Low Risk, High Reward: Why Venture Capital Thrives in the Digital World
TheFamily Papers #003
By Nicolas Colin (Co-Founder & Director) | The Family
William Janeway’s book, Doing Capitalism in the Innovation Economy, is the Bible when it comes to understanding the digital economy and how it is financed. Of particular interest is an idea about the risks investors are willing to take. Venture capital exists in the digital economy not because it is riskier, but because it is in fact less risky than other parts of the economy when it comes to research and development. Venture capitalists are willing to finance radically innovative companies not because they’ve suddenly embraced the idea of risking everything, but because some traditional risks were eliminated forever by various public and private investments, notably in the technological platform that is the Internet.
Technology, really?
Contrary to what is widely assumed, technology is not an issue for a tech startup — at least not in its early stages. Several factors have turned digital technology into a commodity for any company operating at a small scale:
- the Internet itself is a technological infrastructure. GPS is another. Did you ever wonder why it was so cheap and convenient to geolocate people? Thank the American military;
- open source software is more robust, always up-to-date and considerably cheaper than proprietary software, thus lowering the technological barrier to entry. Open source models are now entering other technological fields such as hardware and biotech;
- cloud computing provides digital startup with cheap access to powerful computing resources, allowing them to avoid programming their own code. Those resources are easy to mash up with less generic ones. The cost starts at zero, and remains linear when the business starts to take off;
- finally, programming languages are now higher level, which makes them easier to use for less skilled programmers. You can create a website without writing a single line of code. You don’t need 15 years experience in advanced computer programming to create simple applications.
As a result, the very word “technology” has become a misnomer. We continue to talk about “tech startups”, “tech companies” or “the tech industry”, when in fact technology doesn’t play a key role at the earlier stages in a startup’s life. Too many people still figure that the digital economy, because it is supposedly so technological, is about solitary, genius scientists manufacturing flux capacitors in a house full of stuffy inventions, using technologies that no one understands. But there are no Emmett Browns in the digital economy, which is more about 140 characters than about flying cars.
The fact that technology is commoditized doesn’t mean that it’s a non-issue. You still need hacking talent to implement the small amount of technology necessary for launching operations. Most importantly, at some point, a company’s needs go beyond what commoditized technology is capable of. As Jeff Bezos explained in his 2010 letter to Amazon’s shareholders:
“While many of our systems are based on the latest in computer science research, this often hasn’t been sufficient: our architects and engineers have had to advance research in directions that no academic had yet taken. Many of the problems we face have no textbook solutions, and so we — happily — invent new approaches.”
It so happens that most cutting-edge technologies in the digital field were invented and implemented by large digital companies confronted with their limitations while trying to serve hundreds of millions of users: MapReduce is a case in point, as are NoSQL databases or the OpenStack technology powering Facebook’s data centers.
Traditional companies
To understand why containing scientific and technological risks made venture capitalism possible in the digital economy, we need to first understand what it takes to scale up a traditional company. The graph below represents a model designed to understand the relative weight of two categories of risk at each stage in a traditional company’s history.
At the beginning, technological risks and marketing risks are of equal weight: it is necessary to both create the product and make it known to prospective investors, employees, and customers. That is the theory. In practice, public spending (in academic research or via research tax credits) covers the cost of some product research and enables the company to get to the development phase faster, thus offsetting a substantial part of the early stage technological risk. It reduces the need for seed capital, which often comes from an individual investor or from an existing company’s free cash flow. Just look at how Joe MacMillan, in Halt & Catch Fire, hacks Cardiff Electric to get hold of its cash flow and create a portable computer.
At the next stage, the weight of scientific and technological risk reduces to zero, and all risk is now on another front: marketing and distribution. That is because the product is completely designed and developed even before it reaches the market (as seen when Joe MacMillan strikes a deal with a retail store chain only after the product is ready). Therefore the stake is now to find a market for the product. In the traditional economy, private investors are reassured by a product they can see (for instance at the COMDEX computer exhibition, again from Halt & Catch Fire) and any market study that somehow concludes that it will be a success. They allocate capital to finance marketing and distribution.
If the initial marketing and distribution efforts have been successful, the company finally reaches product-market fit and a new stage: growing into a dominant position. This is usually the time to do an IPO (remember, in the traditional economy, companies had their IPO quite early in their history). The company goes through different phases, attracting different kind of investors: first, those who make money with capital gains on growth stocks; then, the ones who demand dividends from value stocks. It finally reaches a dominant market position.
