Deep Nishar was a senior product person at Google and then ran all of product for LinkedIn before he became a venture capitalist. And Deep and I did some work a couple of years ago where we interviewed a lot of product leaders and this is one of the questions we asked them and there were.
There are two things you really are have to answer. Here was one is the timing. If you're an entrepreneur, when do you bring in the first PM? And then who? What is, what's the skill profile? And those turn out to be dependent too soon with timing and you can actually waste money. Right, a PM is going to be expensive and if you don't need them because the founder is essentially doing the PM job, the wrong PM can slow things down, can bring in process prematurely into an early stage startup.
The wrong PM can elicit a strong reaction from the engineers who at this stage love working with the founder. Right there's that camaraderie and it's very loose and flexible. And now we got somebody who's sort of going to insist on tracking things and you know and process and so forth. And it's often the case that a founder will.
Maybe they're told by the venture capitalist they need a PM, but they really don't want one. They're not ready to share product leadership with somebody. So those are things to watch out for, of course, too late the big problem is the founder becomes the bottleneck. You know the founder, in addition to shaping the product and sharing that vision with the engineering team, has to do all the things the founder CEO has to do: raise money, figure out what to do about, go to market etc.
Etc so you know you want to hit the Goldilocks spot between too fast and too slow. And then what Deep and I, um um decided after talking to a lot of product leaders, is there's three scenarios for when you make that hire. Um a good one is post product market fit, when you've started to scale, and uh, then um.
Then the question of who. We use the term a full stack PM, and by that we meant somebody who can both execute a vision but also be a partner and with a founder in developing the, the product strategy, the product vision. And if you bring that person in um, if the, if the founder has product experience or good instincts for it has time, maybe because there's a co-founder that's doing some of the other stuff: the money, the fundraising or the go to market.
If um, if um, you get to product market fit early. This is a time when the scenario works and sometimes you have a vice president of engineering that actually can do a lot of the things a product manager would do. So under those circumstances, um you you um you can bring in the PM after product market fit.
You don't want to bring them in prematurely because, as you're searching right the, the PM should be tuned to the nature of the business and if you're zigzagging around pivoting, you may end up hiring the wrong person prematurely because you, you don't really have somebody that fits the and the beauty of doing it post market fit. We talked about how the there's a difference between what the early adopters want and what the mainstream may want. It's often the case that even after product market fit, you're doing a lot of changes and now the founder's way too busy to sort of think, to do a new round of customer listening, sort of thinking through the feature set.
Now that has to evolve. So full stack post product market fit scenario. One failure modes there is: you wait too long to do it, you um, you get a backlash again from engineering or again the founder's not willing to hand off. Second mode is you do it really early and you bring in the executor pm, not a full stack pm, but somebody who can take the founder's vision and turn it into product right, prioritize the work, track results and so forth.
And then the third pattern is a full stack pm very early, and you you might want that under two conditions. The first is you got a founder who really doesn't have a vision and unfortunately they're out there and they need a partner to sort of say this is what we should be building. The second is a founder with a strong vision, but they're the kind of creative person you think of.
How many creative duos are lennon and mccartney right? Or in business jobs? And and was neac? Some of us just do better with a sounding board, a partner, and that that full stack individual can can help the right founder or a founder who lacks product vision. So yeah, timing not too soon, not too late.
And then which of the scenarios do you fall into?
Melissa Perri: Yeah, one of the things that you mentioned too i just wanted to like poke in on, is talking about the ceos who are not ready to give up tinkering with the product. And i've seen this pattern as well, where you have and i have a lot of empathy for founders like i don't want this to come off weird, but like i have seen, ceos are really good at building and getting the things off the ground. Then their company starts scaling and they move from being the product person, the person with the vision, into now the ceo role of running the company.
And then you also see this with people who are like the first engineer and then they get promoted to the cto because they've been there for a long time and now the company is scaling. You've got a couple hundred people. You're starting to bring in other executives that are like have done this before.
