Chris Dixon

One thing that was puzzling about the “web 2.0 boom” from 2003-2008 was how irrelevant the East Coast, and particular New York City, was compared to the first dot-com boom.  There were a few big hits – Right Media comes to mind – and a big near miss – Facebook – which started in Boston but moved to the West Coast.

I was mostly checked out of the internet scene in the 90s (in perpetual grad school), but from everything I’ve read and heard, New York City and the East Coast in general was much more competitive with the West Coast.  One interesting supporting data point: Matrix Partners in Boston had the best return of any VC fund in the 90s (an astounding 516% IRR).

I think it’s fairly easy to explain what happened to Boston in the 2000′s.   In the 90′s much of the action was around infrastructure and enterprise software – and Boston (led by MIT) tends to be very infrastructure and enterprise oriented.   I am told Boston is still relevant in biotech and cleantech, and perhaps infrastructure and enterprise IT will have a resurgence, although even those areas seem to now be dominated by the West Coast.

But the question that has puzzled me is:  why did New York City lag behind the West Coast this decade so much more than last decade?  Especially since the internet in the 2000′s has been more than ever about consumers, media, and advertising – traditional New York City strengths?

I think the only explanation is that the finance bubble of 2003-2008 was a giant talent suck on the East Coast.  The people I knew graduating out of top engineering or business programs on the East Cast were all trying to work at hedge funds or big banks or else felt like fish out of water and moved west.   Money was flowing so freely in the finance world that there was no way the risk/reward trade off of startups could compete.  Eventually it just became downright idiosyncratic to be a startup person on the East Coast.  The Larry and Sergey of the East Coast were probably inventing high frequency trading algorithms at Goldman Sachs.

But this is why New York City now seems poised for a technology startup boom. The finance bubble has burst and the industry will hopefully return to its historical norm, about half its bubble size.  The traditional advertising and media businesses are in disarray.  The people who work in them will no doubt find new applications for their talents.

There is also a nice ecosystem developing in New York City.  Union Square Ventures is one of the best VC’s in the country, with early stage investments in companies like Twitter and Etsy (that were followed on by top West Coast VCs at significant markups).   Bessemer is an old firm that has a managed to stay relevant with investments in Yelp, Skype, and LinkedIn among others.  There is also a new wave of scrappy Boston firms spending a lot of time in New York City – specifically Spark, General Catalyst, Flybridge, and Bain Ventures.  First Round Capital out of Philadelphia is extremely active in early stage investing in New York.  There are a bunch of veteran entrepreneurs actively investing in and mentoring seed stage startups.  Google has a big office here and many people seem to be leaving to go start companies.

But most importantly, the engine of the startup economy, young engineers, will be returning to doing something besides shuffling money around.  As Obama said:

…Wall Street will remain a big, important part of our economy, just as it was in the ’70s and the ’80s. It just won’t be half of our economy. And that means that more talent, more resources will be going to other sectors of the economy. And I actually think that’s healthy. We don’t want every single college grad with mathematical aptitude to become a derivatives trader. We want some of them to go into engineering, and we want some of them to be going into computer design…

That’s why I don’t just want to see more college graduates; I also want to specifically see more math and science graduates, I specifically want to see more folks in engineering. I think part of the postbubble economy that I’m describing is one in which we are restoring a balance between making things and providing services…

New York City has many of the same strengths as Silicon Valley – merit-driven capitalism, the embrace of newcomers and particularly immigrants, and a consistent willingness to reinvent itself.   Silicon Valley will always be the mecca of technology, but now that people here are getting back to, as Obama says, making things, New York City has a shot at becoming relevant again in the tech world.

The biggest mistake entrepreneurs make when pitching VCs is to argue that their startup is likely to succeed.  Instead, they should argue that there is a small probability their startup could be a billion dollar or greater exit.  There is a big difference between these arguments – the mean of the return distributions might be the same but what VCs care about is right side tail of the distribution.

