Skip to Main Content
HBS Home
  • About
  • Academic Programs
  • Alumni
  • Faculty & Research
  • Baker Library
  • Giving
  • Harvard Business Review
  • Initiatives
  • News
  • Recruit
  • Map / Directions
Working Knowledge
Business Research for Business Leaders
  • Browse All Articles
  • Popular Articles
  • Cold Call Podcast
  • Managing the Future of Work Podcast
  • About Us
  • Book
  • Leadership
  • Marketing
  • Finance
  • Management
  • Entrepreneurship
  • All Topics...
  • Topics
    • COVID-19
    • Entrepreneurship
    • Finance
    • Gender
    • Globalization
    • Leadership
    • Management
    • Negotiation
    • Social Enterprise
    • Strategy
  • Sections
    • Book
    • Podcasts
    • HBS Case
    • In Practice
    • Lessons from the Classroom
    • Op-Ed
    • Research & Ideas
    • Research Event
    • Sharpening Your Skills
    • What Do You Think?
    • Working Paper Summaries
  • Browse All
    In the Platform Economy, Upwork Searches for Better Matches in the Cloud
    25 Jun 2019Cold Call Podcast

    In the Platform Economy, Upwork Searches for Better Matches in the Cloud

    Welcome to the Platform Economy, where business strategists must learn a new form of competition. David Yoffie and Michael Cusumano explain choices facing Upwork, which matches jobs with job hunters in the cloud.
    SUBSCRIBE ON iTUNES
    More Ways to Listen
    Google Podcasts
    Spotify
    LinkedIn
    Email

    Brian Kenny: What do the following companies have in common? InGen, Cyberdyne Systems, Omni Consumer Products, and Tyrell Corporation? Any idea? Well, for one thing, they're all fake. The stuff of Hollywood. And that's good, because for another thing, they are all evil mega-corporations who have used technology to take advantage of naïve consumers in their insidious quest for world domination. Techno thrillers have been surefire hits at the box office for decades, and in the age of big data, these cautionary tales seem all too plausible. Fortunately for humanity, reality is much less thrilling and much more hopeful. In fact, the internet and the rise of social networks have made us more connected than ever, opening the door for a new generation of companies whose goal is to make our lives easier and more productive, by connecting businesses and customers in new ways. Today we'll hear from Professor David Yoffie of Harvard Business School, and Professor Michael Cusumano of MIT Sloan School of Management, about the case study entitled Upwork: Creating the Human Cloud. I'm your host, Brian Kenny, and you're listening to Cold Call, recorded live in Klarman Hall Studio at Harvard Business School.

    David Yoffie's research focuses on competitive strategy, technology, and international competition. Michael Cusumano of MIT Sloan School specializes in strategy, product development, and entrepreneurship in computer software, as well as automobiles and consumer electronics. Together, they are two of the authors of the newly published book, The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power . The third author is Annabelle Gawer, from the University of Surrey. We did not have the funding to bring her over from England today, so it's just us three gentlemen. Welcome.

    David Yoffie: Glad to be here.

    Michael Cusumano: Hey, thank you.

    Brian Kenny: Great to have you both here. I really enjoyed this case about Upwork. I had no idea, actually, about Upwork, until I read the case, but they are a big player in this business of platforms, which you're going to talk a lot more about. David, let me ask you to start. Just how does the case begin? Who's the protagonist, and what's on their mind?

    David Yoffie: Upwork is a platform for connecting companies to freelance labor. In 2016, there was a new CEO, Stephen Kasriel. Or, Stephane Kasriel. He's French. The company had just hit a billon dollars in transaction, making them one of the biggest players in the gig economy. And they have a problem, which is they have to figure out how to connect more clients to more freelancers, because his objective is to grow to $10 billion in transactions over the next few years. In addition, he has to figure out how to avoid the biggest problem on the platform, which is customers, the clients, corporate clients, often go around the platform after they meet their freelancer on the platform, and take the business offline. Then, the related problem to that is, they have to figure out how to make money. They've been around for 20 years. They obviously do a lot of business. But in 20 years, they've never made a profit.

    Brian Kenny: What led you to write the case, and how does it relate back to your research?

    David Yoffie: Michael and I started working on this book with Annabelle Gawer a few years ago, and so we wanted to have a case that would help us think about the problem of developing a business model that worked. This is a company that's been around for a long time, and yet they have a business model which has always been struggling to figure out how to actually make money. You would think that a company that's got 20 years of history, of fairly significant scale, and is all digital, should make money. That's the common wisdom. And yet here we are, and they struggled with it. So, this was a great opportunity to leverage our research on how platforms work into an in-depth study of a particular company.

    Brian Kenny: Great. We're going to continue to refer back to Upwork as we go through the conversation, but I really want to surface some of the ideas in your book. Michael, could you tell us: what is the platform economy?

    Michael Cusumano: It's a term we hear quite a bit these days. It really refers to the economic activity that a large number of companies, that use the internet for primarily selling products and services, or gathering up innovation... By gathering up, I mean incentivizing third parties, programmers, for example to write applications that make some core product more valuable. So, we've seen platform types of businesses for about 30 years. Or at least those driven by personal computer technology or internet technology. Going back to Microsoft, Intel, and then another generation around the internet, and then more recently in social media.

    Brian Kenny: It seems to have taken off, I guess. Clearly your book is an indication that there's a lot more activity here, and you guys are making the bold prediction that these are the companies of the future. What's driving that?

    Michael Cusumano: Certainly technology is driving part of it. The availability of smartphones, wifi technology, internet accessibility from different kinds of devices. We've also seen changes in the economy, and in population. For example, a lot of young people, they don't like to buy things. They don't like to buy cars. They don't necessarily want to buy homes or vacation homes. So, we see companies like Airbnb come about. They're a critical part of the platform economy. Or Uber, particularly in urban areas. Or home household tools, or even pets. I mean, you can use the platform economy to borrow somebody's pet, or borrow somebody's drill. Do just about anything. And a lot of that we associate with the gig or sharing economy, of which Upwork is a part, for providing access to skills that are not normally accessible through, let's say traditional staffing companies, or other kinds of directories.

    Brian Kenny: And certainly, Uber is a great example of the part of the drivers behind the gig economy.

    Michael Cusumano: Well, it's a good and a bad example, in many ways. We talk about Uber a lot in the new book. And it's, for the most part, platforms are bringing two or more sides... We call them our actors in a market... together. And if there's an imbalance in supply and demand for those two sides, and a platform works really well, and the internet or digital platforms are important, because their scale is essentially global, and access is immediate... But in the case of Uber, it's difficult to get drivers and to keep drivers, and it's difficult to get riders, unless you have a price that's below taxi cabs or other means of transportation. So, Uber ends up what we call subsidizing both sides of the market it's trying to bring together, whereas Airbnb charges both sides. It charges people that rent rooms, and it charges people that have rooms to rent. So they're both gig economy types of platforms, and critical parts of the platform economy, but they operate very differently.

    Brian Kenny: So David, let's go back to Upwork for a minute. The book talks about two types of platforms. I'm wondering what those are, and which one Upwork is?