We've got like you know many, many years of experience and stuff. And that's where i see, you know the ceos one get frustrated because they're like: i just want to tinker with the product and it's like you can't at this point you got to go do other things right. And then you also see the cto's sometimes out of their depth because they've never had a skill or architect the platform like the ones that they're building.
What types of things have you seen with those, you know, with founders kind of staying in those positions, and how it leads to failure and what can people do to like mitigate that?
Tom Eisenmann: Yeah, and i think you've you've really nailed. I mean, it could be its own failure pattern here which is hanging on to the reins too long and not having so. You know, at some stage, if you take, if you're scaling and you've taken a lot of venture capital, in most instances you bring in a new outside, a new member of the board of directors with each major vc round, and at some point the investors are going to outnumber the, the management on the board, and the most important job of any board, especially a startup board, is to hire and fire the ceo.
And so it turns out that only 40 percent of founder ceos are still ceo at series d. Which would be, i don't know, maybe six, seven years into the life of a typical startup, and so it's quite common to replace. But you know, we've got a, and sometimes it needs to happen earlier when the when the ceo still either has control of the board and can't be fired by the investors or some investors are just especially.
We went through a phase it's not as true for the past year, but for a long time bcs were founder friendly and really you know, we're going to make the founder successful and a founder friendly vc is much slower to replace a founder ceo so you can leave the wrong person in place for a long time. It's really, you know, hopefully a good, a good founder has self-awareness. I'm not good at this new stuff.
I don't like newing, doing the new stuff and the new stuff is really important to the success of the enterprise. So it's time and and there's probably some role inside the organization that i can keep doing the things i'm good at and like doing. But but i gotta give up the ceo job and that's just brutally difficult for any founder, right?
Yeah, it's like giving up your baby.
Melissa Perri: Exactly
Tom Eisenmann: And that's hard
Melissa Perri: And i, i do think too, like on the other side, have you seen, we've got the mark, mark suckerbergs out there, we've got the people who have done it. They were the founders and became ceos. What do you think makes them successful and allows them to still hold on? If somebody was like i really just want to see my baby grow and i do want to do this spot, i want to be in it for the long haul, what do they have to do so? I mean, there are some amazing
Tom Eisenmann: Human beings out there and you know it's it's easy to focus on a steve jobs or a zuckerberg, and not all of us are are going to have that dna. So one thing you can do is just be born lucky and be as smart and flexible and fast at learning as as some of these amazing ceos. You can have a really strong number two and respect that person and and and give them so the sandberg cheryl sandberg to zuckerberg is an example and yeah, i think, i think i think that's the only obvious solution is sort of recognize that you need a peer who respects your vision and makes space for you to do.
You know, zuckerberg would regularly step in and and and reshape the product and sandberg was smart enough to make space for that to happen and sort of, you know, cause ripples in the organization, figure out how to how to tamp things down.
Melissa Perri: Yeah that definitely sounds like a a nice approach for it i wish founders saw it too as that it's not failure if you do have to step down and bring somebody else and it's still success because it's still your baby at the end of the day you still have equity you still want to grow it
Tom Eisenmann: So yeah exactly and and for a lot of founders it's a relief because what they really are good at and love doing is creating something new so some of them will stay in the organization in some different role and many of them will go on to do the next thing yeah just good
Melissa Perri: Well thank you so much tom for being with us today where can people go and find your book why startups fail
Tom Eisenmann: Uh, you can find it at any online books, thank you, any online bookstore, amazon or you know. If you want to support your brick and mortar retailer, that's a good thing. And they can find me at on twitter. My last name, eisenman e-i-s-e-n-m-a-n-n with a t in front of it, tisenman at twitter, and. But i'm spending more time on linkedin these days, so as we all are, it isn't twitter anymore. Right, it's x now, so but i'm still there.
Melissa Perri: Great well i hope you all enjoyed this episode if you want any of those links they will be linked in the show notes after this so definitely head to our show notes at productthinkingpodcast.com to get all of those links and find tom's book we will be back next wednesday with another episode of the product thinking podcast make sure that you subscribe so that you never miss one