When the company is dominant, its scale is large enough to keep competition at a distance. Broadcast advertising, preferably through mass media, makes it possible to keep the customers under control. Moreover, the company can use its operations and influence to elevate a barrier to entry that alleviates competitive pressure and forgo any effort to implement radical innovation. This is how it keeps the scientific and technological risk at a low level: by not letting the market reach the stage where radical innovation would be the only way to survive.
Technological risk may be on the rise here, but it is still below the 50% threshold: it is the risk the company takes to implement efficiency innovation (thus liberating capital and making more money for its shareholders) or renewal innovation (shipping new products that help keep competitors at a distance). At this point, the only danger is a competitor willing to bear a higher level of technological risk and break down the barrier to entry with a radically innovative product, such as was seen with Japanese car manufacturers entering the American markets in the 1970s.
It should be noted that at every point in this (somewhat simplistic) history, few outside investors are willing to cover a high (relative) level of technological risk: the first stage is partially financed by the state; the intermediate stages are marked by a much higher level of risk on the marketing front; at the latest stage, efficiency and renewal (both moderately difficult technological challenges) are financed by free cash-flow.
Digital companies
Now let us turn to the case of digital companies (otherwise known as “tech companies”). There are several major differences with traditional ones as to the relative weight of risks at various stages.
The first difference concerns how low the technological risk is at the earlier stage. The whole phase of developing the technological resources necessary to create a new product is covered by commoditized technological resources such as key infrastructures, open source technology, and cloud computing platforms. As a result, the process begins much later as compared to product development in the traditional economy: founding a company is located at a point where the level of technological risk is already very low (even if a bit of technological innovation is still needed).
A second difference is the fact that product-market fit happens much earlier than in traditional companies. This is because tech entrepreneurs have turned customer development into a science: growth hacking and even crowdfunding are powerful tools that can hasten product-market fit. Instead of conquering the market with mass marketing (delayed product-market fit), it suffices to find early adopters (early product-market fit) and expand from there (= crossing the chasm).
The dominant position arrives earlier, for a very simple reason: winner takes all. In the digital economy, all it takes to be perceived as the dominant player is to have faster growth. And that’s when increasing returns kick in. There are at least four reasons why digital companies tend to grow exponentially.
The economy of scale — As centuries of business have taught us, the bigger you are, the lower your marginal cost: lowering unit costs is what economies of scale are all about. (Yet an economy of scale reaches its limit soon enough. As commodities get scarcer, factories reach their peak capacity, distribution routes get longer, and prospective customers become more difficult to convert, scale ceases to be an advantage and turns into a liability. This is why most companies fail to grow beyond a certain market share). Amazon’s falling cost of sales (see graphic above) illustrates economy of scale at work in a traditional industry such as retail.
Network effects — Most tech businesses connect their users with one another, enabling communication between them either directly (sharing content with our Facebook friends) or indirectly (reading another user’s review on an Amazon’s product page). Such connections turn users into nodes and trigger powerful network effects. When these are at work, the value created for each user increases exponentially as the number of users gets higher. The more a business grows, the easier and cheaper it is to acquire new users. Additionally, the more users an application has, the easier it is to retain current users. The result is more lock-in and thus a lower unit cost per user.
Data — The more a business grows, the more data it can collect from its various stakeholders, especially from its customers. This data can be put back into the company supply chain to train algorithms that constantly grow in terms of accuracy and processing speed. In other words, the bigger you are, the more data you collect, and the more you fine-tune your operations through ever-improving machine learning. Machine learning has thus become a central feature of the scalability at the core of tech companies’ business models.
Virality — As Sangeet Paul Choudary wrote in a useful reminder, virality is not the same thing as network effects. Network effects exist when additional users increase the value of the product, thus helping to attract even more users. Virality is more about users taking charge of marketing and selling the product. For instance, there are network effects in the Dropbox business model, but their main scalability feature is the virality that comes with existing users obtaining extra storage when inviting new users to use the product.
Increasing returns considerably lower the risks of marketing and distribution: you don’t need to finance marginal growth like in the traditional economy. Increasing returns spare companies part of the marketing effort, but they ignite two new kinds of technological risks: the infrastructure has to bear up under a rapidly scaling business, and some technology has to be developed to nurture increasing returns, especially machine-learning algorithms. Hence there is a higher level of technological risk at later stages than in the traditional economy.
Finally, as technological risk gets higher, dominant positions are more fragile. Increasing returns tend to protect you, but i) they peak when you’ve conquered the whole market (see this enlightening McKinsey paper), and ii) as your customers can leave in the blink of an eye, you have to improve performance, create new features, launch new products and constantly improve the experience to retain your users: this is the Red Queen Hypothesis, another reason why technology is crucial for dominant digital companies (but, again, not so much at earlier stages).
