Joe Haslam on transitioning from startup to scale-up

To learn more about when and how to grow, I hosted Joe Haslam on our podcast. Joe’s the Executive Director of the Owner’s Management Program at IE Business School in Madrid, which is ranked by the Financial Times as the top school in Europe for entrepreneurship. His academic focus is on how to scale companies, something he’s done quite a bit of himself, most recently as Chairman and co-founder of Hot Hotels.

Our chat covers the pitfalls that lead to premature scaling, paths to growth outside the Silicon Valley norm, the pros and cons of getting big, and more. If you enjoy the conversation check out more episodes of our podcast. You can subscribe on iTunes or grab the RSS feed in your player of choice. Short on time? Below are four key takeaways:

1. One of the biggest errors a startup can make is premature scaling, which means adding sales and marketing before product/market fit is fully achieved. The result is an over-promise and under-deliver scenario that’s hard to recover from.
2. There are three typical models for scaling. Blitzscaling, the concept of chasing VC and being totally focused on growth, is most common in Silicon Valley, but much of the rest of the world takes a different approach.
3. An added emphasis on operations is a key part of scaling. Whereas startups excel in generating demand, operations will help ensure you’ve got the supply to match it.
4. Growing your company comes with the benefit of a wider talent pool from which to hire. Often great people won’t join a company until some of the chaos is tamed and structure is in place.

John Collins: Joe, welcome to the show. How did you go from running startups to looking at the issue of scaling companies?

Joe Haslam: I was one of those consultants who left a good job during the dot-com era to go set up companies. Myself and five others started a company called Marrakech, which went on to raise $75 million. I was looking at where all the money was going, where we were spending it, and we really didn’t have a clue. We can be honest about that now, although the company still exists today. It got me thinking, there must be a way to do this that is less random and less haphazard. As my journey has gone on I’ve founded six companies in total, and in some ways I see some of the behaviors repeated.

One of the things that has changed in startups, with Steve Blank and Alex Osterwalder and Eric Ries, is that we know a lot more about how to start companies. You don’t just lock yourself away in a room and then come up with a product and think people will buy it. We’ve gotten a lot better at that. However, in the world of scale-up, which is what comes after startup, we’re still kind of blind.

There have been some theories – Bob Sutton at Stanford University is doing a lot of work in the area – but in general, there isn’t a lot known about scaleups. We’re finding there’s an assumption that once you raise money, once you start to get bigger, that the time of survival is over, but actually it’s even worse. It’s even harder.

A Scaleup is not a Startup

from

Joseph Haslam

When is your product ready to scale?

John: If you were to read only Medium or Hacker News, you’d assume that once you reach product/market fit, you’re home and hosed. You pour petrol on the flames and away you go. But as you found, you have to put a lot of things in place before you can scale. Is it just a case of get to product/market fit and then get some deep pocket investors? What are the mistakes people make in the transition from startup to growth phase?

Joe: You’ll hear, “Get to product market fit, then add sales and marketing”. Maybe that’s worked for some people in the short term, but might not necessarily work in the medium term.

Alex Schultz of Facebook says that’s the biggest mistake people make. They want to have product market fit so desperately that they pretend to have it. I’ve drawn the analogy to getting married. When you’re getting married you desperately want this to be the love of your life and sometimes you overlook (smaller problems). What you should do before you get married is put someone in front of a slow computer and see how they behave. Anyone who can’t survive that environment is probably not a good person to marry.

Nothing kills a bad product
like good marketing.

The thing about product/market fit is that people want desperately to believe it’s true, so they will look at the evidence that it’s true but ignore the other stuff. The big error then is this thing called “premature scaling”, which is when people believe they have product/market fit when they don’t, and they add sales and marketing. Nothing kills a bad product like good marketing. In other words, if you have a terrible product and you set people up for huge expectations and then you don’t deliver, they generally don’t give you a second chance. We’re seeing that a lot. So don’t pour on the sales and marketing, until you’re sure you have product/market fit.

John: How do you know if your product is ready? What are the sort of things you’re going to see in your product?

