Tolstoy

Intuition and Conviction

The Everything Store, the story of Amazon by Brad Stone, has a number of interesting stories that contain great insights, one in particular about Bezos’s time at the hedge fund D. E. Shaw.

David Shaw saw the opportunity in the internet early and tapped Bezos to help him investigate. (“Intrigued by Shaw’s conviction about the inevitable importance of the internet, Bezos started researching its growth.“)  

It was only then that Bezos learned from the February 1994 issue of Matrix News, a monthly newsletter with facts and analysis about the internet, that from January 1993 to January 1994, essentially the first year of the internet, the number of bytes transmitted over the internet had increased by a factor of 2,057. Another fact was that the number of packets had increased by a factor of 2,560.

Bezos summarized the two facts to say that the internet had grown by a factor of about 2,300 in its first year. 

Shaw and Bezos investigated four areas that led to a number of interesting opportunities:

• Email. They created a free, advertising-supported email system called Juno, which went public in 1999 and merged with rival NetZero.

• Online trading. Shaw created FarSight Financial Services, an early E-Trade, in 1995 and sold it to Merrill Lynch. 

• The Everything Store. They also discussed e-commerce, the idea of "an Internet company that served as the intermediary between customers and manufacturers and sold nearly every type of product, all over the world.”

Bezos dived into “the everything store” idea and concluded that such scope would be impractical at first. He listed twenty categories, including software, office supplies, apparel, and music, and concluded that books were the ideal starting point. It was then that Bezos decided to leave D. E. Shaw to pursue the idea.

What I find fascinating about this story is that it’s actually not what common lore about Amazon’s founding leads you to believe. That legend says that Bezos was led down the Amazon path when he saw the 2,300 times growth.

In fact, it was David Shaw that saw the opportunity first. It was conviction first, research and facts later. Shaw saw the opportunity because of his technology orientation and his framework for D. E. Shaw:

While the rest of Wall Street saw D. E. Shaw as a highly secretive hedge fund, the firm viewed itself somewhat differently. In David’s estimation, the company wasn’t really a hedge fund but a versatile technology laboratory full of innovators and talented engineers who could apply computer science to a variety of different problems. Investing was only the first domain where it would apply its skills. 

Framed differently, others had access to the same data in Matrix News that Bezos saw. It was those facts and analysis overlaid on the framework and mindset from Shaw that compelled the idea.

This echoes what I’ve seen elsewhere in early stage companies: conviction emerging from experience and intuition matter more than facts and analysis. In fact, almost by definition with early stage opportunities, the facts and analysis won’t justify the opportunity. 

Capabilities

Earlier this year, there was a bit of an uproar when it emerged that Chamath Palihapitiya had sold his 10 percent stake in Tinder to IAC for $500 million. 

(Chamath was a key person at Facebook, leading its growth team. He made a bit of cash and became an investor, starting the venture capital firm Social + Capital. And Tinder, of course, is a popular dating app, incubated at the internet company, IAC.)

Well, it turned out the reports weren’t true, and while the exact figure wasn’t known, it’s clear the stake had been sold and, while far lower, for a pretty decent sum nonetheless: later reports put his stake at 11 percent and the sale price at $55 million.

As I read more, I was fascinated, and I’m writing it down now, almost four months after the fact, because I find a lot to admire in what the previous moves imply about Chamath’s thinking.

This is the story:

• Tinder was jointly developed by IAC and a Toronto-based mobile development firm called Xtreme Labs. IAC retained control and the lion’s share of equity, with Xtreme Labs retaining, it seems, 11 percent. The app was developed inside a joint venture between the two called Hatch Labs, which was shuttered in late 2013. 

• Chamath bought a majority stake in Xtreme Labs in late 2012 for $20 million of his own money (i.e., not via his venture firm). The co-founder and CEO of Xtreme Labs, Amar Verma, and Chamath had known each since college, and Chamath had worked with Xtreme on some projects at Facebook. 

Palihapitiya told me the deal makes sense in light of the current scarcity of good mobile developers. It will be worth it to him to be able to use Xtreme’s spare time to help with Social+Capital projects, and to spin out interesting start-ups. And Xtreme is now working on open-source frameworks that will bring its native app expertise to a broader audience. (Source: All Things Digital)

• The structure was $6 million up front and $20 million (total? in addition?) over the next three years.