Investor sentiment, the old saying goes, is a horse race between fear and greed.  The fear and greed in venture capital is all about investing in or missing out on the next Google.  No VC stays up at night worrying about missing the next startup that’s flipped to Google.  The way you get VCs interested is to convince them there’s a small but non-negligible chance you’ll create a billion dollar (valuation) business.

I’ve learned this lesson first hand on both sides of the table.  One example:  A good friend of mine was starting a company a few years ago.  I was excited about the idea and tried to help him raise venture money.  After the entrepreneur pitched some VC friends of mine, I was surprised when the they came back to me to say they are passing because “it seems like a smallish, ‘lifestyle’ business.”

The entrepreneur had made a very good pitch for why his product was valuable, why he could create a profitable business, that he was very smart and well prepared, and so on.   What he needed but failed to do was leave the VC with the nagging thought that this could be the “the next big thing.” Part of this was because of the entrepreneur’s natural modesty.  Some people don’t have the chutzpah to aggressively assert that their idea is the next big thing, even when, deep down, they truly believe it.  In everyday life, this kind of modesty is a virtue. When pitching VC’s, it is the single worst thing you can do.  (If deep down, you don’t believe your idea will be the next big thing – don’t raise VC money.  Once you raise VC you are committed to going for the billion dollar exit whether you like it or not.)

I don’t know if this obsession with the upside outlier case is a good strategy from the VC’s perspective or not.  Granular VC return data is hard to come by.  I tend to think it is a good strategy – one Google or Facebook (and a lot of other billion dollar exits that aren’t nearly as famous) can make up for a ton of misfires.  And the anecdotal return numbers I’ve heard from VCs suggests it works.   But I don’t really know.

As this article by Alex Wright in the New York Times last week reminded me, when the mainstream press talks about artificial intelligence – machine learning, natural language processing, sentiment analysis, and so on – they talk as if it’s all about algorithmic breakthroughs.  The implication is it’s primarily a matter of developing new equations or techniques in order to build systems that are significantly smarter than the status quo.

What I think this view misses (but I suspect the companies covered in the article understand) is that significant AI breakthroughs come from identifying or creating new sources of data, not inventing new algorithms.

Google’s PageRank was probably the greatest AI-related invention ever brought to market by a startup.  It was one of very few cases where a new system was really an order of magnitude smarter than existing ones.  The Google founders are widely recognized for their algorithmic work.  Their most important insight, however, in my opinion, was to identify a previously untapped and incredibly valuable data source – links – and then build a (brilliant) algorithm to optimally harness that new data source.

Modern AI algorithms are very powerful, but the reality is there are thousands of programmers/researchers who can implement them with about the same level of success.  The Netflix Challenge demonstrated that a massive, world-wide effort only improves on an in-house algorithm by approximately 10%. Studies have shown that naive bayes is as good or better than fancy algorithms in a surprising number of real world cases.  It’s relatively easy to build systems that are right 80% of the time, but very hard to go beyond that.

Algorithms are, as they say in business school, “commoditized.”  The order of magnitude breakthroughs (and companies with real competitive advantages) are going to come from those who identify or create new data sources.

Like a lot of things we think are obvious today, financial options were first invented by a philosopher:

There is the anecdote of Thales the Milesian and his financial device, which involves a principle of universal application, but is attributed to him on account of his reputation for wisdom. He was reproached for his poverty, which was supposed to show that philosophy was of no use. According to the story, he knew by his skill in the stars while it was yet winter that there would be a great harvest of olives in the coming year; so, having a little money, he gave deposits for the use of all the olive-presses in Chios and Miletus, which he hired at a low price because no one bid against him. When the harvest-time came, and many were wanted all at once and of a sudden, he let them out at any rate which he pleased, and made a quantity of money. Thus he showed the world that philosophers can easily be rich if they like, but that their ambition is of another sort.