Joe: The classic thing is engagement. You need to pass the toothbrush test, which is what they call it in Google. It’s the idea of something you use every day. Of course you want an early enthusiastic amount of users, rather than lot of users. In other words, a small number of people doing it every day is much better than a large number. What gets really interesting is the scale-up situation, where you have to double down on that, and also the tension that comes with your early enthusiastic users, who tend to be people that like the personal service and things like that.

When you start to scale, it’s not possible to deliver that personal service. Ebay is an example. Their early dedicated user group were almost in tears as the company said okay, there’s more money to be found elsewhere, so that’s where we’re going. That’s a real tension and something you have to manage. You need your early users to test everything from your technical scalability, to how many users you can support at the same time, to things like usability, because as you grow bigger and as you get away from the early adopters, people are less and less tolerant of things that don’t work. That’s one of the things about scale-up that really differentiates it from startup – how you treat your users and what kind of things they’re looking for.

The three models for growth

John: But not all scaling is the same, there are different ways to approach this. You’ve talked about a few different models of how you quickly grow your company. What are the main ones? It seems the US has a particular model; maybe in China they’ve approached it differently.

Joe: Terminology alert here. The first one would be Blitzscaling. There’s a magnificent course available on YouTube that Reid Hoffman has put on Blitzscaling. It’s the kind of thing you have to sit down and literally watch it first to say “wow, I need to learn this”, then watch it a second time with a pen, and watch it a third time as a podcast. The thing about it, of course, is that Silicon Valley isn’t everywhere. At IE Business School, we’re not a Silicon Valley influenced place in the way that Stanford is. Most of our people come from family business. Maybe they had a family business that was working very well in Italy but now have competition everywhere and they’re realizing they have to get to a certain scale to be competitive, so they send the third generation to us and we talk about that.

John: So Blitzscaling is taking in venture capital and being totally focused on growth?

Joe: Absolutely, totally focused on growth. So you have Blitzscaling, which is Silicon Valley. China then has something called Hyperscaling, which is absolutely fascinating. China’s just so big.

John: In population and addressable market.

Xiaomi: an example of a Chinese hyperscaling company

Joe: Yeah, so in other words, you can have an app that says, “We’ve got to close it down, it only has 15 million users.” The way the Chinese people are organized, a lot of people are living away from their villages and they can get hooked on something really, really quickly. Xiaomi would be an example of a Hyperscaling company. What they’ve been able to do week in and week out for three or four years continues to astound me. Amazon is often talked about being a very difficult environment, but Xiaomi is much more intense.

So that’s the Hyperscaling environment. Then you could talk about the rest of the world, which (moves) slowly and profitably rather than a dash for growth. Profit is still attached to a lot of the rest of the world. There wouldn’t be this kind of load up on VC, as long as we have the engagement, go public as quickly as you can and then incentivize your employees with share options approach, which is the Blitzscaling model.

There are enormous companies in Germany and Italy that you probably never heard of, who are still private. They have fantastic growth, but we haven’t heard of them. One of the reasons that Silicon Valley gets the attention is because they provide the information.

John: Are there any models from the natural world that are useful? A lot of people talk about hypergrowth as if it’s something unnatural, but where does scale exist in the natural world?

As you scale, you have to keep the product/market fit, always.

Joe: One of the things we’re poor at in the tech industry is the successful people tend to have an unbelievable ability to focus, and it means we know very little about the rest of the world. There’s a story about Mark Zuckerberg when he came to Madrid, all he could talk about was Facebook. He was even asked about Barack Obama and all that he could say about Obama was that Obama was on Facebook. This focus is a lot of what makes these companies super successful, but it doesn’t mean that you read recreational physics or stuff like that.

Patrick Collison at Stripe is an exception. One of the reasons he’s liked by a lot of people is he seems to be a more rounded figure. But the reality of scaling is that you should be reading books in biology about how a forest fire both starts and is put out, because what you’re trying to do is sustain product/market fit. As you scale, you have to keep the product/market fit, always. If it doesn’t, you don’t make sales. The product/market fit changes.