• The most recent comment in the All Things Digital piece, which about captures the general confusion around the purchase (and that from the small number that cared at all), read: “This is the worst thing I’ve seen an investor do. Are you serious? This is a development shop with low margins. I know this team, and I know this space incredibly well. Just pull out of this deal ASAP or reduce your stake for Jesus sakes. Wow. I had respect for Chamath once. Horrible." 

• In late 2013, Xtreme Labs was sold to Pivotal Labs (EMC) for $65 million cash plus incentive compensation to the staff of 300 or so. But Chamath kept the equity stake in Tinder for himself as part of the deal. Exactly what Chamath earned on the sale is a function of how much of the total $20 million committed he ended up investing and what exactly "majority stake” means, but regardless it’s a decent multiple. Let’s say it was $20 million for 80 percent. The $65 million sale would have netted him $52 million for a 2.6 return with a holding period of about a year. 

• Then, about six months later, Chamath sold the stake in Tinder for $55 million if we believe the reports. The $52 million from Xtreme plus $55 million for Tinder yields $107 million for a 5.4x return in about a year and half. Not bad. 

This is what I think is noteworthy:

• Forget the economic return, which alone is interesting. 

• When everyone else, including Chamath himself via his venture firm, is investing in applications, Chamath bought a development firm. 

• My guess is that, having looked at a large sample of companies and given his own experience, he did the qualitative math on number of possible opportunities and the scarcity of good development teams. He said to himself that there’s a scarcity of good developers. I’d word it differently. There’s a scarcity of good development engines, groups that can work together to put out good product.

• A slightly different lens is that there even fewer good development engines with great optionality. Yes, good teams come together and start companies. The thesis is defined and the direction largely set. People do pivot, and often there is a lot of optionality. But there aren’t many truly experimental development engines. Chamath saw Xtreme Labs as a way to learn and experiment. And that’s fascinating. He ended up selling and netting a great outcome, but there’s an argument that he may have sold too soon.

• What I like most about this story–and why I wrote it down to make sure I capture the insights–is the unconventional, first-principles thinking. He was willing to follow the logic of his analysis all the way into core development talent. He is investing in teams and companies via his venture firms, but he was also building core capabilities and learning via the investment in Xtreme Labs. 

Invent the future

“We are still the masters of our fate. Rational thinking, even assisted by any conceivable electronic computors, cannot predict the future. All it can do is to map out the probability space as it appears at the present and which will be different tomorrow when one of the infinity of possible states will have materialized. Technological and social inventions are broadening this probability space all the time; it is now incomparably larger than it was before the industrial revolution—for good or for evil.”

“The future cannot be predicted, but futures can be invented. It was man’s ability to invent which has made human society what it is. The mental processes of inventions are still mysterious. They are rational but not logical, that is to say, not deductive.”

- Alan Kay, Inventing the Future

Be Guided by Beauty

Jim Simons in a nutshell:

Dr. Simons received his doctorate at 23; advanced code breaking for the National Security Agency at 26; led a university math department at 30; won geometry’s top prize at 37; founded Renaissance Technologies, one of the world’s most successful hedge funds, at 44; and began setting up charitable foundations at 56. (Source: The New York Times)

In the video below, he shares some guiding principles with MIT students. I liked one piece of wisdom in particular:

Be guided by beauty. Everything I’ve done has had an aesthetic component to me. Building a company trading bonds…what’s aesthetic? If you’re the first one to do it right, it’s a terrific feeling and a beautiful thing to do something right, like solving a math problem.

My other notes and takeaways from the talk:

• Be first: do something no one else is doing. 

• Powerful insights on data everyone else has = success. Powerful insights on data no one else has = tremendous success. 

“In those days, we sent people down to the NY Federal Reserve to copy histories of interest rate numbers. They didn’t exist in the ‘70s. You couldn’t buy data, and there certainly wasn’t online delivery. To build the original models, you had to collect a lot of data by hand, which we did.“

The main fund, The Medallion Fund, was remarkably successful: Started in 1988. In 1993, closed to new outside investors. In 1992, they started buying out outsiders. By 2005, there were no outsiders. 