There’s obviously a correlation between very high competitiveness in the digital economy and the small amount of technological risk that is taken at the earlier stages. Marketing risks are very high because potential customers are so empowered and so difficult to grab, as opposed to what happened in the traditional economy. Hence entrepreneurs have to minimize technological risk and utilize ready-to-use technology provided by the Internet or open source software. Conversely, the technological barrier to entry is so low that it frees up capital to take even more risks on the marketing front: hence the competitive pressure that exists on every digital market, no matter how dominant a firm can be at any time. Those are the reasons why software eats the world.
A traditional company would cope with that pressure by erecting (or contributing to) a barrier to entry. Alas, in the digital economy, it is not that easy. Or, to be more accurate, there is an arbitrage: erecting a barrier to entry (for instance, a tangible infrastructure) comes at the expense of maximizing your increasing returns (see issue #001 of TheFamily Papers). Amazon imposes a higher barrier to entry because of its tangible infrastructure, but it also has less increasing returns than its intangible counterparts such as Google or Facebook. Netflix also imposes a barrier to entry because it produces original content, but then again it’s operating on a market where increasing returns are hard to sustain, mostly because of negotiations with rights holders and regulations. The tech method of erecting barriers to entry relies on two pillars: the closed ecosystems built for instance by Google (Search, Gmail, Maps, Chrome, YouTube) or Apple (iPhone, App Store, iTunes); and the double-sided platform business model that has been developed by companies such as Google (users / advertisers), Amazon (sellers / buyers), and Uber (drivers / riders).
How do tech companies cope with unprecedented levels of technological risk at a large scale? This is yet another difference with traditional companies. Because barriers to entry are not as high as in the traditional economy, companies can’t rely on efficiency and renewal only: they have to get serious when it comes to radical innovation at later stages. This means they have to attract and retain talent, hence the talent war pointed out by John Doerr:
“Google, Facebook, Amazon, Apple — I think those are the four great horsemen of the Internet. They’re really setting the pace. They’re not limited by market. They’re limited by their ability to execute — to get great, smart people.”
And because it is so hard to implement radical innovation inside the company, dominant tech companies have to keep on buying innovative startups: this is why acquisitions are more frequent in the digital world as compared to the traditional one.
Biotech companies
Finally, there are the biotech companies. It’s a long story, but if you’re into the whole brevity thing: a scientist does some research using grants (such as NIH grants in the US); based on this work, the same scientist or another Entrepreneur founds a biotech startup to try and make an effective drug based on that work, and this startup is financed by venture capitalists for the long period of scientific research and technological development. Finally, if the new drug is approved by the authorities and covered by health insurance, everybody lives happily ever after.
As Janeway put it, the situation of biotech companies is very different from that of digital companies, even though both rely on venture capital:
“Because demand is funded by third party payers and is consequently inelastic, a plausible projection of revenue can be projected contingent, of course, on successful clinical trials and approval by the Food and Drug Administration. Thus, a biotech startup is unique: only in this instance is it possible to estimate a fundamental value, the present value of the net cash flows from the investment — if, and it is a huge if — the scientific and regulatory hurdles to market entry are overcome. The fact that investors have repeatedly chosen to bet on that contingency demonstrates, as well, the weight that the risks of marketing bear versus scientific and technological risks: the biotech exception exemplifies the value attached to the minimization of marketing risks in a domain where scientific and technological risks are enormous.”
Biotech is the only other sector outside the digital economy where there exists venture capital. Early stage investors are willing to renounce liquidity and take a higher level of scientific and technological risk because the companies they invest in have no risk on the marketing and distribution front (because a drug that works and that is paid for by social insurance is very easy to market, since sick and dying people will want the drug no matter what). You don’t have a business if you don’t have the drug. Conversely, if you have the drug, you can make a fortune. Hence the importance, as this Index Ventures blog post puts it, of focusing on one thing (and one risk) only: developing a single molecule.
Key takeaway
Venture capitalists exist in two conditions: i) when there’s a financing need for only one category of risk (either marketing in the digital economy or technology in biotech) and ii) when any success is so huge it pays off the amount of risk that has been taken in the previous phase at a portfolio’s scale (with potential success being huge due to increasing returns in the digital economy and third-party payment in biotech).
You can buy Bill’s book on Amazon here.
(This is an issue of The Family Papers series, which is published in English on a regular basis. It covers various areas such as entrepreneurship, strategy, finance, and policy, and is authored by The Family’s directors and team members as well as occasional guest writers. Thanks to Oussama Ammar, Blake Armstrong, Kyle Hall, Laetitia Vitaud, and Antoine Zins for reading drafts of this and contributing suggestions.)