Generally that definition gets expanded. You might have product/market fit that’s for this group of users in these circumstances at this time. They are the optimum for your product. What will change is, for example, perhaps you’ll move from women between 30 and 50. You’ll extend that. So it changes, not a lot, but it does change. You’ll also go narrower. You’ll dig even deeper. You’ll double down, triple down, and that’s how you scale.

The best example when talking about scale is Ryanair. I was talking to the Deputy CEO and I asked, “What does the Deputy CEO do?” I’d never heard of a Deputy CEO. He says, “I stop bad ideas.” In other words people that try to deviate from their central plan, anyone who jumps up and says, “Oh, we should be doing this”. He’s saying focus, focus, focus. The key to scale is focus. You get even more within what you’re doing rather than expanding it.

Of course that puts you in much more risk, and that’s why you’re at bigger risk as a scaling company of going out of business.

Avoiding common pitfalls

John: We talked about successful examples, but are there good examples of companies that chose the wrong path, whose mistakes we can learn from? I think straightaway maybe of Lego, who expanded way too far away from the core toys and nearly went bankrupt.

Joe: There’s a saying that any company that moves into a new headquarters, you should short them immediately, and Lego had just moved into a fantastic new headquarters. Changing your logo or moving into a new headquarters are shortable offenses. I’m going to talk about Uber. There’s a professor at NYU, Aswath Damodaran, who’s known as the high priest of valuation, and said very early that the numbers don’t add up. He was criticized for it it. Notably Bill Gurley wrote a very good post, and made some good points, which talked about how they were creating a new category, so it’s impossible to value something that’s new.

Is Uber creating a new category where the old rules don’t apply?

But if you read Naked Capitalism or even the FT columns now, they’re talking about a number of things about Uber. Their culture doesn’t scale for one, but also their subsidy model, and even the idea that they don’t have any moats. Warren Buffett is famous for talking about a defense of a moat. I was in Saudi Arabia recently and Careem is completely dominating what’s going on. There are local, Uber-type services that are much stronger. And we all know that eventually they had to get out of China as well.

John: You touched on the importance of operations. It’s not the sexy thing that people want to write or talk about, but surely there’s a lot we can learn from other industries. What do you think the tech industry can learn?

Joe: There’s an almost total disconnect between operations and the time they have spent understanding how to fulfill and how to keep your demand flexible. The model in startups is about demand whereas operations is coming from the supply point of view. If the Inuit have 50 words for snow, operations has 50 words for time. The classic engineering way of thinking about problems is the magic triangle of good, quick and cheap. I can give you any two, but I can’t give you all three. The modern operations thinking is much more that I can give you all three, and Amazon is a huge example of that.

When I give a speech, people ask, “At what number of people should I hire an HR person? At what number of people should I open an office?” They ask you very specific questions and there are as yet no magic answers to these, but operations have been at least thinking about these problems.

One of the things about Uber of course, is they have no operations. They have no COO, which is another raging red flag (Editor’s note: and no CEO since this interview was recorded). One of the funniest things I see is people will talk to me about “okay, we’re scaling up, what things do we need to do?” And I’ve talked to them about culture and things like that and then I see pictures on Facebook, whatever, where some people are joining the company and how they have welcome packs waiting for them, and I see the journey they’ve made.

Culture at scale

John: What happens to culture as you grow up? Are you trying to protect your original culture or do you just have to resign yourself to the fact that it’s going to change?

Joe: The first thing is this idea of hiring slowly, particularly at the start. Patrick Collison talks a lot about this in the Blitzscaling class.

John: He basically says your first ten hires are your first hundred, because you’re going to hire from their own networks.

Joe: What you need to do isn’t so much hire the person, but hire the person who hires the person. So that works two ways. If you want to build a good technical team, you get one technical guy and then Pied Piper will get the rest. But the point that Patrick was making is very much this idea that you have to get the right 10 people because that’s a tenth of your company and that if you have the wrong person, they’ll set the wrong values.