"What’s the secret to our success?”

• People. “We start with great scientists. We start with first class people that have done first class work, or that we have reason to believe will do first class work. Because I was there at the beginning with a few other people that were pretty good at math and science, we had pretty good taste.”

• Infrastructure. “We provide people with a great infrastructure. It’s easier to get to work here than anywhere else.”

• Open environment. “The most important thing we do is have an open atmosphere. My belief is that the best way to conduct research on a broad scale is to make sure as much as possible that everyone knows what everybody else is doing. (At least as quickly as possible. Sometimes you want to keep an idea to yourself for a bit so you don’t look like an idiot.) The sooner the better, start talking to other people about what you’re doing. Because that’s what will stimulate things the fastest. No compartmentalization. Everybody meets once a week. Any new idea gets brought up, discussed, vetted, and hopefully put into production. It’s an open atmosphere.”

• Alignment to firm success. “And people get paid on the overall profits, not on their own work. Everyone has an interest in everyone else’s success.”

“Those policies—no one of which seems remarkable—turn out to be a pretty winning combination: great people, great infrastructure, open environment, and try to get everyone compensated roughly based on overall performance.”

My guiding principles:

• Different. “Do something new. I love to do something new. I don’t like to run with the pack. For one thing, I’m not such a fast runner. If you’re one of n people working on the same problem in different places, I know if it were me I’d be last. I’m not going to win that race. But if you can think of a new problem or a new way of doing something, that other people aren’t all working on at the same time, maybe that would give you a chance." 

• People. "Collaborate with the best people you possibly can. When you see a person, or get to know a person, that seems like a great guy or a great gal to work with at something, try to find a way to do it. Because that gives you some reach and some scope. And it’s also fun to work with terrific people.”

• Beauty. “Be guided by beauty. Pretty much everything I’ve done has had an aesthetic component, at least to me. Now you might think, building a company trading bonds—what’s so aesthetic about that? What’s aesthetic about it is doing it right. Getting the right kind of people, approaching the problem, and doing if right. If you feel you’re the first one to do it right—and I think we were—that’s a terrific feeling. It’s a beautiful thing to do something right. It’s also a beautiful thing to solve a mathematics problem or create some mathematics that people hadn’t thought of before. 

• Persistence. "Don’t give up. At least, try not to give up. Sometimes it’s appropriate to be at something, trying to do something, for a hell of a long time.”

• Luck. “Hope for some good luck." 

Judging people

“If one wants to know any man well, one must consider him gradually and carefully, so as not to fall into error and prejudice, which are very difficult to correct and smooth out later.”

- Fyodor Dostoevsky, Crime and Punishment

Remember the Obvious

I read this today by Charlie Munger, and it echoed something I’d been thinking about:

[We] to try more to profit from always remembering the obvious than from grasping the esoteric. … It is remarkable how much long-term advantage people like us have gotten by trying to be consistently not stupid, instead of trying to be very intelligent. There must be some wisdom in the folk saying, `It’s the strong swimmers who drown.’

Business Education

I’m slowly making my way through Clay Christensen’s latest piece in HBR, “The Capitalist Dilemma,” which seeks to understand why “despite historically low interest rates, corporations are sitting on massive amounts of cash and failing to invest in innovations that might foster growth.”

Christensen, along with co-author Derek van Bevers, lays out a nuanced argument, detailing three different types of innovations–performance-improving, efficiency, and market-creating–and argues that various incentives have combined to drive “companies [to] invest primarily in efficiency innovations, which eliminate jobs, rather than market-creating innovations, which generate them.”

Christensen and van Bevers explore the reasons driving this shift and also lay out four proposed solutions. One called “Rebalancing Business Schools” caught my eye because it echoes something I’ve been thinking about and am starting to believe is a significant shortcoming in how we think about businesses.

From that section:

Much as it pains us to say it, a lot of the blame for the capitalist’s dilemma rests with our great schools of business, including our own. In mapping the terrain of business and management, we have routinely separated disciplines that can only properly be understood in terms of their interactions with one another, and we’ve advanced success metrics that are at best superficial and at worst harmful.

