Why Some Start-Ups Fail to Scale

CURT NICKISCH: Welcome to the HBR IdeaCast from Harvard Business Review. I’m Curt Nickisch.

As a startup founder, it’s got to feel exhilarating to see new customers streaming in and paying for your product or service. To get to this point, you’ve gone from conceiving your idea, building a small team and getting those early funders to help you test it in the world. Considering how many startups fail, watching customers put money down can make you feel like you’ve slayed a dragon. But watch out. There is another dragon waiting around the corner. Even fast-growing startups that get glowing reviews from customers and the media often end up flaming out. Despite all that positive momentum and growth, they’re just not able to stay profitable at scale.

This scale-up phase of entrepreneurial ventures is a huge challenge, and today’s guest has researched the stumbling blocks to long-lasting success. He says you can overcome them by expanding your business model while at the same time systematically removing internal constraints on growth.

Jeffrey Rayport is a senior lecturer at Harvard Business School, and he’s a co-author with Professor Davide Sola and Martin Kupp at ESCP Business School of the HBR article, “The Overlooked Key to a Successful Scale-up.” Jeffrey, thanks for joining.

JEFFREY RAYPORT: Curt, thank you so much for having me.

CURT NICKISCH: So what is the scale-up phase and how do startups know when they’re in it?

JEFFREY RAYPORT: So that is a great question, and it’s really where we began. The startup world orients very much around these ideas of zero to one, getting from the proverbial two guys in a garage or the group of folks in front of a whiteboard to something that has what in lean methodology terms is called product market fit.

And much of the startup world, after 25 years of internet entrepreneurship, focuses very much on that incredibly hard problem of how you stand something up. We think of the scaling phase, we meaning Davide and Martin and I, I’m so glad you mentioned my colleagues in Europe, we think of the scaling phase as beginning with the confirmation of product market fit.

And that means you’ve got an opportunity. It does not mean that you have much in the way of revenues, and most folks at that point have negative profitability. So for us, the scaling stage begins with confirmed product market fit and moves up a very steep growth curve to the point where a venture can declare that it not only has product market fit, but also profit market fit. Often business models are borne out, not just on a unit economics, but become profitable by dent of scale. And so those two critical things need to happen in the scaling stage, and those we would argue are where real value creation occurs.

CURT NICKISCH: I love that term profit market fit. What does that look like?

JEFFREY RAYPORT: So for us, the key initial insight was that what it looks like is not the middle of a curve. And what I mean by that is that the received wisdom from the academic literature, going all the way back to an article by James G. March in the early 1990s, was that there are effectively two stages in the development and evolution of any business enterprise.

And stage one is famously the period of exploration, roughly coincident with this idea of searching for some kind of market demand or market traction. And then the second stage was that ventures and corporations move into the stage called exploitation. And it’s a very, very elegant model, and intuitively it makes sense. You look for opportunity. When you have find opportunity, you figure out how to exploit it for all of its potential economic value.

For us, the problem with that is that we have been seeing, as I think anyone in the business world these days who have got an eye on the startup space, but especially the tech space, a lot of companies where they can be very successful on the road to product market fit, but the wheels come off the bus in some way during the scaling phase. The academic world hasn’t commented a lot on the question of what it takes to get from that zero-to-one phase to the stage of scale. We have deemed that middle stage, not exploration, not exploitation, but the extrapolation phase.

CURT NICKISCH: Where do the wheels come off the bus in this phase?

JEFFREY RAYPORT: It’s interesting. As my colleague Tom Eisenmann has written about in his book Why Startups Fail, there are clearly many, many modes, and Tom has created a typology of them, of why startups fail. He’s looking largely at earlier stage businesses.

CURT NICKISCH: That’s not a lot of solace to people who’ve been there. It’s like, now I know how to name this failure.

JEFFREY RAYPORT: Exactly, exactly. That’s right. That’s some form of consolation as you go through the bankruptcy process, and I shouldn’t joke about that because we’re seeing a lot of that happening right now. We’re seeing a lot of significant meltdowns, most recently the FTX implosion, very much in the scaling phase of a business. And there are many ventures that by dent of their success and moving up that steep portion of the curve, cannot keep up with the demand that they’ve generated. So some of this is wheels coming off the bus because they actually can’t source enough supply to keep up with demand.