There’s another interesting theory, which is the Bill Belichick theory of “do your own job”. In the Super Bowl, if you have guys who are dreaming about making the final play, you’ll never get to the final play. What’s your job? It’s literally they sit at this post. José Mourinho says never cross the halfway line if you’re a fullback, you have this notion of do your job. That’s quite interesting. When you start up, you just need someone who’s flexible, who’ll do everything. But after a while, you need people who’ll do their job.

One of the great reasons for scaling is when you get to a certain size, you can hire great people, but great people won’t come to you until you’re a certain size, because they’re like “I can’t deal with the chaos, I need to have the certain things. I’m very good at my job, but I need to have this structure”.

Great people won’t come to you until you’re a certain size.

So culture is one of those absolutely crucial things, and every company has a culture – the question is whether you have sat down and created it. Is it written down, is it acted on? Culture is defined as how people behave when the boss isn’t in the room. It’s also defined as the worst behavior the boss is willing to tolerate. The high priest of this is Tony Hsieh as Zappos. He’s written a book about it and talks about how when they used to interview people they would send a Zappos employee to pick the guy up at the airport in Vegas. How the candidate treated the taxi driver would tell you so much about how they would treat other people.

It’s not all about unicorns

John: Personally I’ve never been comfortable with the whole unicorn tag. A billion-dollar valuation is a nice round figure, but it’s pretty arbitrary. There’s a big focus on unicorns at the moment, but there’s clearly a lot of venture in latter stage finance around. Meanwhile equity markets aren’t that open to tech companies. What are the things we really need to think about as we’re trying to finance scaling. You talked about the Blitz model, is that the only way to go when you need to get venture to scale?

Joe: The first thing is how much of your company you have to give away. Aaron Levie, co-founder at Box, was down to about 2% by the IPO. I don’t think he’ll have any difficulty paying the bill if you go out for a drink with him, but at the same time, Amancio Ortega still owns 59% of Inditex. He built it slowly and had a different model. So that’s the first issue, giving away your company.

The unicorn is a very Silicon Valley thing. In Spain for instance, where I live, you have big companies on the stock exchange, and you have small companies, what you don’t have is a middle, and the real debate is in the middle. Germany has these fantastic Mittelstand companies that are like 800 people, the very best at what they do in the world. They’re clear leaders.

I like to talk about scaling as encouraging entrepreneurs to make the leap from 10-15 people to 100, rather than this idea of Blitzscaling or Hyperscaling to be a unicorn. And if you want to make a social point, that’s where jobs come from. Jobs come from new, fast growing companies.

The advantages of scale

John: Just so we don’t totally focus on the downsides of getting big, what are the advantages of scale? You touched on being able to attract better quality people, but what can you do at scale that you can’t do when you’re in that startup phase?

The number-one inhibitor to scaling is bad management.

Joe: Well, survival. Obviously you’re under threat, but eventually you can withstand more stuff. With Hot Hotels, for instance, Facebook is now starting to sell hotels, Google have been selling hotels indirectly for quite a while. Amazon tried to sell hotels and then stopped. One of the issues is your defensibility and keeping people away. If you look at interviews with (co-founders of) Dropbox and Airbnb, they talk about being afraid of the people in Germany opening clones, so that’s why they Blitzscaled. They wanted to get big quickly to deter other people from getting into the market. So that’s the first reason you need to do this.

The second is that you can hire a better type of person who’s better at their job. The number-one inhibitor to scaling is not lack of money or lack of talent, although they’re certainly factors, it’s bad management. I spend half my time running companies, and there are a whole pile of things that we think we are good at that we’re not actually good at, including recruiting. Even the very best recruiters will tell you they get it right maybe five out of ten times. That’s even the best people. Maybe six times out of ten, but certainly not eight or nine. Someone who’s never studied recruiting, what chance do they have?

A big kind of developmental thing is in life, you find out what you’re good at and get even better at it. Whatever you’re bad at, just let someone else do it. Getting big allows you to deal with those factors. Take away the dependencies and that will allow you to focus on doing what you’re really good at.

John: Joe, thanks for coming in and talking to us about the problems, challenges and enjoyment of scale.

Joe: Enjoyed it, John. Thanks.