Finance is taught independently in most business schools. Strategy is taught independently, too—as if strategy could be conceived and implemented without finance. The reality is that finance will eat strategy for breakfast any day—financial logic will overwhelm strategic imperatives—unless we can develop approaches and models that allow each discipline to bring its best attributes to cooperative investment decision making. As long as we continue this siloed approach to the MBA curriculum and experience, our leading business schools run the risk of falling farther and farther behind the needs of sectors our graduates aspire to lead.

The intricate workings of the resource allocation process often are not studied at all in business schools. As a result, MBAs graduate with little sense of how decisions in one part of the enterprise relate to or reflect priorities in other parts. One of our alumni noted, “The only way we learned what projects to invest in was in FIN I [the introductory finance course at HBS].” A whole host of questions goes unasked—and unanswered: How do I identify conditions that signal opportunity for long-term, growth-creating investment? What proxies for estimated future cash flows can I use in evaluating an investment that is pointed toward a new market? How do we identify and build innovations that will help noncustomers perform jobs they need to get done? When are the traditional metrics of IRR and NPV most appropriate, and when are they likely to lead us astray? Since the functions of the enterprise are interdependent, we should mirror this in our teaching.

And lest one think that this is an academic question, removed from reality, Charlie Munger, Vice Chairman of Berkshire Hathaway, articulated a similar idea at the 2011 annual meeting of Berkshire Hathway:

Costco of course is a business that became the best in the world in its category. And it did it with an extreme meritocracy, and an extreme ethical duty—self-imposed to take all its cost advantages as fast as it could accumulate them and pass them on to the customers. And of course they’ve created ferocious customer loyalty. It’s been a wonderful business to watch—and of course strange things happen when you do that and when you do that long enough. Costco has one store in Korea that will do over $400 million in sales this year. These are figures that can’t exist in retail, but of course they do. So that’s an example of somebody having the right managerial system, the right personnel solution, the right ethics, the right diligence, etcetera, etcetera. And that is quite rare. If once or twice in your lifetime you’re associated with such a business you’re a very lucky person.

The more normal business is a business like, say, General Motors, which became the most successful business of its kind in the world and wiped out its common shareholders… what, last year? That is a very interesting story—and if I were teaching business school I would have Value-Line-type figures that took me through the entire history of General Motors and I would try to relate the changes in the graph and data to what happened in the business. To some extent, they faced a really difficult problem—heavily unionized business, combined with great success, and very tough competitors that came up from Asia and elsewhere in Europe. That is a real problem which of course… to prevent wealth from killing people—your success turning into a disadvantage—is a big problem in business.

And so there are all these wonderful lessons in those graphs. I don’t know why people don’t do it. The graphs don’t even exist that I would use to teach. I can’t imagine anybody being dumb enough not to have the kind of graphs I yearn for. [Laughter] But so far as I know there’s no business school in the country that’s yearning for these graphs. Partly the reason they don’t want it is if you taught a history of business this way, you’d be trampling on the territories of all the professors and sub-disciplines—you’d be stealing some of their best cases. And in bureaucracies, even academic bureaucracies, people protect their own turf. And of course a lot of that happened at General Motors. [Applause]

I really think the world … that’s the way it should be taught. Harvard Business School once taught it much that way—and they stopped. And I’d like to make a case study as to why they stopped. [Laughter] I think I can successfully guess. It’s that the course of history of business trampled on the territory of barons of other disciplines like the baron of marketing, the baron of finance, the baron of whatever.

IBM is an interesting case. There’s just one after another that are just utterly fascinating. I don’t think they’re properly taught at all because nobody wants to do the full sweep.

(Source)

Arguments

I read the following advice in Daniel Dennett’s book Intuition Pumps and Other Tools for Thinking and have found it very useful:

How to compose a successful critical commentary:

• You should attempt to re-express your target’s position so clearly, vividly, and fairly that your target says, “Thanks, I wish I’d thought of putting it that way.

• You should list any points of agreement (especially if they are not matters of general or widespread agreement).

• You should mention anything you have learned from your target.

• Only then are you permitted to say so much as a word of rebuttal or criticism.

The goal is to get yourself and the other party into a frame of mind where each is open to the other’s viewpoint. 

