As Software Eats The World

Tech people like me can sometimes come across as presumptuous/arrogant regarding the disruption of other peoples’ industries. It is possible this is an understatement. I am sure that some of my Twitter friends will expand on this for me.

From this side of the aisle, though, it’s less smugness, It’s more the result of hard experience and learning from our own lives and careers. In tech, our *own* businesses are disrupted by technology changes and new competitive entrants at whiplash-inducing rates. It’s shocking how quickly you can go from the hot disruptive upstart to the stodgy disrupted incumbent in tech frequently within 5 years.

I’ve probably been on the receiving end of disruption 30 times in the last 20 years, almost as many times as I’ve been on the giving end. Now, on the one hand, you might say, “How can people live like that? What’s wrong with a little stability?” But, what we see is: Frequent disruption is the handmaiden of rapid progress and it’s a blast to create and work amid rapid progress.

It’s not just the rapid progress of tech. It’s also the rapid growth of companies, and even better, rapid development of *people* and their talents. It’s hard to stay in tech for any period of time and not get good at rapid adaptation, skill acquisition, and new product creation. As software eats the world, the same disruption dynamics always present in tech are now applying to many more industries, fields, and professions. Rather than superiority/contempt, what a lot of us feel is deep sympathy/understanding, even if that’s not always how it comes across! Now we all have the opportunity to learn together, to make many parts of industry/life more innovative/dynamic, which is better for everyone.

[tweet  align=”center”] [tweet  align=”center”] [tweet  align=”center”] [tweet  align=”center”] [tweet  align=”center”] [tweet  align=”center”] [tweet  align=”center”]

Source Tweets: 1,2,3,4,5,6,7,8,9,10,11,12,13

A Few Thoughts On Timing and Staging Of Capital Into Modern Tech Startups

A few thoughts on timing and staging of capital into modern tech startups. Start with the fact that 2003-20011 seed rounds were ~$500K-1M. As points out, now you see more startups raising $2-3-4M or even more in “seed” financing, often in multiple tranches. So then, think about the startup that’s raised $3-4M or even $5-6M in “seed” funding that goes to raise a “Series A” from VC firms.

Venture capitalist looks back across the table: “You’re not raising a Series A, you’re raising a Series B. You already raised your Series A [in seed $]”. This can take the startup by surprise, because it really affects how VCs think about progress and milestones, is key to raising new round. VC’s assume Series A is to build, product and get first beta customers; Series B is to build the business around the product and get to revenue.

So a startup that raised as much cash as a Series A in seed funds, but hasn’t achieved actual Series A milestones, can be in real trouble. VC says: “You said you’re raising A but you’re actually raising B, and you haven’t accomplished enough to merit a B. Thank you, but pass.” So the risk of calling $3-4-5-6M “seed” raises “seed” is that the founder can fool himself/herself heading into the first real VC raise.

The rise of the “New Series A” by firms like A Capital is intended to address this issue head on. In effect, a $3-5M seed round or a $3-5M “New Series A” is a recreation of the original conception of an A round from historical VC. Takeaway for founders? Think very hard about timing and staging of capital versus progress and milestones. This matters a lot for raising A/B/C.

Source Tweets: 1,2,3,4,5,6,7,8,9,10,11,12

BusinessWeek Illustrates Who’s Coming To Silicon Valley

Have to post this again because it’s just too good not to. A graphic illustration from BusinessWeek showing who’s coming to Silicon Valley.

Image originally published on <a title="" href="">BusinessWeek</a>

Image originally published on BusinessWeek

One third of Silicon Valley startups are founded by Indian-Americans. As of 2010, Asian-Americans are the majority of Silicon Valley tech workforce: 50% vs 40% for Caucasians. Possibly eye-opening sources of talent in quantity: Japan, Middle East, Vietnam, France, Pacific Islands, Africa, Caribbean.

Silicon Valley is a powerful successful example of the “melting pot” theory: ignore origin and ethnicity, come together to do big things.

It’s why we must continue to push to expand access/openness/inclusion to all origins/ethnicities/genders/religions: there are huge opportunities ahead.