    David Yoffie: Actually, we talk about three types of platforms. We talk about transaction platforms, innovation platforms, and hybrid platforms, which are a combination of transaction and innovation. Basically, a transaction platform is driven by what we call cross side network effects. It's the opportunity of bringing one side of a market on. It makes it more attractive to the other side of a market. So, it's more buyers leads to more sellers, more sellers leads to more buyers, more passengers for Uber leads to more drivers coming on, more drivers coming on actually creates a cross side network effect to enhance the opportunity for more passengers. Now, innovation platforms operate very differently. They are a technological foundation for building complementary products. Things like an operating system, where the purpose of the platform is to get third parties, such as application developers, to write applications which will add value, and will directly be connected to the other side of the market. Which, in this case is often consumers or workers or enterprises. A hybrid is a company that actually combines the two. So, you can think about Google, for example. They have a transaction platform with Android, as well as with Google Search, where they connect advertisers to end users. But they also have an innovation platform around Android, where people write literally millions of applications for a smartphone, that enable you to get all kinds of new capabilities onto that smartphone that Google doesn't provide directly. And they do both. That's what we call a hybrid, because they're doing both innovation and they're doing transactions. In the case of Upwork, Upwork is a classic transaction platform. What they're doing is, they're connecting freelancers anywhere in the world to corporate clients, and they're trying to help companies figure out how can they find someone who can do a very specific, skilled job, that will enable them to do the work for them, without necessarily being a full-time employee.

    Brian Kenny: Okay. This sounds a little bit like some... I don't know. It's like when I think of Manpower, or I think of some of the other staffing agencies that have been out there for a long time... How is Upwork different than those?

    David Yoffie: Manpower is actually very different. I actually met with the CEO of Manpower last week, and he's different, because they actually employ 600,000 people. They literally are the employer of record. And what they do is, they take the people they employ, and they put them into companies on a temporary basis. In the case of Upwork, the freelancer actually works directly for the company. That's why we think of Upwork as a platform, and we would think of Manpower as much more of a distributor and employer of record. So they are actually very different kinds of strategies, but they're all part of the staffing ecosystem. Which is, there are thousands of players who are in broadly the same business as Upwork. Recruiters, temp agencies, other platforms. You can even think of Uber as a competitor in some sense, because what are drivers? They are temporary freelancers, who are working with Uber in order to directly connect to an individual passenger. Now, Upwork primarily focuses in on two areas, which is technical areas... Things like software development and web design, as well as creative design. But they actually allow freelancers to work in hundreds of categories, ranging from writing and marketing, to legal skills and translation services. So, they basically enable any freelancer, anywhere in the world, in almost any possible skill area, to connect with a client who needs something in their specific expertise.

    Brian Kenny: Okay. So, they've chosen a specific vertical, it sounds like, which is in the creative space. So it's not industry-specific, but it's function-specific.

    David Yoffie: Well, there are two verticals that they work with, which are primarily, like I said, in the technical space. A lot of what people do with Upwork is they will do the web application or the web design, or write code on a very specific project. Then they have a whole set of creative people, who will do anything from logo design, to doing your job. To doing marketing.

    Brian Kenny: Are there characteristics of this platform that make it work, and could make it work for other specialties as well?

    David Yoffie: The answer is yes. As I said, they literally have hundreds of verticals, though they have two dominant ones. One of the challenges that Upwork as a company has... And this is one of the challenges we discuss in the case... is that very specialized platforms in vertical spaces tend to be their competitors. So, you can imagine one who is just doing legal work, for example. The question for a company is, if I'm a small business, and I need a lawyer to do something very specialized, will I go to Upwork to find a potential lawyer, or will I go to a legal platform to find a special lawyer? So, they compete with many verticals who are specialized in particular categories, and they also compete broadly across a horizontal spectrum, of essentially any kind of work that you can imagine.

    Brian Kenny: Michael, what are some of the challenges that are common to platforms?

    Michael Cusumano: Well, the first challenge is how do you get started?

    Brian Kenny: Right.

    Michael Cusumano: In the case of Upwork, for example, you can't line up companies and provide skilled workers for them if you don't have any workers online, or if you don't have any companies lined up. We call this a chicken or egg problem. You've got to get started somewhere. Generally, platform businesses, they need to figure out what are the sides of the market they're bringing together. They need to figure out which one is more important than the other. In other words, which one will draw in or pull in the other side, if you have some momentum there. They have to figure out which side they can charge, or which sides they can charge. There should be some kind of a business model under this, and eventually they have to figure out how to manage or govern the ecosystems that they create. In the case of Facebook, for example, as we've seen, we had a lot of nefarious activity from the Russian government, or pseudo government agencies, come in and influence elections. So there's lots of things that can take place on these platforms if they're not governed properly, and that's a huge challenge, once they're big enough to be important.

    Brian Kenny: Which brings me to another question. I mean, the numbers even at Upwork were pretty staggering. Those were huge numbers. You mentioned the sort of workaround that was happening, and that one of the challenges they've faced was that sometimes their customers and their workers were cutting side deals, I guess. Can a platform get too big for its own good?

    Michael Cusumano: Well, it certainly can, in some ways. People often worry about a company like Amazon, which is another characteristic platform economy company. Although it's more complex than that. Amazon has always had a very strong physical presence, as well. But Amazon has so many businesses, and it kind of links them all through data and a common view of a customer, kind of a digital database that tells the company what people are buying. So, it can sell groceries. It could sell books. It has a Prime membership system, which not only gives them data, but also discourages people from shopping anywhere else. And so, the big platforms tend to get bigger. And this is part of the network effects. Which, they're like positive feedback loops. And so, if Amazon has 500 million products, why would you go anywhere else? And if Upwork has millions of potential workers, and all the skills you possibly need, why would you go anywhere else? So this race to get big, and get big relatively fast, is an important part of how platforms have been designed, and part of their strategy. Now, if they get too big, obviously they can come up against antitrust laws or other kinds of laws. But one of the problems we see in Amazon, for example, is they don't really have dominant market share in too many markets. They do in electronic books. Maybe you can say online retail is... But it's not really a market. It's a channel. So, you know. They only have four percent of groceries, but they scare the hell out of every grocery company in the United States, because of what potentially they could do.

    Brian Kenny: David, as you think about where Upwork is in their evolution... They've been around for quite a while, you mentioned... are they sort of at a point where they're poised to take that next leap, and do they run the risks of getting too big?

    David Yoffie: They don't run the risk of getting too big. Just to give you some context, Manpower is a $21 billion company. Upwork is a $200 million company, at the time of the case.

    Michael Cusumano: Well, in terms of revenue.