Full stack startups

I’m excited to see the increased discussion around “full stack” startups–startups that take on the entire challenge of offering a service, rather than just offering a layer of technology that other service providers use. 

Think Tesla and Uber.

In Tesla’s case, rather than sell technology to incumbent automobile manufacturers, Elon Musk took on the thornier challenge of creating an automobile company built from the ground up on new technology. The challenges were far greater, but he didn’t have to deal with incredibly long sales cycles, revenue concentration, or customizing the technology to fit existing, and often outdated, technologies and practices, among other benefits. 

By going full stack, he was also able to address many of the other inconveniences in the auto purchasing and ownership experience, namely the dealership-based purchasing and servicing process, which he detailed in a blog post related to New Jersey’s decision to ban Tesla sales in company-owned showrooms. 

I first heard the idea of “full stack” startups from Glenn Kelman, CEO of the real estate company Redfin, and I wrote about it in an earlier post. His mental model for the idea was “change the game”. In other words, rather than walk into the existing game, with proverbial hat in hand, offering to fit a disruptive idea into the confines of the existing industry, one should consider changing the game entirely. Abstract the discussion one level higher to ask: What is the fundamental service being provided by the industry and how might it be improved dramatically with new technology and by being re-thought completely?

Chris Dixon of Andreessen Horowitz wrote about the idea earlier this month, coining (I believe) the phrase “full stack startups”.

And then today I read this incredible series of tweets by Balaji Srinivasan, also at Andreessen Horowitz, which prompted me to write this post to collect my thoughts:

Veeva is Different

I read once that achieving 10x outperformance is easier than achieving 2x outperformance. The reason is that 2x requires you to do better than the alternative on already established dimensions, whereas to generate 10x, you need to re-create the game.

This idea led me to recall something similar I heard from Mike Maples. He pointed out that Microsoft raised $1 million of capital and that eBay raised $5 million (video). These huge, game-changing companies didn’t raise much capital. They created products for which, to quote Maples, “the world had a high give-a-shit factor”. The world needed their products badly, letting them fail, iterate, and improve, all the while begging them to continue to do business with them. And when these companies do get it right, their product gets pulled into the market. 

Veeva Systems, whose S-1 became available last week, appears to be that sort of company.

After initially glancing through it’s S-1, I tweeted how remarkable it was that it raised only $7 million and was on track to clear $200 million of revenue, having doubled revenue consistently year after year.

Since sales and marketing tend to be the most significant use of capital in SaaS companies, I focused on that, and it appeared they performed far better in that dimension. In my tweet, I estimated they were about two to three times better than other SaaS companies.

Turns out I was wrong. Veeva Systems was 3.6 times better than the average SaaS company. For every $1 they spent on sales and marketing, they generated $3.05 in annual recurring revenue. They made three times their money in just the first year. The average for all other currently public SaaS companies in their pre-IPO years was $0.85. These companies lost money in the first year, making it up in later years due to the subscription nature of the revenue.

Digging one layer deeper, what’s remarkable is how far beyond the typical range Veeva Metrics is in this regard, standing out even from the second best-performing company, Demandware.

How does this happen?

I don’t have any proprietary insight today into how this came about, but there are hints in the S-1:

We sell our solutions through our direct sales organization and had sales representatives in 13 countries as of July 31, 2013. Our sales force is managed regionally by general managers in North America, Europe and Asia Pacific who are responsible for all sales, professional services and customer success in each of their geographies. We believe this provides for an integrated view of the customer relationship as well as higher levels of local and regional focus on our customers.

Life sciences companies are typically organized by the major functions of research and development for the creation and development of new solutions, and commercial, for the sales and marketing of those solutions once they are approved for use. In large life sciences companies, research and development and commercial business lines may also have separate technology and business decision makers. Accordingly, we market and sell our solutions to align with the distinct characteristics of the research and development buyer and the commercial buyer. Within each region, we have research and development and commercial sales teams. Each of these teams is further divided to sell to the largest global pharmaceutical companies and to smaller life sciences companies.

