Author Archives: pmarca

I think I can, I think I can

…but if I think I can too well, maybe I can’t…

From a new paper in the Journal of Economic Psychology:

High failure rates and low average returns suggest that too many people may be entering markets as entrepreneurs. Thus, anticipating how one will perform in the market is a fundamental component of the decision to start a business.

Using a large sample obtained from population surveys conducted in 18 countries, we study what variables are significantly associated with the decision to start a business.

We find strong evidence that subjective, and often biased, perceptions have a crucial impact on new business creation across all countries in our sample.

The strongest cross-national covariate of an individual’s entrepreneurial propensity is shown to be whether the person believes herself to have the sufficient skills, knowledge and ability to start a business.

In addition, we find a significant negative correlation between this reported level of entrepreneurial confidence and the approximate survival chancesof nascent entrepreneurs across countries.

Our results suggest that some countries exhibit relatively high rates of start-up activity because their inhabitants are more (over)confident than in other countries.

Ning passes 100,000 social networks

As I’ve previously discussed, my new company Ning exists to give everyone the ability to create your own social network for anything — in less than two minutes, for free — with the ability to customize your network any way you want.

Today we passed 100,000 social networks on Ning.

This chart shows the number of social networks on Ning since we rolled out the current version of our service earlier this year:

As you can see, 70,000 networks have been created in the last seven months alone!

What does it mean to have 100,000 social networks?

As you might expect, the 100,000 networks on Ning follow a power law curve for any metric you choose to apply: number of members, say, or number of page views.

Internally we think of the networks on Ning as falling into three buckets:


  • Big networks — the top, say, 200 networks at any point in time that each have a large number of members.


  • Long tail networks — smaller networks with, by definition, more than 1 member but fewer members than the big networks — could be dozens, hundreds, or thousands of members.


  • Throwaway networks — networks that were created, maybe have a few members, but are not being used.

A couple of interesting things about those buckets:

First, we are not as dominated by big networks as you might think. In fact, well over half the page views on Ning are generated by networks other than the top 200 — the long tail is alive and well on Ning!

Second, it is our goal to have a meaningful number of throwaway, or unused, networks on Ning at any point in time. The whole point of Ning is ease of creation— how easy it is to create your own social network, and play with ideas around that concept.

Think of how easy it is to sign up for a Yahoo account — that’s how easy we want it to be to create your own social network on Ning, and a natural consequence of that is going to be a certain number of throwaway social networks, just like Yahoo has a certain number of throwaway user accounts.

This has been misunderstood by some otherwise very bright people, but think about it: what would be good about a service that’s so hard to use that there are no throwaway networks?

Third, they’re not static buckets — new networks pop into the first category regularly, older networks fall out of the first category into the second category, and sometimes networks that were created for a specific purpose (say, a fundraising drive) stop being used altogether. That is all good — even in the last case, those networks’ users have been introduced to Ning and many of them will go on to create their own networks for other purposes.

Why is number of networks a significant metric?

This rapid growth rate of number of networks is significant to us because the number of networks on Ning is the leading indicator for all of our other key usage metrics, including registered users, unique users, and page views — plus our key revenue metrics: advertising impressions and premium services fees.

This is because Ning is different than your typical social networking service.

On a typical social networking service, users join a single large social networkdesigned and run by the service’s owner. Users then invite other users to join that same single large network — this is the viral adoption loop by which the network grows.

On Ning, users both join existing user-created networks — one of the 100,000+ networks that already exist — and/or create their own networks. This is a double viral loop. Loop one is users being invited to join a network created on Ning. Loop two is some percentage of those users creating their own new networks and then inviting other people to join those new networks.

On Ning, these two viral loops feed one another: the more networks on Ning, the more users who are getting invited to join — and the more users who join, the more new networks that get created, leading to even more users being invited to join. And the loops repeat over and over again.

And as a result, the key driver of this whole dual viral adoption cycle is new networks being created — on average, existing networks continuously bring in new users over time, and those new users create new networks, which add to the base of existing networks bringing in new users, ad infinitum.

And in fact, although as a private company we are quite happily exercising our freedom to not publish all of our internal metrics, the ones we focus on are all growing nicely:


  • Page views are growing about 40% month over month — i.e., roughly doubling every two months. (Our internal definition of page view corresponds to the number of ad impressions we will be able to serve.)


  • Registered users are growing as a big healthy multiple on the number of networks. This is despite the fact that Ning, unlike most other social networking services, does not require users to register in order to view most content. On Ning, you only have to register to participate directly in a network or to create your own network. (This is both because this is a friendlier model for users, and because we believe page views is the key revenue driver, not registered users.)