Friendster would be a good example of that, the earliest of the major social platforms. The demise of Friendster had not a lot to do with the question of whether they had product market fit. It had everything to do with whether they could actually serve the tens of millions of simultaneous users who were coming onto the platform. So one is the issue of being killed by your own success.

Another, of course, is that for businesses that have not yet achieved profit market fit, the issue of securing capital to finance your way to the point at which the business becomes cashflow positive is very significant. And if you have not planned carefully as to how much capital you need to get there or what milestones you need to hit, that’s another way to see a venture run out of fuel before it actually gets to the point of sustainability.

We have seen ventures fall apart on a human level, meaning on the level of organizations not having sufficiently coherent culture, in order to assure that as they go from 50 to 500 to 5,000 people, that people are on mission and that the efforts are aligned. And then I suppose there are ventures, and it seems crazy to say this, but there are ventures and entrepreneurs who attempt to scale into a market that’s not big enough to justify the scale that is part of their vision.

If it turns out that you can win 90% of the share in the market and you’re still a small venture, then actually you’ve got a ceiling on growth and scalability, which means you have nowhere to go. We see that happen quite a lot. We see organizations who think they have a sound go-to-market strategy in order to access the consumers they need to serve, again, to drive scale, who find out that they don’t have either practical means or economically feasible means to reach the customers who are their target market. So it is all over the map. There are lots of ways to fail.

CURT NICKISCH: So you’ve given another term to help explain the scale-up phase, calling it extrapolation. Can you just explain what extrapolation is?

JEFFREY RAYPORT: Well, maybe the best way to bring it to life is to talk about one of the companies that we’ve spent a lot of time with, and that’s King Digital Entertainment. It’s a London-based game maker. Many people will know it. It’s been recently acquired by Activision Blizzard. King, people who will not know the corporate name will know Candy Crush Saga, Candy Crush Soda Saga, the mobile casual games that they produce. When they introduced Candy Crush, it became a hit unlike anything they’d ever seen, and their revenues grew 12-fold. Now that meant that in order to keep up with that pace, they had to significantly expand headcount. They had to significantly expand infrastructure.

So extrapolation is often an order of magnitude increase in top line revenues in a relatively short period of time. The companies that we looked at, several dozen companies we’ve written cases about collectively, tend to do this in a relatively compressed period of one to three years. And so for anyone who has managed steady growth at 10 to 20% a year, think about the idea that over a period of one, two or three years, your revenues go up 10X or 20X or 30X. It starts to describe the unique challenges, not just of growth, but of exponential as opposed to linear growth.

CURT NICKISCH: Your operating costs are going up exponentially too, perhaps.

JEFFREY RAYPORT: Absolutely. And even before the crises of today, think about a company like Uber, which in theory looks very much like the kind of platform dynamics with increasing returns that we just talked about. But the reality is that Uber, by dent of its relentless pursuit of growth, competition in the marketplace from Lyft and others and so forth, managed to move up that growth curve, get all the way to an IPO, establish a public market valuation before the tech meltdown of $100 billion market cap, and they still had not found profit market fit. They had not made the business model work.

That’s changed in the last couple of years as they’ve pursued profitability, as many tech ventures are these days with the change in market sentiment. But what you say is absolutely right, which is the fact that you have a platform does not guarantee that you are achieving profit market fit. That has to be part of what happens during the extrapolation phase, not by accident, but by design.

CURT NICKISCH: What do you need as a company then to begin extrapolating?

JEFFREY RAYPORT: The sufficient conditions, meaning that you’ve got some foundational attributes of your opportunity, but now you actually want to see it come to life. One is understanding, back to what we talked about just a moment ago, that you’ve got an effective way to get to the target customers, that large proportion of the large market that you need to reach to be successful, that you actually have a so-called go-to-market strategy that you believe in that can be successful, that gets you the access. And by the way, in economic terms and unit economic terms, you can support.