We believe the combination of our industry­ focus and commitment to customer success provides strategic advantage and allows us to more efficiently market and sell our solutions as compared to horizontal cloud­based companies. Our awareness, demand generation and sales cultivation programs are focused and designed to be cost efficient because we target only the life sciences industry buyers. We believe that we further benefit from word­ of­ mouth marketing as customers endorse our solutions to their industry peers. This allows us to focus our sales and marketing efforts without the need for a larger number of sales executives.

(Emphasis mine.)

In short, the company is sub-divided into regions with general managers seemingly running mini organizations targeting the companies in their regions. One person owns sales, professional services, and customer success. It’s a powerful alignment.

Combine that with the industry focus, where dramatic success with a few clients will spread through word-of-mouth, and it isn’t hard to imagine inbound requests just started to flood into the organization.

Professional services drove customer success

Professional services is something else that jumps out in the S-1. Veeva spent an incredible amount on professional services. Professional services was 50 percent of its revenue in early years, declining to 30 percent more recently. It was probably even higher in the years not detailed in the S-1. This was really low margin, difficult-to-manage and -scale business. They spent $50 million on professional services between 2009 and 2013.

Customer success drove incredible renewal and upsell rates

Professional services drove customer success and had a powerful impact on future revenue from those customers. This is demonstrated in its churn metrics, what Veeva Systems measures as its renewal rates.

At the end of each year, Veeva Systems would take the total annualized revenue from the full list of customers at the end of the prior year and divide by the annualized revenue from those customers at the end of that prior year.

If some customers left and others spent about the same, this metric would be below 100 percent.

If, on the other hand, customers were not only sticking around but buying more from Veeva Systems, renewal rates would be above 100 percent.

For 2010, 2011, and 2012, renewal rates were 192, 159, and 187 percent, respectively. Specifically, on average across all customers that were customers at the end of 2011, for every $100 of subscription revenue, they had increased that to $187 by the end of 2012.

Having seen many such metrics for SaaS companies, I know these are amazing, particularly at the revenue levels to which they apply. Such an organic uplift from the established revenue base makes for very efficient sales and marketing because that revenue required almost no additional work.

Update: Subsequent to writing this post, David Skok and Pacific Crest put together a great survey of private SaaS company metrics, one of which has renewal rates as measured by Veeva here. It further illustrates how unusually high Veeva’s net renewal rates are compared the median 110 percent below:

Further, I suspect this is skewed higher by companies with smaller revenue bases. For other metrics, David Skok and Pacific Crest separated those, finding a skew towards the high end, driven by the lower revenue companies. I’m confident that Veeva’s performance here is mind-blowing for any company with more than, say, $10 million in annual recurring revenue. 

Lessons:

• Make products people can’t live without.

• Make someone responsible for the entire customer cycle, from sales to implementation to success.

• Focus on a small part of the world, where word of mouth can amplify success.

__

Notes:

1. The exact calculation I used to assess sales and marketing effectiveness is below. I calculated this for every quarter highlighted in the S-1 of every SaaS company that went public, whether currently public or acquired. I then averaged the quarterly metrics to come up with one metric per company to estimate pre-IPO sales and marketing effectiveness.

2. I’d be remiss if I didn’t mention the folks from whom I learned this. They include Rory O’ Driscoll at Scale Venture Partners in a series of insightful posts on what he calls the Magic Number here and here, Phillippe Botteri when he was with Bessemer Venture Partners writing about CAC ratio here, and David Skok at Matrix Partners in a tremendous series on SaaS metrics here.

3. One thing worth commenting on is that Veeva Systems is built on top of Salesforce’s Force.com platform. This isn’t free. Salesforce gets a cut. I don’t know the exact arrangement, and I didn’t see it outlined specifically in the S-1. But it’s embedded in the gross margins, which, while lower than other SaaS companies, aren’t dramatically lower. Veeva’s 2012 gross margin on subscription revenue was about 75 percent. SaaS companies will ideally have 90 percent subscription gross margins. So this is significant. But it’s not 50 percent, which would be eye opening. So Botteri’s CAC ratio approach above, which uses GMs instead of revenue, would be one way to refine this. Back of the envelope adjustments indicate, however, that you’d still find Veeva Systems holding a significant lead in the areas I explored so I don’t think this changes my conclusions.