    David Yoffie: They're a little bit bigger today... In terms of revenue. But even in terms of total transactions, today Upwork is only about a billion and a half, compared to, let's say Manpower is $21 billion. So, they have a lot of upside growth potential. In fact, one of the things that's great for them is that the freelance labor market is exploding. There is an increasing number of freelancers, and increasing demand for freelancers. So, they have an open runway to grow. But again, they've been doing this for 20 years, and they haven't figured out how to make money. So the bottom line for them is, how do they ultimately find a way to drive real profits? And they have, really two or three problems they've got to solve to do that. Number one, the customer acquisition is very expensive, most of that customer acquisition is Google AdWords. They have to advertise heavily in order to get companies to understand that there's a real opportunity, and it's expensive. In addition, they spend a lot of money on R&D, because they're trying to improve the matches. If you're a platform, one of the critical issues is, "How do I get each side of the market to figure out who's the best group to match with?" You have literally hundreds or thousands or tens of thousands of companies trying to find a very specific skill, and then you've got 12 million freelancers, and you're trying to figure out which freelancer is the right one.

    Brian Kenny: Let me just interrupt for a second. So if I'm a customer, and I need a logo designed or something, how does it work? What are the steps I take, and then how do I get matched up with somebody?

    David Yoffie: You'll post on to Upwork what your job requirements are, what you want to do, then Upwork will try to match you with someone whose skills would seem to reflect what you're trying to do. There's a limited number of referrals that Upwork will give to the client. Because they don't want to bombard them with hundreds or thousands, because that would be…

    Brian Kenny: Defeats the purpose.

    David Yoffie: So, it's very important for Upwork to figure out, "How do we make it an efficient transaction platform? What they've said their goal is, is they would like to find the one perfect transaction match for you, and match you with that individual. That's very hard to do. And that's why they're spending a lot of money on R&D, in order to figure out how to make that match better. But then we have this problem, that I find you the absolute right match, and the company and the freelancer start talking to each other, and they actually do it on the platform. And they say, "You know, why are we doing this with Upwork? Upwork is charging us... " At the time of the case, 10%. "Why don't we just take this offline, and we'll come up with a deal that we'll both end up making more money?" That's a real problem for the platform.

    Michael Cusumano: It also happens for other platforms, too. Like Airbnb. If they have a repeated guest for the second time, the guest will go around or disintermediate the platform. So it is a problem, but there are ways of figuring out pricing mechanisms to get around it, at least at some point. But to the extent that they have that disintermediation, it adds to their customer acquisition costs, because they have to keep acquiring new customers.

    David Yoffie: And that's exactly the problem. The problem is, if you lose a customer because they take the transaction offline, and you want to double sales... Or in this case, they're trying to increase their transactions 10x, it's very expensive to keep on acquiring new customers.

    Brian Kenny: Is it expensive for them to acquire the talent side of the equation?

    David Yoffie: No. In fact, it's the opposite. Which, one of their problems is they have something like seven to 10 times more freelancers who want to come on the platform, that they are not necessarily willing to accommodate, because it just makes the matching problem too hard to do.

    Michael Cusumano: And so far, they only accept two percent of applicants. Whereas, there are some other matching staffing platforms that accept anybody. So, one of their, let's say differentiating factors, is that they screen who their staff are very carefully, so there's a quality associated with staffing through Upwork that seems to work well. And, you know, looking at their R&D, most of it is in artificial intelligence, machine learning algorithms, and they probably don't have to spend at their current level forever. One would hope that they would build out that system, like Airbnb did, and then eventually be able to make a profit. I mean, I don't see profitability as being too far in the future for them. So they're a pretty good platform, compared to some others we've seen out there.

    David Yoffie: They're not like Uber. They're close to break even. The business challenge is, after 20 years, you hope to be better than break even. But as Michael said, their R&D levels should go down over time, as a percentage of sales. And their customer acquisition cost should decline, if they can find other ways to create loyalty by their clients and by their freelancers, through either a change in pricing, or changes in the services they provide, that say, "It's worth it to stay on the platform, and the price that we're paying is more than fair."

    Brian Kenny: As you think about the future for platform enterprises, what are some of the things that they should be thinking about?

    David Yoffie: Well, one of the arguments we make in the book is that platforms fundamentally are double-edged swords. We call it a double-edged sword because, I like to say it's the best way to build a trillion dollar business. And yes, that's trillion with a T. In fact, the only trillion dollar businesses that exist in today's world, at the moment, are platforms.

    Brian Kenny: Wow.

    David Yoffie: At the same time, we also know that it is also a great way to lose a fortune. It is also a way that you could potentially undermine democracy, as we've seen Russia being able to do with the Facebook platform. Or to promote violence, like we saw with the New Zealand video on Twitter, YouTube, Facebook... And so, it really is a double-edged sword. The possibility of bad actors getting engaged in a platform, but also the possibility of creating enormous efficiencies, and generating spectacular amounts of value. So, that's the fundamental dilemma. In the book, what we try to argue is that there are ways you can solve these problems as you get big, and as you get dominant. Upwork is not at that level yet. But over time, the same lessons would apply to them. I mean, for example, one of the things we argue is that in an environment where there are very strong network effects, you have the potential for a winner take all outcome, and that leads to the obvious debate around antitrust. We've been hearing that debate frequently today, with regard to Facebook. Part of the problem is that a lot of these very big firms become bullies. Once you become very powerful and very dominant, it's very easy to bully your suppliers, bully each side of the platform. Google has been fined more than 9 billion dollars, largely for bullying behavior, by the European community. And what we argue in the book is, you don't need to be a bully. Once you actually become powerful, you are most likely going to get the business without being a bully. But there's a natural tendency to just use your power, and leverage it in a way that potentially drives the authorities towards an antitrust solution. You know, similarly, we argue that a lot of these platforms are abusing their labor. Uber obviously doesn't have full-time employees for their close to three million drivers that they employ, and the average wage, we know, is somewhere close to $10, if they're lucky. Again, in many cases, when these people are working 40 or 50 hours a week for Uber, we argue once you actually get to that level, you actually should be an employee. And that most of these platforms have the capability and have the resources to afford that. The same thing would be true for Upwork. Now, Upwork, I should also say, they practice what they preach. They use a lot of freelancers to manage their business.

    Michael Cusumano: The platform economy has maybe a little bit of a dark side to it, on the business element, but those things will be sorted out. I mean, the platforms that are really efficient and contributing useful products and services, they're going to stay, I think. The other ones, unfortunately, have attracted a lot of venture capital. That's what has kept them going, even though they're really not viable businesses. So again, Uber, the bigger it gets, the more money it loses. You really don't see a turnaround scenario for them.

    Brian Kenny: So if I'm a traditional business, and I want to become a platform business, can I do that, or am I just out of luck?

    David Yoffie: In the book, we talk about a number of different strategies for traditional businesses. The answer is no, you're not out of luck. But it's hard. It's very hard. We basically have a very simple framework, which we call buy, build, or belong. If you're a traditional business, the first step you can take is try to belong to an existing platform. For many small businesses in particular, that has become a very viable strategy today. You just look at the number of small businesses that have belonged to the eBay platform, or belong to the Amazon platform. Today, third parties represent 60 percent of Amazon's volume. So, there are a lot of very successful small businesses that operate on the platform. They were traditional businesses, brick and mortar, and they recognize that by being on the Amazon platform, it's possible to get a much broader reach.

    Michael Cusumano: There are also some negative cases, like Toys R Us, which was a disaster. Toys"R"Us sold toys through Amazon. Amazon looked at the data and said, "We can sell toys, too. We don't need Toys'R'Us." Basically put them out of business. So, it's also a double-edged sword if you're going to play with a giant like Amazon, and become reliant on them for your internet marketing channel.