Source Tweets: 1,2,3,4,5,6

The Productivity Puzzle Of Robots Eating All The Jobs (Or Not)

Progressive and smart economist Jared Bernstein on the productivity puzzle of robots eating all the jobs (or not):

Productivity growth was up 1% last year and has averaged 0.8% since 2011, a smooth trend through the numbers. The trend suggests that the pace of productivity growth has decelerated since the first half of the 2000s which begs an important question.


I keep hearing about ‘the end of work‘ based on the assumption that the pace of labor-saving technology such as robots and artificial intelligence has accelerated. Maybe it has. There’s lots of good anecdotes to that effect, most recently that geeky-looking Google self-driving car.


But the robots-are-coming advocates need to explain why a phenomenon that should be associated with accelerating productivity is allegedly occurring over a fairly protracted period where the [productivity] trend in output per hour is going the other way.


A shave with Occam’s razor [explanation] would be weak demand and its corollary, weak capital investment [nothing to do with robots]. Until someone can convince me what’s wrong with the above argument, I don’t want to hear that automation is precluding full employment.


My own take: We’re still coming out of a severe macroeconomic down cycle, credit crisis, deleveraging, liquidity trap. The prevailing pessimistic economic theories (death of innovation, robots eating all the jobs, crisis of inequality) will fade with recovery.

For bonus points, identify the other tech-driven economic force that could explain low productivity at a time of great tech advancement. My nomination: Tech-driven price deflation; lowers prices, reduces measured GDP and productivity, while boosting consumer welfare.

Source Tweets: 1,2,3,4,5,6,7,8,9,10,11,12

The Short Leash Of Highly Paid Executives

When a top executive gets fired and the public reasons are unclear or confusing, the most common explanation is that he or she lost support of his or her direct reports. It is generally impossible for any board or CEO to leave an executive in place when that executive’s direct reports are collectively revolting. What’s interesting: This is true in practice almost regardless of what you think of the reasons for the revolt.

If the reasons for the staff revolt are valid, then clearly a mistake has been made and must be fixed, the executive has to go. If reasons for the revolt are not valid, then there’s an even bigger cultural problem at that point, with an even broader cleanup required. But even if reasons for the revolt are not valid, the executive under fire is still right in the middle of it, has lost confidence from their troops, still goes.

Executives get paid the big bucks because their decisions impact lives and careers of 100s/K’s/10K’s of employees. Thus the short leash is justified. Another theory for high executive compensation is precisely to make it easier to take dramatic action when needed. It’s an implicit safety net. None of the preceding is intended to diagnose any specific situation, just the general pattern. I’m also not excusing any kind of bad behavior.

All of this double-underlines the importance of wise governance, leadership development, management training, and cultural integrity.

[tweet align=”center”] [tweet align=”center”] [tweet align=”center”]

Source Tweets: 1,2,3,4,5,6,7,8,9,10

Theories on Valuing Companies

The conventional view of how to value companies:

(1) Analyze the company + its financials + future cash flows;
(2) Calculate the correct valuation.

What actually happens:

(1) Observe current market valuation;
(2) Construct theory and model to explain that valuation.

In this way, George Soros’s theory of reflexivity is exactly correct. Fundamentals influence prices which influence fundamentals which influence prices which influence fundamentals… ad infinitum.

At the cyclical top, high prices drive creation of theories to explain infinite future glory; negative investors and analysts get fired. At cyclical bottom, low prices drive creation of theories to explain permanent future misery; positive investors and analysts get fired.

Therefore, a boom in theories of how everything’s a bubble and certain to crash is evidence of a cyclical bottom, not a cyclical top. Therefore, Efficient Market Hypothesis is correct if for “all information” you substitute “all information, theories, noise, and bullsh*t”. Since we are social animals, the challenge of actually standing outside of the herd is brutally hard. Pressure to conform is constant/intense.

Further Reading:

Famous paper well worth reading: “The Limits of Arbitrage

Another famous paper well worth reading: “Noise” by the great Fischer Black

Tweets – 1,2,3,4,5,6,7,8,9,10,11


Prevailing Beliefs on Tech Innovation

Prevailing beliefs that I do not share:

  • Tech innovation is dead.
  • Tech innovation is dead except for the part that will kill all the jobs.
  • Tech innovation is dead except for the part that will kill all the jobs and give all the money to the 1%.
  • QE is hopelessly distorting the economy.
  • Hyperinflation is right around the corner.
  • Hyperinflation is right around the corner and interest rates are about to skyrocket.
  • Five billion more people are getting access to the most amazing tools for education, information, creation, and access to markets ever…and yet they will figure out how to do… absolutely nothing with them, and are doomed to lives of spiraling poverty and despair.