It’s also true that you’ve got to have some view, if you don’t have it on day one, to what your approach to monetization will be. And Curt, I know you know the audio industry well, and we both as consumers, I’m sure, have spent many, many delightful hours listening to two of the major streaming services on the planet over the last 10 years, one being Spotify and the other being SoundCloud.

Several years ago, Davide Sola and I had the pleasure of spending time with the leadership team at SoundCloud in Berlin and then in New York trying to understand their business. This was at a time when the consumer listenership of SoundCloud was around 200 million monthly active users, and that made it, at the time, on a consumer basis, larger than today’s industry leader, Spotify.

So what was the difference in the outcomes of those companies? SoundCloud at that time was effectively giving away music free of charge to consumers and focused on a very small and vibrant community of musicians who wanted to use SoundCloud as a hosting platform. So several hundred thousand musicians paid SoundCloud monthly or annual fees to host and stream their music on the platform. It’s very clear, if you’re thinking about market size and scalability, that a market at that time of 300,000 musicians who are monthly active users is a lot smaller than 200 million monthly active users who are consumers or listeners.

Spotify focused first on the hundreds of millions of consumers, and SoundCloud got there late. And even though they managed to do a bunch of the deals they ultimately needed to do with the record labels to clear the copyright protections on that audio content, it was too late, in effect, to save the business from a dramatic recapitalization down round and effect turnaround that has been true of its story to this day.

So understanding how it is you’re actually going to monetize what you’re doing is enormously important, and then there are other attributes that we’ve talked about. There are increasing returns, dynamics to that economic model, understanding how the business will make the most of and leverage network or density effects, things that people refer to as virality or the viral coefficient in the market. And then, of course, none of this comes without capital investment, and that means that if it’s a big, bold, ambitious strategy, it’s going to take capital across multiple rounds of venture funding to get there. And there’s got to be a strategy and a view as to how that capital will come into the enterprise, which means knowing what it’s going to take in terms of milestones to score those additional rounds of capital until you move up the curve and hit that point of profit market fit, at which point the enterprise or the venture, in theory, moves towards sustainability.

CURT NICKISCH: Well, let’s talk about this process then of extrapolation, and what are some of the successful ways you’ve seen startups navigate this phase?

JEFFREY RAYPORT: One we talk about is a business started by a couple of friends of mine, our next door neighbors here in Back Bay in Boston, Niraj Shah and Steve Conine, who are the founders of Wayfair. Anyone who is in the business of refurnishing or furnishing their homes knows that there are only a couple of places to go online to buy furniture and furnishings, and Wayfair is lead among them with these days 14, 15 million SKUs with essentially a platform that is comprehensive to the home category.

Wayfair had a very interesting start back in the early days of Google, days when Yahoo was a big search portal and so forth, they noticed that there were these category sites, people who sold very specific products online, and the very first business that Wayfair established was called racksandstands.com. It was a site, I would call it a product.com site that sold only that category of product, or products plural. They went on to other amusingly named sites like allgrandfatherclocks.com, and they went category by category, niche by niche until the business was nearly a decade old and they had about 250 such sites that sold product very successfully online at competitive prices.

Growth was directly correlated to their ability to stand up additional categories. And by the time you got 250 categories, you could argue that would be a diminishing return strategy. And what clearly would allow you to grow the business was to find a satisfied customer at racksandstands.com and convince her that the next time she needs a grandfather clock, she should go to allgrandfatherclocks.com. But that kind of repeat purchase did not happen because there was no way to know from any one site that they were part of a larger store.

So the point at which they raised real money was at the time when they recognized that the only way to get that kind of repeat purchase and cross-category sale was to put all of these sites on a common platform under a common brand, and that is the point at which they spent two years migrating what was then eight or 10 million SKUs from the 250 product.com sites over to the common platform ultimately branded as Wayfair and spent a good deal of money in marketing and media in order to build that brand.