    David Yoffie: The second thing we talk about is you can buy a platform, and buying can work really well. If you think about Facebook's purchase of Instagram, for example, which has become an incredibly successful business. It can also work very badly. Remember, MySpace was bought by News Corp, and that was a disaster. The example we use in the book is Walmart. Walmart, for almost 17, 18 years, really struggled to try and create a platform to compete with Amazon, and it is finally starting to get some traction because it purchased Jet.com. I think part of the reason that worked is that they gave all of the responsibility for the platform to Jet. And if you look, over the last two years, they've been averaging something close to 40 to 45 percent growth in their e-commerce business.

    Michael Cusumano: But that's largely adding an e-commerce channel, whereas Amazon has that. It has an online store, which is equivalent to Walmart.com, but it's built up the Amazon Marketplace since roughly 1999, 2000, and that is bringing third parties together. So, high volume goods, where Amazon thinks it can make a profit, it buys them in bulk and resells them on its platform. But seamlessly on the platform, you also see the Marketplace. So, if you want to buy the books that David and I wrote in the 1980s or early 1990s-

    Brian Kenny: Big sellers, I think.

    Michael Cusumano: ... you can see them on Amazon. You don't even know that Amazon doesn't hold them in inventory, but it knows who has them, which little bookstore in southern California. And so, they will make that match for you, so the long-tail items. So, that's a superior part of Amazon that I think Walmart hasn't really figured out how to match, but they don't need to match that. What they really needed to do is have an internet presence for retail.

    David Yoffie: What Jet has done--this is the change--before they bought Jet, there were only about 500,000 SKUs on the platform from third parties. Today, there's 75 million.

    Brian Kenny: Wow.

    David Yoffie: Again, they're not Amazon. Amazon's got 500 million. But they literally went from 500,000 to 75 million, in the last two years. So, they're still tiny compared to Amazon. They're not anywhere close to being a serious threat to Amazon directly. But, the purchase has gotten them into the game.

    Brian Kenny: You both taught this case. I don't want you to give away any secrets. But just, any surprises as you discussed it with students, many of whom obviously use these platforms and know them well?

    David Yoffie: Well, the fun thing about this case for students is that in every class, there's always some students who have hired people on Upwork for a startup that they're trying to build, or they've been a freelancer, and have been hired through Upwork.

    Brian Kenny: You get both perspectives.

    David Yoffie: So you get both sides. And some students love the experience, some students hate the experience. Some students have circumvented Upwork, as they've found clients on the platform. So, there's always a very strong point of view. I think that part of the attraction of the case is trying to figure out this problem of how do you actually make real money? As I started out by saying, the normal thing everybody expects is it's a digital business. Of course they should make lots of money. And here, they've been around for 20 years, a billion in transactions, and it's just close to break even. So, that's a great challenge for students to figure out. There must be a way to make money, given the digital nature of this business, and the fact that they are well established. They're a well-known brand. They operate globally. They have 12 million freelancers all around the world. They have five million corporate clients. Why can't you make money?

    Brian Kenny: Michael, any surprises from your perspective?

    Michael Cusumano: We looked more carefully at the comparison of Upwork to other companies that had tried to turn their product into a service. And in almost all cases, the common theme is that they keep losing customers, and they have to keep replenishing those customers. So, very few software as a service companies make any money, because their customer churn rates are so high. That's why we kind of liked Upwork more than some of these other companies, because they had a valuable service. It may be one time. If they can figure out a pricing model that goes forward into the future... And that's where we debated, and I won't give away the conversation I had with Stephane afterwards. But his thinking is very much like the debate we had in our class, that they have to do something to build loyalty with their corporate clients. The workers all want to be on the site, and there's no cost to them. But to get the companies to stay with them multiple times, and not disintermediate, or allow disintermediation, was the critical thing. But we came up with a number of pricing solutions that Upwork can do. That would mean they make less money on repeated engagements, but they have loyal customers. So, they need to think their corporate customers in terms of like a customer lifetime value model.

    Brian Kenny: Now I'm going to ask a question that might be unfair, but my last question to each of you: What's the next big platform that we should be looking for?

    David Yoffie: In the book, we talk about four next big platforms. One that I've been particularly interested in, and wrote a case on, is on voice. Alexa versus Siri versus Hey Google and a variety of others. Voice is right now one of the hot platform battles of the internet giants, and that's going to continue for quite some time. We're still at the very early stages of that. A second next-generation platform is, what is going to happen in a world of driverless cars? How does that change the economics of an Uber or Lyft or Waymo, or any of the other players who are likely to enter the space. That is also a platform that is likely to evolve quite significantly.

    Michael Cusumano: Yeah, two more forward-looking platforms that also have some current traction were genetic engineering, particularly CRISPR, the set of gene editing tools, and also quantum computing. You have other firms emerging that are building tools, others with building diagnostic tests, others building therapies. Some universities are involved in this as well, building libraries. So, we think there's some potential there. In the case of quantum computing, we have a lot of startups. It's actually quite hot right now. It's also a very difficult technology. We see this as a new kind of innovation platform, where you'll have some core technologies, and then increasing numbers of application developers building applications. There's some important progress in cryptography, and also secure communications, using quantum computers. So, governments are interested. CIA, China. A lot of communications companies.

    Brian Kenny: Sounds like it could be another techno thriller being developed.

    Michael Cusumano: It could. We should write a book, The Business of Quantum Computing. It's been slowly evolving, because the science is complicated. But even at MIT, we have maybe four or five quantum computers running in the labs of several of our professors. We have several classes on how you program quantum computers. I've been teaching this case... Which is an HBS case, although it was written originally by an MIT faculty and a Harvard faculty... I've been teaching it for about 15 years. So it's moving slowly, but just in the last two or three years, we've seen dozens of companies jump in. A lot of startups, venture capital coming as well. So, it's going to be a new innovation platform for solving particular types of very complex, combinatorial optimization problems, some other simulation problems, and also cryptography problems. It's also a dangerous area. Here's another double-edged sword. Because some of these computers, if they really play out as we think they will, they will be able to build cryptographic keys that we cannot break.

    David Yoffie: By the way, equally dangerous is that those quantum computers, should they succeed, will probably be able to break all existing cryptography that we have today.

    Michael Cusumano: And steal all the financial assets.

    Brian Kenny: That sounds just like the intro that I read about the techno thrillers. We've come right back. That's full circle. Thank you both for being with me today.

    David Yoffie: Thank you for having us.

    Michael Cusumano: Thanks very much.

    Brian Kenny: The book, again, is called The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power. If you enjoy Cold Call, you should check out our other podcast from Harvard Business School, including After Hours, Skydeck, and Managing the Future of Work. Find them on Apple Podcasts, or wherever you listen. Thanks again for joining us. I'm your host, Brian Kenny, and you've been listening to Cold Call, an official podcast of Harvard Business School, brought to you by the HBR Presents Network.