Source: tweets – 1,2,3,4,5,6,7,8,9,10


The Smart Person Fallacy In Two Easy Steps

The Smart Person Fallacy in two easy steps. “I’m smart, I can learn about situation X and figure out the way it should work.”

First step: Situation X is likely far more complex with far more moving parts and confused causes and effects than you can imagine. Autodidacts and polymaths are highly prone to this. True experts in a field are often far more skeptical about their own ability to understand. Often suffered by professional writers, journalists, commentators, columnists, analysts, investors and venture capitalists.

Second step: In many complex situations, logic matters far less than other factors — with incentives at the top of the list. So thinking your way to the answer may well be counterproductive or worse. Often suffered by intellectuals, academics, theorists, paternalistic left-wingers and venture capitalists.

Bonus step: For any mandated change to situation X, unintended consequences are likely to dominate the long term effects.

Source Tweets: 1,2,3,4,5,6,7,8

Private Valuations In Tech

The third aspect of the valuation of tech companies often misunderstood is private valuations set by venture capitalists and other private investors. This topic was recently explored by @Jessicalessin in this excellent article.

A private company in which a sophisticated investor has bought a minority stake for $X/share is not actually worth $X * total number of shares.

First, the entire company has not traded hands, just a small slice of it. So we don’t actually know what the whole company is worth. Second, most financing rounds are for preferred shares, which have special rights. Other shares don’t have those rights and are worth less. Smart VCs think about startup shares less as stock than as options — options with limited (1x) downside and unlimited (1,000x+) upside.

A share of preferred startup stock ~= A long-dated out-of-the-money call option, paired with a long-dated contingent put option. The contingent put option is the liquidation preference in preferred stock. It increases the odds of getting cash back in a downside sale of the company. Plus, in some high-valuation late-stage rounds, there are additional downside protections like ratchets, which can be highly valuable and preferred stock brings with it governance rights and access to information not available to normal investors. Those have value too.

So you can’t extrapolate the value of an entire company from a minority sale of preferred stock. It’s better just to focus on cash raised. In my view, there is WAY too much discussion of private valuations in tech. Fuzzy numbers matter way less than real company substance.

The best book to read as a followup to this post is Venture Deals: Be Smarter Than Your Lawyer and Venture Capitalist by Brad Feld.

Source Tweets: 1,2,3,4,5,6,7,8,9,10,11,12,13

Common Fallacies About The Valuation Of Public and Private Technology Companies

First, ask any MBA how to value tech companies, she’ll say “discounted cash flow, just like any other company”.

Problem: For new and rapidly growing tech companies, up to 100% of value is in terminal value 10+ years out, so the discounted cash flow framework collapses. You can run as many discount cash flow spreadsheets as you want and may get nothing that will help you make good tech investment decisions. Related to the fact that tech companies don’t have stable products like soup or brick companies; future cash flows will come from future products. Instead, the smart tech investor thinks about:

  • Future product roadmap/oppurtunity
  • Bottoms-up market size and growth
  • Talent and skill of the team.

Essentially you are valuing things that have not yet happened and the likelihood of the CEO and team being able to make them happen. Finance people find this appalling, but investors who do this well can make a lot of money, but spreadsheet investing is often disastrous. It doesn’t mean cash flow doesn’t matter, in fact the opposite is true: this is the path to find tech companies that will generate tons of future cash.

Corollary: For tech companies, current cash flow is usually useless for forecasting future cash flow, a lagging not leading indicator. This trips up value investors (Prem Watsa!) all the time; tech companies with high cash flows often about to fall off a cliff. Because current cash flows are based on past products not future products and profitability often breeds complacence and bureaucracy.

Always, always, always, the substance is what matters: WHO and WHAT. WHO’s building the products and WHAT products are they building?

Brand will not save you, marketing will not save you, channels will not save you, account control will not save you. It’s the products. Which goes right back to the start: Who are the people, what are the products, and how big is the market? That’s the formula.

Source Tweets: 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15