- Aristotle, Politics, Book 1, Part XI

/* aggregate_options_strike_price = your options strike price per share * number of shares you own
company sale price is 1) if private transaction: amount paid by acquirer plus any funds in startup returned to investors,  2) if IPO = market capitalization.
note: if you assume all financings were 1x preferred, investor preferences == total amount of money the company has raised
to do:  add condition for participating preferred, graph various scenarios

*/

function my_exit_payout(  company_sale_price, your_percent_ownership, your_aggregate_options_strike_price, investor_preferences, investors_ownership_percent)
{

if (investors_ownership_percent * company_sale_price < investor_preferences) investor_converts=FALSE;
else investor_converts=TRUE;

if (investor_converts) return your_percent_ownership * company_sale_price – your_aggregate_options_strike_price;
else {
common_stock_proceeds = company_sale_price – investors_preferences.
your_percent_common = your_percent_ownership / ( 1 – investor_ownership_percent );
return common_stock_proceeds * your_percent_common – your_aggregate_options_strike_price;
}

}

The one number you should know about your equity grant is the percent of the company you are being granted (in options, shares, whatever – it doesn’t matter – just the % matters).

Number of shares:  meaningless.

Price of shares:  meaningless.

Percent of the outstanding option pool:  meaningless.

Your equity in relation to other employees:  meaningless.

Strike price of options: meaningless.

The only thing that matters in terms of your equity when you join a startup is what percent of the company they are giving you.  If management tells you the number of shares and not the total shares outstanding so you can’t compute the percent you own – don’t join the company! They are dishonest and are tricking you and will trick you again many times.

I find it really depressing how often employees, especially engineers who are so smart about other mathematical issues, don’t get this.  I felt forced to post this after talking to a friend today who told me about how a prominent NYC startup has been telling hires the number of shares they are granted but won’t tell them the percent those shares represented (“it is company policy”), or the number you need to compute the percent – the total outstanding shares.  It’s really amazing people are getting away with this simple and incredibly cynical trick.

I’ve seen many companies “split the stock” 10-1 so that instead of, say, 10M shares there are 100M shares outstanding so the absolute number of shares granted sounds really big to naive hires who don’t understand that all that matters is the percent they own.

I think every engineering school in the country should have a week-long course on the basics of the capitalization of startups.  There are other things that matter too, but far less (like the number of preferences outstanding).   I’ll try to write about these other things in later posts.

Engineers – here’s how equity is paid out in a normal company sale/IPO (assuming a “good” outcome – in the downside cases it’s more complicated as investors have preferences which act like a max() function).  You get the percent you own multiplied times the price the company was sold for (or the market cap after IPO).  That is why percent ownership is the only equity number that matters.  Don’t work for someone who tells you otherwise or won’t tell you what percent you own.

The last step to raising venture capital is normally a 1 hour pitch to the whole partnership during their weekly monday meeting.  This is often described to entrepreneurs as a formality, but at least in my experience, for early stage deals, I would say there is probably a 25% chance of you getting a term sheet afterwards and a 75% chance of you getting rejected (although it will rarely come in the form of an actual “no”) .

The reason the odds of you getting dinged are that high are:

1) In most VC firms all it takes is one partner to say “This is really stupid – I hate it” to kill a deal.

2) Although by the time you pitch, the lead partner has probably told the other partners about you and probably sent around a memo, the non-lead partners probably didn’t pay attention, and only really do when you are presenting.

Good VCs have a much lower post-partnership ding ratio, because they work hard to socialize a deal and really get their partners to focus on it before asking the entrepreneur to present.   For example, I used to work for Rob Stavis at Bessemer and he had a much lower post-meeting ding rate.  This was because he spent a lot of time talking to his partners beforehand (“socializing the deal”), and if they had good objections he got them early on.  (Ps. Hopefully the VC will work extra hard to pre-sell the deal if they ask the entrepreneur to drop everything and fly across the country.)

The very worst thing that can happen in a partnership meeting is what I call the “partner ambush.”  Basically this is when the partner who brought you in (the “lead” partner), who you’ve met with for many hours and fully understands your company and is excited about investing in it, realizes midway through the meeting things are going badly and decides to try to save face by turning on the entrepreneur.