So this is interesting. So this eliminated one significant constraint to growth. They had another constraint which was next on the list, which is that people tend to buy from eCommerce platforms where they have incredibly delightful and satisfying experiences, as defined by an Amazonian gold standard. You get the product quickly, it’s beautifully packaged, the box is clean, what you ordered is actually what’s in the box and so forth. One of the barriers for Wayfair with its dropship model, and even today at your top line of 15 billion and up, the company still is selling roughly 85% of what it offers on its site via dropship, meaning it ships directly from the manufacturer.

Two problems with that at the time. One was that manufacturers are not in the business of serving or fulfilling one-to-one orders. They ship on pallet loads to warehouses that go out to retailers who break down the lots. So one issue was these folks were not terribly skilled at one-to-one order fulfillment. They didn’t do it quickly, and they didn’t actually have much skill, capability or background in how to package up the product, let alone to put a Wayfair brand on the box.

The guys, Niraj and Steve, recognizing that this was their next constraint, then built a logistics network, something that they call CastleGate, in which they went to their suppliers, their manufacturers, and offered to do two things. One was to educate them in state-of-the-art packaging of product and shipping logistics, so that they could fulfill orders directly dropship in a more consumer or end-user friendly way.

But even more importantly, and this is where CastleGate came in, was to forward position, meaning to move the best selling products on offer into CastleGate owned by Wayfair, so Wayfair distribution points across the country, so that those products could be shipped with lightning speed directly to consumers. And hence, you now have a rising level of satisfaction or net promoter score, NPS, among your users.

CURT NICKISCH: And returns go down. Right.

JEFFREY RAYPORT: That’s right. Returns go down, satisfaction goes up, loyalty increases, people come back and make three more purchases that year instead of two, so all of a sudden lifetime value begins to grow. The third constraint was that they’re an incredibly commoditized category, as you know from walking into any furniture store. In general, this is one of the last big categories of consumer durables where pretty much everything you look at outside of a Knoll or a Herman Miller showroom is largely unbranded. And because furniture is unbranded, especially at the manufacturer level, they don’t have a lot of pricing power.

So one of the things that Wayfair did to increase pricing power for their suppliers and for Wayfair as a retail platform or a marketplace, was to establish a huge number of private label or house-branded lines, essentially to take, whether it’s sofas or it’s mattresses or wall coverings, Wayfair created a bunch of house brands which allowed them to market these as branded products, establish higher price points and hence more gross margin for them as the retail platform to pocket, as well as to share with the suppliers. So one of the things that we see as a success factor is applying this ruthless and disciplined process to the ways in which you take off the limiters on how scalable the business could be and how large it could become.

CURT NICKISCH: What about your team and your people and your culture? What do you have to do there to make sure that you’re able to reach product market fit from a functioning organizational perspective?

JEFFREY RAYPORT: I am so glad you asked that because that human element is as important, if not more so, than all of the strategy go-to-market and operating dynamics we’ve just been talking about, the support positive economics. What we have seen, of all the ways in which successful scaling CEOs and founders that we’ve studied have delivered the dream here that we’re talking about, is that their version of design for scalability is to put disproportionately heavy emphasis on cultural issues early on. I’ve always thought, in the businesses I encounter in my own experience in the business world, that there are two fundamental ways of thinking about corporate culture, and one is that culture is like the weather. We have no control over it. It’s just something that happens to you. And five years out, if you wind up in Dilbert land with a bunch of cubicles and depressed employees, my God, what went wrong? But it just happened.

And much of large-scale enterprise defaults to that kind of cultural environment, not because they want it, but because it just happens. One of the beauties of being in the startup space, of course, and one of the motivating factors for entrepreneurs is you get to invent the world anew. You get to dream a dream and then live inside it, and it’s your dream. The folks who manage to do that at scale, meaning to get to scale, do it by making some very clear decisions about what kind of culture they want at the very start and driving toward it with mindful investment over time.

There are many, many positive examples of this. One company we’ve spent a good deal of time is the circular commerce company that is a platform selling pre-owned apparel called Thredup. And James Reinhart and a graduate of the school started that venture with a very clear eye to a culture of curiosity, of ambition, of performance, to the point where James would periodically stand up in front of the workforce and essentially let folks know, in a way that I suppose might sound very Elon-like in terms of what’s happening at Twitter right now, that with a human face, he would say, “Look, here is what we’re all about. And if this is not what you are all about, we’ve got an incredibly generous severance plan that will aid you in departing from the enterprise.”