     Read more

    Brian Kenny: What do the following companies have in common? InGen, Cyberdyne Systems, Omni Consumer Products, and Tyrell Corporation? Any idea? Well, for one thing, they're all fake. The stuff of Hollywood. And that's good, because for another thing, they are all evil mega-corporations who have used technology to take advantage of naïve consumers in their insidious quest for world domination. Techno thrillers have been surefire hits at the box office for decades, and in the age of big data, these cautionary tales seem all too plausible. Fortunately for humanity, reality is much less thrilling and much more hopeful. In fact, the internet and the rise of social networks have made us more connected than ever, opening the door for a new generation of companies whose goal is to make our lives easier and more productive, by connecting businesses and customers in new ways. Today we'll hear from Professor David Yoffie of Harvard Business School, and Professor Michael Cusumano of MIT Sloan School of Management, about the case study entitled Upwork: Creating the Human Cloud. I'm your host, Brian Kenny, and you're listening to Cold Call, recorded live in Klarman Hall Studio at Harvard Business School.

    David Yoffie's research focuses on competitive strategy, technology, and international competition. Michael Cusumano of MIT Sloan School specializes in strategy, product development, and entrepreneurship in computer software, as well as automobiles and consumer electronics. Together, they are two of the authors of the newly published book, The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power . The third author is Annabelle Gawer, from the University of Surrey. We did not have the funding to bring her over from England today, so it's just us three gentlemen. Welcome.

    David Yoffie: Glad to be here.

    Michael Cusumano: Hey, thank you.

    Brian Kenny: Great to have you both here. I really enjoyed this case about Upwork. I had no idea, actually, about Upwork, until I read the case, but they are a big player in this business of platforms, which you're going to talk a lot more about. David, let me ask you to start. Just how does the case begin? Who's the protagonist, and what's on their mind?

    David Yoffie: Upwork is a platform for connecting companies to freelance labor. In 2016, there was a new CEO, Stephen Kasriel. Or, Stephane Kasriel. He's French. The company had just hit a billon dollars in transaction, making them one of the biggest players in the gig economy. And they have a problem, which is they have to figure out how to connect more clients to more freelancers, because his objective is to grow to $10 billion in transactions over the next few years. In addition, he has to figure out how to avoid the biggest problem on the platform, which is customers, the clients, corporate clients, often go around the platform after they meet their freelancer on the platform, and take the business offline. Then, the related problem to that is, they have to figure out how to make money. They've been around for 20 years. They obviously do a lot of business. But in 20 years, they've never made a profit.

    Brian Kenny: What led you to write the case, and how does it relate back to your research?

    David Yoffie: Michael and I started working on this book with Annabelle Gawer a few years ago, and so we wanted to have a case that would help us think about the problem of developing a business model that worked. This is a company that's been around for a long time, and yet they have a business model which has always been struggling to figure out how to actually make money. You would think that a company that's got 20 years of history, of fairly significant scale, and is all digital, should make money. That's the common wisdom. And yet here we are, and they struggled with it. So, this was a great opportunity to leverage our research on how platforms work into an in-depth study of a particular company.

    Brian Kenny: Great. We're going to continue to refer back to Upwork as we go through the conversation, but I really want to surface some of the ideas in your book. Michael, could you tell us: what is the platform economy?

    Michael Cusumano: It's a term we hear quite a bit these days. It really refers to the economic activity that a large number of companies, that use the internet for primarily selling products and services, or gathering up innovation... By gathering up, I mean incentivizing third parties, programmers, for example to write applications that make some core product more valuable. So, we've seen platform types of businesses for about 30 years. Or at least those driven by personal computer technology or internet technology. Going back to Microsoft, Intel, and then another generation around the internet, and then more recently in social media.

    Brian Kenny: It seems to have taken off, I guess. Clearly your book is an indication that there's a lot more activity here, and you guys are making the bold prediction that these are the companies of the future. What's driving that?

    Michael Cusumano: Certainly technology is driving part of it. The availability of smartphones, wifi technology, internet accessibility from different kinds of devices. We've also seen changes in the economy, and in population. For example, a lot of young people, they don't like to buy things. They don't like to buy cars. They don't necessarily want to buy homes or vacation homes. So, we see companies like Airbnb come about. They're a critical part of the platform economy. Or Uber, particularly in urban areas. Or home household tools, or even pets. I mean, you can use the platform economy to borrow somebody's pet, or borrow somebody's drill. Do just about anything. And a lot of that we associate with the gig or sharing economy, of which Upwork is a part, for providing access to skills that are not normally accessible through, let's say traditional staffing companies, or other kinds of directories.

    Brian Kenny: And certainly, Uber is a great example of the part of the drivers behind the gig economy.

    Michael Cusumano: Well, it's a good and a bad example, in many ways. We talk about Uber a lot in the new book. And it's, for the most part, platforms are bringing two or more sides... We call them our actors in a market... together. And if there's an imbalance in supply and demand for those two sides, and a platform works really well, and the internet or digital platforms are important, because their scale is essentially global, and access is immediate... But in the case of Uber, it's difficult to get drivers and to keep drivers, and it's difficult to get riders, unless you have a price that's below taxi cabs or other means of transportation. So, Uber ends up what we call subsidizing both sides of the market it's trying to bring together, whereas Airbnb charges both sides. It charges people that rent rooms, and it charges people that have rooms to rent. So they're both gig economy types of platforms, and critical parts of the platform economy, but they operate very differently.

    Brian Kenny: So David, let's go back to Upwork for a minute. The book talks about two types of platforms. I'm wondering what those are, and which one Upwork is?

    David Yoffie: Actually, we talk about three types of platforms. We talk about transaction platforms, innovation platforms, and hybrid platforms, which are a combination of transaction and innovation. Basically, a transaction platform is driven by what we call cross side network effects. It's the opportunity of bringing one side of a market on. It makes it more attractive to the other side of a market. So, it's more buyers leads to more sellers, more sellers leads to more buyers, more passengers for Uber leads to more drivers coming on, more drivers coming on actually creates a cross side network effect to enhance the opportunity for more passengers. Now, innovation platforms operate very differently. They are a technological foundation for building complementary products. Things like an operating system, where the purpose of the platform is to get third parties, such as application developers, to write applications which will add value, and will directly be connected to the other side of the market. Which, in this case is often consumers or workers or enterprises. A hybrid is a company that actually combines the two. So, you can think about Google, for example. They have a transaction platform with Android, as well as with Google Search, where they connect advertisers to end users. But they also have an innovation platform around Android, where people write literally millions of applications for a smartphone, that enable you to get all kinds of new capabilities onto that smartphone that Google doesn't provide directly. And they do both. That's what we call a hybrid, because they're doing both innovation and they're doing transactions. In the case of Upwork, Upwork is a classic transaction platform. What they're doing is, they're connecting freelancers anywhere in the world to corporate clients, and they're trying to help companies figure out how can they find someone who can do a very specific, skilled job, that will enable them to do the work for them, without necessarily being a full-time employee.

    Brian Kenny: Okay. This sounds a little bit like some... I don't know. It's like when I think of Manpower, or I think of some of the other staffing agencies that have been out there for a long time... How is Upwork different than those?

