The structure of the VC industry is changing. This matters not only to entrepreneurs raising capital — but it also impacts the finance industry overall, because companies are staying private longer (fewer IPOs) and public investors (including hedge funds, mutual funds, publicly held corporations) are getting into the VC game, too. So in that sense these changes affect everyone who is in the market.
The changing structure of the VC industry was a huge topic at the recent PreMoney conference — and the focus of a discussion between me and fellow venture capitalist Mark Suster, which you can watch here and here. Mark also just wrote an excellent post discussing the structural changes; I highly recommend you read it. To summarize, some of the key observations include: the rise of small funds (they now account for 67% of all new funds raised in 2014); the concentration of capital among fewer, larger firms (10 firms accounted for 48% of total limited partner dollars raised in 2012, according to the NVCA); the decline in the number of VC firms with $100-500M in assets (this group represents only 18% of funds in 2014) — and the fact that companies are staying private longer (some of my thoughts for why are here).
But the real question is: Why now? What’s really behind the structural transformation happening in venture capital — and by extension, the tech industry? And finally: How do we reconcile the seeming paradox of it being cheaper to start companies today… but at the same time needing to raise more capital?
Much ink has been spilled on the very real decline in the costs of building technology companies. Some estimate as high as 100x decreases in the costs of storage, compute, bandwidth, networking, etc. in just the past decade and half.
With the advent of AWS, DigitalOcean, Rackspace — and cloud-based technologies in general — technology startup costs are significantly lower today. A lot of entrepreneurs can’t even conceive of this now, but there was a giant slurping sound of dollars flowing from VC firms to the leading technology infrastructure companies during the time we were building Loudcloud (2000). Back then, a startup raised $5 million in A round financing (yes, in those days, A rounds were actually $5 million!) and that money traveled from the VC’s bank account to the startup’s bank account directly into the bank accounts of Akamai, BEA, Cisco, EMC, Exodus, Level 3, Oracle, and others. It’s important to note that all of this was up-front capital expenditure.
Today, startups can pay as they go, scaling up their operating costs (vs. CAPEX) as needed, as their business grows. At the same time, advances in developer productivity tools (think Github) and programming languages (like Python) have dramatically increased the efficiency of software development. In many ways, what previously took 100 developers can now take just 10.
In addition, a consumer-facing startup that deploys its applications via the internet or as a mobile app gets customer distribution mostly free, at least initially. While Apple does charge a 30% tax on revenue that flows through the App Store, this is a success-based fee rather than an up-front sunk cost that must be borne regardless of revenue generation.
Meanwhile, the proliferation of platforms such as Google, Facebook, and Twitter as customer acquisition channels makes it that much easier for new companies to experiment — with relatively small amounts of invested capital — on a product to see whether it gets initial market traction. This is not completely free, but it is still a very targeted and a more incremental (vs. up-front) way to acquire initial customers at relatively low cost.
All of this explains the rise of small funds: It simply takes less money to get a company from conception to initial product launch to early customer traction (or not) than it has at any other time in history. But how does this explain the concentration of capital dollars in larger ($100-500 million) VC funds? Doesn’t that contradict the argument about how much cheaper it is to start companies today?
Well, it turns out that, while it is in fact cheaper to get started and enter the market, it also requires more money for the breakout companies to win the market.
End-user markets are much bigger than in previous decades. The internet population (now exceeding 2.5 billion) is 10x what it was in 1999, and rapidly growing around the world. Tablets have overtaken PCs as the compute medium of choice, but the real transformation is in smartphones. Of all the cellphones in use today, only 30% are smartphones. If you believe, as we do, that there will be a wholesale conversion of this remaining 70% to smartphones — which are really broadband-enabled supercomputers that we carry around with us at all times — the opportunity for successful companies to reach ever larger markets is unprecedented.
This is is not just a consumer trend: Enterprise adoption of mobile technologies and the proliferation of SaaS as a software delivery vehicle means more users within companies and government agencies will deploy more applications than ever before. And the security, identity management, and core infrastructure needs to support these applications will grow accordingly.
But the biggest transformation of all is in who can be reached. With potentially 5.9 billion users coming online — largely due to the developing world — we have the ability to reach global markets at a scale never before witnessed. Already more people have more access to mobile telephones than they do toilets.
At the same time, the number of international markets that were once off limits due to geopolitical issues or high costs of entry is significantly smaller. The widespread penetration of mobile devices — and therefore applications — has also leveled the playing field. For example: 80% of Twitter users are international, and Facebook arguably has more international than domestic users if you include its reach through WhatsApp. The direction of innovation is starting to turn, too: Spotify successfully entered the U.S. market after having starting outside the U.S., and Alibaba, Baidu, and Tencent are likely to do so as well over the coming years.
Finally, we are seeing the rise of what my partner Chris Dixon has metaphorically described as the “full-stack” startup, where entrepreneurs want to own more of the full customer experience and are therefore building out expertise in areas beyond just tech. For instance, in a non-full stack world, AirBnB would have been a software company that developed a better customer matching and booking engine and then licensed that to multiple hospitality purveyors who would then own the customer relationship. Instead, the “full-stack” AirBnB owns the customer relationship end-to-end, and thus must build expertise in customer service, global operations, regulatory affairs, and so on. Perhaps more significantly, in a full-stack world, software companies can compete with their physical-world counterparts and build new and unexpected platforms that others can then build on. [More on full-stack startups here, here, and here.]
Ultimately, today’s winners have a chance to be a lot bigger. But winning requires more money for geographic expansion, full-stack support of multiple new disciplines, and product expansion. And these companies have to do all of this while staying private for a much longer period of time; the median for money raised by companies prior to IPO has doubled in the past five years. It’s vastly more efficient for startups that anticipate needing significant capital to seek partners who have the capital and the resources to grow with their needs.
* * *
Much of the above explains why we are seeing the structural changes in the VC industry that Suster and others have described.
But it’s important to note that these structural changes are not unique to the VC industry — this is a natural evolution that has occurred in nearly every other services-based business over the last 50 years. That is, a “death of the middle” occurs over time as service industries bifurcate into a smaller number of large, fully integrated, full-service institutions on one end and a larger number of smaller, niche-oriented institutions (with a focus on stage, industry, or specialized skillset) on the other.
Look no further than investment banks, law firms, accounting firms, advertising agencies, buyout firms, talent agencies, and recruiting firms to see how this phenomenon has played out in mature services-based business. Interestingly, we are witnessing this right now in a non-services business as well: the U.S. retail market. Long-dominant department stores such as J.C. Penney and Sears are giving way to big-box retailers and etailers (Amazon, Best Buy, Target) at the large end complemented by large numbers of specialty, boutique stores at the smaller end. [Though in retail, betting “narrow” can be big as the internet allows companies to better segment and address previously unaddressed markets.]
Despite how much venture capital firms invest in the most disruptive industries, the VC industry has itself been relatively staid from a structural perspective since its founding. With current trends as our compass and history as our guide, we can expect the industry to look a lot different in the next 10 years than it has in the previous 50.
So, what should we make of all of this? For VCs and LPs, venture capital is alive and well and, more significantly, the return potential to private investors generally looks very bright. For entrepreneurs, this is the most exciting time not just to get capital from multiple sources — but to build companies with a long-term game in mind.