I had this happen to me when I was raising money for my last startup, SiteAdvisor.   Basically what happened is me and my co-founder Tom Pinckney walked into this big, well known VC firm at 4pm to a room of very tired looking guys (yes, they are all male) who had been hearing back-to-back pitches all day (side note:  always try to present in the morning).  No one introduced themselves or said hello, which was a bit unnerving.   The first questions were clearly hostile to the very idea of a consumer security startups (for a bunch of bad reasons, most VCs vastly prefer enterprise to consumer security – especially on the east coast and back in 2005).   One of them literally laughed at the idea of marketing via search engines (this is the east coast – believe it or not many VCs our here still don’t know what (white hat) SEO is and how important it can be).   Then the partner who brought me in said “Well, Chris, why not make SiteAdvisor into an enterprise product” basically turning on me and the whole concept of the company.  Things went downward from there.  To add insult and injury, the lead partner never even bothered to call me to ding me afterwards – in fact I haven’t heard from him to this day.

In retrospect, that would have actually have been a very good investment for the VC if they had actually given our pitch a fair hearing.  Which gets me to my final point:  I think VCs are making a mistake by putting so much emphasis on the partnership pitch.  There is some positive correlation between presenting to a room full of (sometimes hostile) VCs and building a successful startup, but not a very high one.

Besides missing good investments, the emphasis on the partner pitch leads VCs to invest in bad companies.  An investor friend of mine was recently talking about a failed startup he invested in:

Toward the end of the company, when things were going very badly, I went in and spent a day sitting with the entrepreneur and watching him work.  At that point I realized his one skill in life was pitching investors.  He had no idea how to manage people, build a product, get stuff done, etc.

The current early-stage VC process is optimized to favor people who are good at pitching partnerships, not necessarily people good at creating successful startups.

Bill Gurley posted a really nice summary of one of the main problems with the venture capital industry, and Fred Wilson responded here.   I totally agree with their analysis, but would add one more major problem with the venture industry to the list.  The fact that most VCs get rich via “management fees” just by showing up every day.

For those who don’t know, most VC’s get paid by so-called 2 and 20.  The 2 refers to the 2% of the fund they use to cover operating expenses and pay their salaries.  The 20 refers to the (normally) 20% “carry” fee – the percent of the profits they make for their investors that they get to keep.

Now I fully support carry fees – it is very similar to equity in a startup.  VC’s should get paid when they make money for their investors.

The problem is the management fees.  2% made sense back when VC funds were much smaller, but not now that they have gotten so large.  As peHUB said in their email newsletter today, Benchmark had an $85M fund in 1995 but today has a $500M fund.  That seems to be the typical trend for most big VCs.

Let’s do a little math.  2% of $85M is $1.7M.   Assuming 8 partners, that means salaries are in the $100-$200K range.  Much higher than national averages but, by the standards of finance, they aren’t getting “rich.”  2% of $500 is $10M, so each partner is probably getting $1M+ in salaries.   Over the 10 year life of the fund that’s $10M.  Even on Wall Street that is considered pretty rich.  And they get that money even if they make only bad investments and don’t return a dime to their investors.

This is why you see VCs raising bigger and bigger funds, why you frequently hear them say things like “I need to do 2 deals this year” and, worst of all, why you often see VC’s arguing for larger round sizes even if the startup has no productive use for the additional money – and even for the same percentage ownership.   In other words, in many cases VCs argue for a higher valuation just so they can “put more money to work.” Why?  If you raise a $500M fund and tell your LPs you are going to invest it over, say, 4 years, then its pretty hard to go back to them after a year and say “thanks for the $10M in management fees, I decided not to make any investments this year.”

VC’s seem to be a big fan of performance-based compensation when it comes to startups.  They should adopt it for themselves as well.

Products with so-called networks effects get more valuable when more people use them.  Famous examples are telephones and social networks.