It’s something that Reed Hastings at Netflix is famous for. Many people know what their severance is on the day that they take their job. And as a result, essentially the culture is saying that this is not a family, it’s a team, that people are on a team because they have a useful role to play. When the role is no longer useful, then there needs to be a graceful, respectful way to get them into a different role or take them out of the team. But these ideas, again, whether it’s warm and fuzzy or it’s hard driving, whatever it is, being clear and explicit about it correlates with success in the companies that we’ve studied.

CURT NICKISCH: Jeffrey, you’ve talked a lot about the work that startups have to do on their culture, on understanding their operations and the size of the market and making really strategic choices about getting to the profit market fit that you’re talking about. I’m curious if you think a lot of startups fail in this extrapolation phase, in this scale-up phase because they know their voice is telling them to do that and they just don’t have time to do it because of this compressed timeframe that you’re talking about. Or do they just not know and that’s the reason that they fail at making some of these crucial choices?

JEFFREY RAYPORT: We have a very healthy respect for something that great Austrian military strategist, that is von Clausewitz, used to call the fog of war. I don’t think you can underestimate how challenging it is to do any of what we’re talking about, whether it’s in the launching phase, again in searching for product market fit, or it’s now in this incredibly intense phase moving up the curve towards profit market fit and scale.

So Curt, I would say that we’re talking about immensely talented people who are fully aware, clearly or largely aware, one would hope, of the risks involved. But I think the notion that it is hard to see clearly while you’re in the midst of it is very profound. So that’s the fog of war argument.

I think the other part of this takes us all the way back, if you will, to the beginning of the conversation and where this article came from. One of the things that we found very interesting, when we’ve taken the ideas in this article back to many of the people we wrote cases about and in several cases we interviewed and quoted in the article and we said, “Here’s our conceptual understanding of what you did to be successful,” without exception, we got very interesting looks, raised eyebrows, a sense of surprise and delight. Nobody was arguing with us about the fact that this was a robust and high fidelity way to describe what they do, but there’s not a single one of them who said, “Oh my God, of course, I had a roadmap. I knew exactly what I was doing.”

I think it illustrates the fact that without actually circumscribing this phase of growth, meaning that without pretending that we move from exploration to exploitation in a nanosecond, but instead saying, wait a second, there’s this middle phase, and it requires different ways of leading and managing and structuring operations, thinking about economics, thinking about culture, planning for the future, raising capital.

There are a bunch of skill sets and capabilities that are unique to success in this phase as well as approaches that we’ve talked about, none of which we as the business world have been paying attention to or will pay attention to in a rigorous way without saying that this phase in the development of any venture is fundamentally different from what comes before and what comes after.

So I think there is a fog of war version of this, but there’s also the fact that we in the world of practice and the world of academia have a lot to do, we believe, in further understanding this space and further figuring out ways to maximize upside and success while minimizing the very substantial risks that come with the territory.

CURT NICKISCH: Well, Jeffrey, you and your colleagues’ research hopefully will clear some of that fog and give people in that situation some better information for moving ahead. Thanks so much for coming on the show to talk about it.

JEFFREY RAYPORT: Curt, thank you so much.

CURT NICKISCH: That’s Jeffrey Rayport, senior lecturer at Harvard Business School and the co-author of the HBR article, The Overlooked Key to a Successful Scale-Up.

And we have more episodes and more podcasts to help you manage your team, manage organizations, and manage your career. Find them at hbr.org/podcasts or search HBR in Apple Podcasts, Spotify, or wherever you listen.

This episode was produced by Mary Dooe. We get technical help from Rob Eckhardt. Our audio product manager is Ian Fox. And Hannah Bates is our audio production assistant. Thanks for listening to the HBR IdeaCast. We’ll be back with a new episode on Tuesday. I’m Curt Nickisch.