    David Yoffie: Manpower is actually very different. I actually met with the CEO of Manpower last week, and he's different, because they actually employ 600,000 people. They literally are the employer of record. And what they do is, they take the people they employ, and they put them into companies on a temporary basis. In the case of Upwork, the freelancer actually works directly for the company. That's why we think of Upwork as a platform, and we would think of Manpower as much more of a distributor and employer of record. So they are actually very different kinds of strategies, but they're all part of the staffing ecosystem. Which is, there are thousands of players who are in broadly the same business as Upwork. Recruiters, temp agencies, other platforms. You can even think of Uber as a competitor in some sense, because what are drivers? They are temporary freelancers, who are working with Uber in order to directly connect to an individual passenger. Now, Upwork primarily focuses in on two areas, which is technical areas... Things like software development and web design, as well as creative design. But they actually allow freelancers to work in hundreds of categories, ranging from writing and marketing, to legal skills and translation services. So, they basically enable any freelancer, anywhere in the world, in almost any possible skill area, to connect with a client who needs something in their specific expertise.

    Brian Kenny: Okay. So, they've chosen a specific vertical, it sounds like, which is in the creative space. So it's not industry-specific, but it's function-specific.

    David Yoffie: Well, there are two verticals that they work with, which are primarily, like I said, in the technical space. A lot of what people do with Upwork is they will do the web application or the web design, or write code on a very specific project. Then they have a whole set of creative people, who will do anything from logo design, to doing your job. To doing marketing.

    Brian Kenny: Are there characteristics of this platform that make it work, and could make it work for other specialties as well?

    David Yoffie: The answer is yes. As I said, they literally have hundreds of verticals, though they have two dominant ones. One of the challenges that Upwork as a company has... And this is one of the challenges we discuss in the case... is that very specialized platforms in vertical spaces tend to be their competitors. So, you can imagine one who is just doing legal work, for example. The question for a company is, if I'm a small business, and I need a lawyer to do something very specialized, will I go to Upwork to find a potential lawyer, or will I go to a legal platform to find a special lawyer? So, they compete with many verticals who are specialized in particular categories, and they also compete broadly across a horizontal spectrum, of essentially any kind of work that you can imagine.

    Brian Kenny: Michael, what are some of the challenges that are common to platforms?

    Michael Cusumano: Well, the first challenge is how do you get started?

    Brian Kenny: Right.

    Michael Cusumano: In the case of Upwork, for example, you can't line up companies and provide skilled workers for them if you don't have any workers online, or if you don't have any companies lined up. We call this a chicken or egg problem. You've got to get started somewhere. Generally, platform businesses, they need to figure out what are the sides of the market they're bringing together. They need to figure out which one is more important than the other. In other words, which one will draw in or pull in the other side, if you have some momentum there. They have to figure out which side they can charge, or which sides they can charge. There should be some kind of a business model under this, and eventually they have to figure out how to manage or govern the ecosystems that they create. In the case of Facebook, for example, as we've seen, we had a lot of nefarious activity from the Russian government, or pseudo government agencies, come in and influence elections. So there's lots of things that can take place on these platforms if they're not governed properly, and that's a huge challenge, once they're big enough to be important.

    Brian Kenny: Which brings me to another question. I mean, the numbers even at Upwork were pretty staggering. Those were huge numbers. You mentioned the sort of workaround that was happening, and that one of the challenges they've faced was that sometimes their customers and their workers were cutting side deals, I guess. Can a platform get too big for its own good?

    Michael Cusumano: Well, it certainly can, in some ways. People often worry about a company like Amazon, which is another characteristic platform economy company. Although it's more complex than that. Amazon has always had a very strong physical presence, as well. But Amazon has so many businesses, and it kind of links them all through data and a common view of a customer, kind of a digital database that tells the company what people are buying. So, it can sell groceries. It could sell books. It has a Prime membership system, which not only gives them data, but also discourages people from shopping anywhere else. And so, the big platforms tend to get bigger. And this is part of the network effects. Which, they're like positive feedback loops. And so, if Amazon has 500 million products, why would you go anywhere else? And if Upwork has millions of potential workers, and all the skills you possibly need, why would you go anywhere else? So this race to get big, and get big relatively fast, is an important part of how platforms have been designed, and part of their strategy. Now, if they get too big, obviously they can come up against antitrust laws or other kinds of laws. But one of the problems we see in Amazon, for example, is they don't really have dominant market share in too many markets. They do in electronic books. Maybe you can say online retail is... But it's not really a market. It's a channel. So, you know. They only have four percent of groceries, but they scare the hell out of every grocery company in the United States, because of what potentially they could do.

    Brian Kenny: David, as you think about where Upwork is in their evolution... They've been around for quite a while, you mentioned... are they sort of at a point where they're poised to take that next leap, and do they run the risks of getting too big?

    David Yoffie: They don't run the risk of getting too big. Just to give you some context, Manpower is a $21 billion company. Upwork is a $200 million company, at the time of the case.

    Michael Cusumano: Well, in terms of revenue.

    David Yoffie: They're a little bit bigger today... In terms of revenue. But even in terms of total transactions, today Upwork is only about a billion and a half, compared to, let's say Manpower is $21 billion. So, they have a lot of upside growth potential. In fact, one of the things that's great for them is that the freelance labor market is exploding. There is an increasing number of freelancers, and increasing demand for freelancers. So, they have an open runway to grow. But again, they've been doing this for 20 years, and they haven't figured out how to make money. So the bottom line for them is, how do they ultimately find a way to drive real profits? And they have, really two or three problems they've got to solve to do that. Number one, the customer acquisition is very expensive, most of that customer acquisition is Google AdWords. They have to advertise heavily in order to get companies to understand that there's a real opportunity, and it's expensive. In addition, they spend a lot of money on R&D, because they're trying to improve the matches. If you're a platform, one of the critical issues is, "How do I get each side of the market to figure out who's the best group to match with?" You have literally hundreds or thousands or tens of thousands of companies trying to find a very specific skill, and then you've got 12 million freelancers, and you're trying to figure out which freelancer is the right one.

    Brian Kenny: Let me just interrupt for a second. So if I'm a customer, and I need a logo designed or something, how does it work? What are the steps I take, and then how do I get matched up with somebody?

    David Yoffie: You'll post on to Upwork what your job requirements are, what you want to do, then Upwork will try to match you with someone whose skills would seem to reflect what you're trying to do. There's a limited number of referrals that Upwork will give to the client. Because they don't want to bombard them with hundreds or thousands, because that would be…

    Brian Kenny: Defeats the purpose.

    David Yoffie: So, it's very important for Upwork to figure out, "How do we make it an efficient transaction platform? What they've said their goal is, is they would like to find the one perfect transaction match for you, and match you with that individual. That's very hard to do. And that's why they're spending a lot of money on R&D, in order to figure out how to make that match better. But then we have this problem, that I find you the absolute right match, and the company and the freelancer start talking to each other, and they actually do it on the platform. And they say, "You know, why are we doing this with Upwork? Upwork is charging us... " At the time of the case, 10%. "Why don't we just take this offline, and we'll come up with a deal that we'll both end up making more money?" That's a real problem for the platform.