“Complementary network effects” refer to situations where a product gets more valuable as more people use the product’s complement(s). Two products are complementary when they are more (or only) useful together – for example, a video game and video game console, or an OS and an application for that OS.  Microsoft Windows gets more valuable the more apps are made for it, which in turn makes Windows more popular, which in turn leads to more apps, and so on.  Microsoft Windows is not more valuable simply because there are more copies of Microsoft Windows in the world, but because there are more complements to Windows in the world.

Network effects can be your friend or your enemy depending on whether your product has reached critical mass.  Getting to critical mass in complementary network effect markets is sometimes called overcoming the “chicken and egg problem.”  Back in graduate school (2003), my friend Jeff Rhodes and I wrote a paper titled “Six Strategies for Overcoming the ‘Chicken and Egg’ Problem in Complement-Based Network Effects Markets.”  This is a frequent challenge when launching technology products, yet at least at the time we had seen very few people try to systematically document strategies for overcoming it.  Some of our examples are a bit dated now, but if you are interested in this topic you might like the full paper.

Here is a high level summary of the 6 strategies we describe with a few updated examples.  I’d love to hear from any readers who have more strategies and/or example products.

1. Signal long-term commitment to platform success and competitive pricing.   When Microsoft launched the original Xbox,  they made a big deal of publicly committing to spending $500M promoting the platform, thereby signalling that they were fully committed for the long haul and giving comfort to 3rd party game developers.   Another way to give comfort that your platform isn’t going away is to open source it – this way third parties know that even if the company stops supporting the product, independent developers can continue to do so (e.g. Google Android and Chrome).  Open sourcing also gives comfort that the company isn’t going to raise prices once they’ve reached critical mass.

2. Use backwards and sideways compatibility to benefit from existing complements. Microsoft of course has used backward compatibility very successfully for decades with DOS and then Windows, as have many game console makers.  In our paper we argue that the successful early bill pay (“bill presentment”) companies provided backward compatibility by sending snail mail checks to merchants who had yet to sign on to their electronic platform.

Virtual machines and Bootcamp gave Apple’s hardware some sideways compatibility with Windows.  Sun’s invention of Java could be seen as an attempt to introduce sideways compatibility between its shrinking server market and its competitors (Windows, Linux) by introducing a new, cross-platform programming layer.

3. Exploit irregular network topologies. In the last 90s, most people assumed that dating websites was a “winner take all market” and Match.com had won it, until a swath of niche competitors arose (e.g. Jdate) that succeeded because certain groups of people tend to date others from that same group.  Real-life networks are often very different from the idealized, uniformly distributed networks pictured in economics textbooks.  Facebook exploited the fact that social connections are highly clustered at colleges as a “beachhead” to challenge much bigger incumbents (Friendster).  By finding clusters in the network smaller companies can reach critical mass within those sub-clusters and then expand beyond.

4. Influence the firms that produce vital complements.  Sony and Philips, the companies that oversaw the successful launch of the compact disc technology in the early 1980′s, followed the CD launch with the introduction of the digital audiotape (DAT) in 1987. The DAT offered CD sound quality and, in a significant improvement over CD technology, it also offered the ability to record music.  Despite these improvements, the DAT never gained significant consumer adoption and ended as an embarrassing failure for Sony and Philips.  DAT failed because Sony and Philips failed to reassure record companies who were concerned that the recording capabilities of DAT would lead to widespread piracy.  Sony finally reached an anti-piracy agreement with record companies in 1992, but by that time consumer expectations for the DAT platform were dampened sufficiently to doom the platform.

On the other hand, when Sony and Philips launched the CD, they succeeded because they did a significantly better job influencing complement producers. Most importantly, they addressed the record companies’ primary concern by making CDs piracy resistant (or so it seemed at the time). In addition, Philips was able to influence Polygram, a major record label, to release music in the CD format because Philips owned a 50 percent stake in Polygram. Finally, Sony and Philips provided the record companies with access to their manufacturing technology and plant in order to ensure an adequate supply of complementary products. As a result, nearly 650 music titles were available in CD format when the first CD players were released and the CD format went on to become the most popular music format.

5. Provide standalone value for the base product.  Philips introduced the videodisc player (VDP) in 1979 as a competitor to the VCR. VDPs had slightly better picture quality than VCRs and had potentially lower hardware and software costs, owing to a simpler manufacturing process. However, the VCR had a 3-4 year head start on the VDP and had already developed an installed base of over one million units.

Providing a stand-alone use is the strategy that VCR producers used to achieve a successful launch and avoid fighting the difficult chicken and egg startup problem. Unlike the VDP, the VCR offered the ability to time-shift television programming. In fact, when the VCR was launched this was the only application available because the market for pre-recorded videocassettes had not yet developed. The standalone value for the VCR “time-shifting television programming” was sufficiently strong to get over a million people to purchase the product in the first 3-4 years after its launch. This installed user base of the VCR as a base product was sufficient to entice entrepreneurs to develop a market for pre-recorded videocassettes as complementary products in the late 1970′s. The complement-based network effect that resulted improved the value of the base product, increased sales velocity for the base and complementary products, and ensured that the VCR would be a common feature in most American homes.

A good modern example of this would be del.icio.us, which had stand alone value by storing your bookmarks in the cloud, and also had network effects with its social features.

6. Integrate vertically into critical complements when supply is not certain.  To overcome the chicken and egg problem, companies must find a way to ensure an adequate supply, variety, and quality of complementary goods. By vertically integrating into the complement product as well as the base product, a company can attempt to ensure an adequate supply of both goods.  Nintendo is the leading developer of games for its consoles, and Microsoft and Sony fund many of their most popular games.

Vertical integration is risky – as witnessed by the Apple computer in the late 80s and early 90s. By remaining tightly integrated, Apple precluded market competition from providing the necessary variety of price-competitive complements and base products.

**

Many of the above strategies (especially 3 & 5) apply to regular (non-complementary) network effect products.

One things I’ve noticed over the years is that equity grants given to new employees soon after Series A financings are generally a bad deal for those employees on a risk/reward basis.  (By a Series A financing I’m referring the first round of funding by VCs, where the amount raised is roughly $2M or more).

Here’s how equity is often granted from the very beginning of a company’s formation:

1. Founders decide on mostly equal split over beers.  It’s all just scribbles on a napkin at this point so equity flows freely.

2. In the cold light of day, founders renegotiate, with some founders possibly getting (significantly) more than others.

3. Employees who join pre-funding get reasonably big equity grants.

4. Series A financing occurs.

5. Suddenly equity grants to new employees are sliced an order of magnitude or more from what they were prior to Series A.

(Also, toss in there along the way one founder gets disgruntled and leaves – see founder vesting).

The problem is a Series A financing usually de-risks the company far less than the equity grants drop.  If I had to graph this in a totally unscientific way it would be like this (for successful companies – as represented by the green line going straight up):

Why do the equity grants drop so much after initial VC financings?

1) There are well established norms for post VC equity grants.  Going against them generates a lot of resistence from VCs.  By way of example, here are directionally accurate although probably 2x what I have typically seen post Series A.

2) Compounding this, after a financing the founders probably just got finished arguing for a smaller option pool to reduce their dilution, and it’s seems very hypocritical after that to argue for greater-than-standard equity grants.

3) The company now has an arms length valuation, probably in the multi-millions of dollars.  Suddenly 1% is worth “real money.”

The best time to join a company is at the very beginning – to found or co-found the company.  The second best time is to join before venture financing.  The third best time is when the company has started to ramp sales/traction – at that point your equity grant will be small but at least the company will have a much higher likelihood for success.  The worst time, from my experience, is right after initial (Series A) VC funding.

The flip side of this argument is after the company raises venture financing, an employee is more likely to get a “market” cash salary.  Personally I’d rather see people get bigger option grants post Series A and sub-market (or better yet subsistence) cash salaries – until the company is cash flow positive.  This is pretty much the opposite of Wall Street’s compensation schemes.  To me, as a principle, that means it’s probably a good idea..