    Michael Cusumano: It also happens for other platforms, too. Like Airbnb. If they have a repeated guest for the second time, the guest will go around or disintermediate the platform. So it is a problem, but there are ways of figuring out pricing mechanisms to get around it, at least at some point. But to the extent that they have that disintermediation, it adds to their customer acquisition costs, because they have to keep acquiring new customers.

    David Yoffie: And that's exactly the problem. The problem is, if you lose a customer because they take the transaction offline, and you want to double sales... Or in this case, they're trying to increase their transactions 10x, it's very expensive to keep on acquiring new customers.

    Brian Kenny: Is it expensive for them to acquire the talent side of the equation?

    David Yoffie: No. In fact, it's the opposite. Which, one of their problems is they have something like seven to 10 times more freelancers who want to come on the platform, that they are not necessarily willing to accommodate, because it just makes the matching problem too hard to do.

    Michael Cusumano: And so far, they only accept two percent of applicants. Whereas, there are some other matching staffing platforms that accept anybody. So, one of their, let's say differentiating factors, is that they screen who their staff are very carefully, so there's a quality associated with staffing through Upwork that seems to work well. And, you know, looking at their R&D, most of it is in artificial intelligence, machine learning algorithms, and they probably don't have to spend at their current level forever. One would hope that they would build out that system, like Airbnb did, and then eventually be able to make a profit. I mean, I don't see profitability as being too far in the future for them. So they're a pretty good platform, compared to some others we've seen out there.

    David Yoffie: They're not like Uber. They're close to break even. The business challenge is, after 20 years, you hope to be better than break even. But as Michael said, their R&D levels should go down over time, as a percentage of sales. And their customer acquisition cost should decline, if they can find other ways to create loyalty by their clients and by their freelancers, through either a change in pricing, or changes in the services they provide, that say, "It's worth it to stay on the platform, and the price that we're paying is more than fair."

    Brian Kenny: As you think about the future for platform enterprises, what are some of the things that they should be thinking about?

    David Yoffie: Well, one of the arguments we make in the book is that platforms fundamentally are double-edged swords. We call it a double-edged sword because, I like to say it's the best way to build a trillion dollar business. And yes, that's trillion with a T. In fact, the only trillion dollar businesses that exist in today's world, at the moment, are platforms.

    Brian Kenny: Wow.

    David Yoffie: At the same time, we also know that it is also a great way to lose a fortune. It is also a way that you could potentially undermine democracy, as we've seen Russia being able to do with the Facebook platform. Or to promote violence, like we saw with the New Zealand video on Twitter, YouTube, Facebook... And so, it really is a double-edged sword. The possibility of bad actors getting engaged in a platform, but also the possibility of creating enormous efficiencies, and generating spectacular amounts of value. So, that's the fundamental dilemma. In the book, what we try to argue is that there are ways you can solve these problems as you get big, and as you get dominant. Upwork is not at that level yet. But over time, the same lessons would apply to them. I mean, for example, one of the things we argue is that in an environment where there are very strong network effects, you have the potential for a winner take all outcome, and that leads to the obvious debate around antitrust. We've been hearing that debate frequently today, with regard to Facebook. Part of the problem is that a lot of these very big firms become bullies. Once you become very powerful and very dominant, it's very easy to bully your suppliers, bully each side of the platform. Google has been fined more than 9 billion dollars, largely for bullying behavior, by the European community. And what we argue in the book is, you don't need to be a bully. Once you actually become powerful, you are most likely going to get the business without being a bully. But there's a natural tendency to just use your power, and leverage it in a way that potentially drives the authorities towards an antitrust solution. You know, similarly, we argue that a lot of these platforms are abusing their labor. Uber obviously doesn't have full-time employees for their close to three million drivers that they employ, and the average wage, we know, is somewhere close to $10, if they're lucky. Again, in many cases, when these people are working 40 or 50 hours a week for Uber, we argue once you actually get to that level, you actually should be an employee. And that most of these platforms have the capability and have the resources to afford that. The same thing would be true for Upwork. Now, Upwork, I should also say, they practice what they preach. They use a lot of freelancers to manage their business.

    Michael Cusumano: The platform economy has maybe a little bit of a dark side to it, on the business element, but those things will be sorted out. I mean, the platforms that are really efficient and contributing useful products and services, they're going to stay, I think. The other ones, unfortunately, have attracted a lot of venture capital. That's what has kept them going, even though they're really not viable businesses. So again, Uber, the bigger it gets, the more money it loses. You really don't see a turnaround scenario for them.

    Brian Kenny: So if I'm a traditional business, and I want to become a platform business, can I do that, or am I just out of luck?

    David Yoffie: In the book, we talk about a number of different strategies for traditional businesses. The answer is no, you're not out of luck. But it's hard. It's very hard. We basically have a very simple framework, which we call buy, build, or belong. If you're a traditional business, the first step you can take is try to belong to an existing platform. For many small businesses in particular, that has become a very viable strategy today. You just look at the number of small businesses that have belonged to the eBay platform, or belong to the Amazon platform. Today, third parties represent 60 percent of Amazon's volume. So, there are a lot of very successful small businesses that operate on the platform. They were traditional businesses, brick and mortar, and they recognize that by being on the Amazon platform, it's possible to get a much broader reach.

    Michael Cusumano: There are also some negative cases, like Toys R Us, which was a disaster. Toys"R"Us sold toys through Amazon. Amazon looked at the data and said, "We can sell toys, too. We don't need Toys'R'Us." Basically put them out of business. So, it's also a double-edged sword if you're going to play with a giant like Amazon, and become reliant on them for your internet marketing channel.

    David Yoffie: The second thing we talk about is you can buy a platform, and buying can work really well. If you think about Facebook's purchase of Instagram, for example, which has become an incredibly successful business. It can also work very badly. Remember, MySpace was bought by News Corp, and that was a disaster. The example we use in the book is Walmart. Walmart, for almost 17, 18 years, really struggled to try and create a platform to compete with Amazon, and it is finally starting to get some traction because it purchased Jet.com. I think part of the reason that worked is that they gave all of the responsibility for the platform to Jet. And if you look, over the last two years, they've been averaging something close to 40 to 45 percent growth in their e-commerce business.

    Michael Cusumano: But that's largely adding an e-commerce channel, whereas Amazon has that. It has an online store, which is equivalent to Walmart.com, but it's built up the Amazon Marketplace since roughly 1999, 2000, and that is bringing third parties together. So, high volume goods, where Amazon thinks it can make a profit, it buys them in bulk and resells them on its platform. But seamlessly on the platform, you also see the Marketplace. So, if you want to buy the books that David and I wrote in the 1980s or early 1990s-

    Brian Kenny: Big sellers, I think.

    Michael Cusumano: ... you can see them on Amazon. You don't even know that Amazon doesn't hold them in inventory, but it knows who has them, which little bookstore in southern California. And so, they will make that match for you, so the long-tail items. So, that's a superior part of Amazon that I think Walmart hasn't really figured out how to match, but they don't need to match that. What they really needed to do is have an internet presence for retail.

    David Yoffie: What Jet has done--this is the change--before they bought Jet, there were only about 500,000 SKUs on the platform from third parties. Today, there's 75 million.

    Brian Kenny: Wow.

    David Yoffie: Again, they're not Amazon. Amazon's got 500 million. But they literally went from 500,000 to 75 million, in the last two years. So, they're still tiny compared to Amazon. They're not anywhere close to being a serious threat to Amazon directly. But, the purchase has gotten them into the game.

    Brian Kenny: You both taught this case. I don't want you to give away any secrets. But just, any surprises as you discussed it with students, many of whom obviously use these platforms and know them well?

    David Yoffie: Well, the fun thing about this case for students is that in every class, there's always some students who have hired people on Upwork for a startup that they're trying to build, or they've been a freelancer, and have been hired through Upwork.

    Brian Kenny: You get both perspectives.

    David Yoffie: So you get both sides. And some students love the experience, some students hate the experience. Some students have circumvented Upwork, as they've found clients on the platform. So, there's always a very strong point of view. I think that part of the attraction of the case is trying to figure out this problem of how do you actually make real money? As I started out by saying, the normal thing everybody expects is it's a digital business. Of course they should make lots of money. And here, they've been around for 20 years, a billion in transactions, and it's just close to break even. So, that's a great challenge for students to figure out. There must be a way to make money, given the digital nature of this business, and the fact that they are well established. They're a well-known brand. They operate globally. They have 12 million freelancers all around the world. They have five million corporate clients. Why can't you make money?

    Brian Kenny: Michael, any surprises from your perspective?

    Michael Cusumano: We looked more carefully at the comparison of Upwork to other companies that had tried to turn their product into a service. And in almost all cases, the common theme is that they keep losing customers, and they have to keep replenishing those customers. So, very few software as a service companies make any money, because their customer churn rates are so high. That's why we kind of liked Upwork more than some of these other companies, because they had a valuable service. It may be one time. If they can figure out a pricing model that goes forward into the future... And that's where we debated, and I won't give away the conversation I had with Stephane afterwards. But his thinking is very much like the debate we had in our class, that they have to do something to build loyalty with their corporate clients. The workers all want to be on the site, and there's no cost to them. But to get the companies to stay with them multiple times, and not disintermediate, or allow disintermediation, was the critical thing. But we came up with a number of pricing solutions that Upwork can do. That would mean they make less money on repeated engagements, but they have loyal customers. So, they need to think their corporate customers in terms of like a customer lifetime value model.

    Brian Kenny: Now I'm going to ask a question that might be unfair, but my last question to each of you: What's the next big platform that we should be looking for?

    David Yoffie: In the book, we talk about four next big platforms. One that I've been particularly interested in, and wrote a case on, is on voice. Alexa versus Siri versus Hey Google and a variety of others. Voice is right now one of the hot platform battles of the internet giants, and that's going to continue for quite some time. We're still at the very early stages of that. A second next-generation platform is, what is going to happen in a world of driverless cars? How does that change the economics of an Uber or Lyft or Waymo, or any of the other players who are likely to enter the space. That is also a platform that is likely to evolve quite significantly.

    Michael Cusumano: Yeah, two more forward-looking platforms that also have some current traction were genetic engineering, particularly CRISPR, the set of gene editing tools, and also quantum computing. You have other firms emerging that are building tools, others with building diagnostic tests, others building therapies. Some universities are involved in this as well, building libraries. So, we think there's some potential there. In the case of quantum computing, we have a lot of startups. It's actually quite hot right now. It's also a very difficult technology. We see this as a new kind of innovation platform, where you'll have some core technologies, and then increasing numbers of application developers building applications. There's some important progress in cryptography, and also secure communications, using quantum computers. So, governments are interested. CIA, China. A lot of communications companies.

    Brian Kenny: Sounds like it could be another techno thriller being developed.

    Michael Cusumano: It could. We should write a book, The Business of Quantum Computing. It's been slowly evolving, because the science is complicated. But even at MIT, we have maybe four or five quantum computers running in the labs of several of our professors. We have several classes on how you program quantum computers. I've been teaching this case... Which is an HBS case, although it was written originally by an MIT faculty and a Harvard faculty... I've been teaching it for about 15 years. So it's moving slowly, but just in the last two or three years, we've seen dozens of companies jump in. A lot of startups, venture capital coming as well. So, it's going to be a new innovation platform for solving particular types of very complex, combinatorial optimization problems, some other simulation problems, and also cryptography problems. It's also a dangerous area. Here's another double-edged sword. Because some of these computers, if they really play out as we think they will, they will be able to build cryptographic keys that we cannot break.

    David Yoffie: By the way, equally dangerous is that those quantum computers, should they succeed, will probably be able to break all existing cryptography that we have today.

    Michael Cusumano: And steal all the financial assets.

    Brian Kenny: That sounds just like the intro that I read about the techno thrillers. We've come right back. That's full circle. Thank you both for being with me today.

    David Yoffie: Thank you for having us.

    Michael Cusumano: Thanks very much.

    Brian Kenny: The book, again, is called The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power. If you enjoy Cold Call, you should check out our other podcast from Harvard Business School, including After Hours, Skydeck, and Managing the Future of Work. Find them on Apple Podcasts, or wherever you listen. Thanks again for joining us. I'm your host, Brian Kenny, and you've been listening to Cold Call, an official podcast of Harvard Business School, brought to you by the HBR Presents Network.

     Close

    Cold Call Podcast
    Post A Comment
    In order to be published, comments must be on-topic and civil in tone, with no name calling or personal attacks. Your comment may be edited for clarity and length.
      Trending
        • 25 Jan 2022
        • Research & Ideas

        More Proof That Money Can Buy Happiness (or a Life with Less Stress)

        • 14 Mar 2023
        • In Practice

        What Does the Failure of Silicon Valley Bank Say About the State of Finance?

        • 25 Feb 2019
        • Research & Ideas

        How Gender Stereotypes Kill a Woman’s Self-Confidence

        • 15 Nov 2022
        • Book

        Stop Ignoring Bad Behavior: 6 Tips for Better Ethics at Work

        • 07 Mar 2023
        • HBS Case

        ChatGPT: Did Big Tech Set Up the World for an AI Bias Disaster?

    David B. Yoffie
    David B. Yoffie
    Max and Doris Starr Professor of International Business Administration
    Contact
    Send an email
    → More Articles
    Find Related Articles
    • Market Platforms
    • Competitive Strategy
    • Partners and Partnerships
    • Alliances
    • Web Services

    Sign up for our weekly newsletter

    Interested in improving your business? Learn about fresh research and ideas from Harvard Business School faculty.
    This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
    ǁ
    Campus Map
    Harvard Business School Working Knowledge
    Baker Library | Bloomberg Center
    Soldiers Field
    Boston, MA 02163
    Email: Editor-in-Chief
    →Map & Directions
    →More Contact Information
    • Make a Gift
    • Site Map
    • Jobs
    • Harvard University
    • Trademarks
    • Policies
    • Accessibility
    • Digital Accessibility
    Copyright © President & Fellows of Harvard College