  • Unique users are growing as an even bigger healthy multiple on the number of networks. (Because we don’t force registration to view content, our uniques as a multiple on registered users will be higher than many other services.)

Digging into more specifics about the 100,000 social networks already created on Ning:


  • The networks on Ning range across an unbelievable spectrum. Here you can see a cross-section of the creative, vibrant networks that have been created on Ning over the last few months.


    We love this diversity because we’re building a broad-based horizontal service — and the more different topics people build networks around, the more diverse the group of people who will come and join and create their own networks.



  • And to that point, the diversity of the user base on Ning is enormous. Some social networking systems start by appealing to a specific vertical — undergraduates, or Los Angeles club kids — and grow from there. Ning has been horizontal and diverse from the start. Kids to adults, many socioeconomic backgrounds, many levels of technological sophistication, many countries of origin.


    The theme is that there isn’t a theme — which we love, as confirmation of the universal appeal of this kind of service.



  • Number of users per network is similarly across the board, from single digits to tens of thousands.


  • As I’ve discussed previously, Ning is a platform that allows infinite customization of social networks that users create. On Ning, you can customize practically ever aspect of your network, from the look and feel and feature set if you’re a normal user, all the way down into the source code if you’re a programmer.


    As a consequence, the 100,000 networks on Ning span the gamut from completely uncustomized, to extensively customized, to completely reprogrammed.


I’d like to close with a few general comments about where we’re at and what we’re doing.

First: interestingly, rapid growth of all of these metrics on Ning has coincided with the explosion of public interest around Facebook — as people learn about social networking on a system like Facebook, they get even more interested in creating their own social networks and engaging in targeted, topical social networks on Ning. Or, put more simply, the market is very large and growing fast — people love this stuff.

Second, without tooting our own horn too much, passing 100,000 networks and growing rapidly means that we’re the clear and overwhelming leader in the business of enabling people to create new social networks.

In particular, in the last several months there have been a number of companies announced that are in the business of providing so-called white-label social networks for businesses. Ning is different — we’re building a broad-based consumer service where anyone can create a new social network for anything, in less than two minutes, for free. The white-label services, in contrast, typically charge businesses tens or hundreds of thousands of dollars per network, and it takes a lot longer than two minutes to tilt them up once you pay the fee.

Ning’s success is going to make it very hard for these business-oriented white label services to exist. We already have a broader feature set and more customizability than all of the white label services that we know of, and that gap is likely to only get wider as the superior economics of our self-service consumer model let us develop more features and more platform capabilities faster than, and operate at much higher scale than, any white label service that has to acquire customers one at a time through an expensive direct sales effort.

Third, to make our service better and to ensure that everything I just said isn’t a load of hot air, we have an incredibly aggressive slate of product development projects ahead of us over the next 6-12 months. More on those shortly. We will in no way be standing still.

Paging George Orwell, Joppatowne Maryland edition

Bear in mind that this is a public American high school

In late August, Maryland’s Joppatowne High School became the first school in the country dedicated to churning out would-be Jack Bauers. The 75 students in the Homeland Security and Emergency Preparedness magnet program will study cybersecurity and geospatial intelligence, respond to mock terror attacks, and receive limited security clearances at the nearby Army chemical warfare lab.

The new school is funded and guided by a slew of federal, state, and local agencies, not to mention several defense firms. Officials say it will teach kids to understand the “new reality,” though they hasten to add that the school isn’t focused just on terrorism. [Of course not — they also cover surveillance.]

School administrators… refused to be interviewed for this story. But it’s no secret that the program is seen as a model for the rest of the country, with the Pentagon and other agencies watching closely.

Students will choose one of three specialized tracks: information and communication technology, criminal justice and law enforcement, or “homeland security science.” David Volrath, executive director of secondary education for Harford County Public Schools, says the school also hopes to offer “Arabic or some other nontraditional, Third World-type language.” [Perhaps Busuu, or Njerep, or Zaparo? Nah, I’m guessing mostly Arabic.]

…[I]t’s not clear how many Joppatowne grads will be on track to join the upper echelons of the intelligence community and how many will wind up as airport screeners. “We do want to encourage higher education,” Volrath says. “We also want to be realistic. Some of these defense contractors will have huge security needs, and the jobs won’t require four years of college.”

From Mother Jones, courtesy of Marginal Revolution.

That’s it, public company accounting is broken

One of the dirty little secrets — or rather, dirty huge non-secrets — of Wall Street is that public company accounting has been diverging further and further from cash accounting — which is to say, reality — over time.

Over the last several years, a whole series of new laws and rules have larded up income statements and balance sheets with all kinds of fictional, non-cash components to the point that you basically can’t conclude much about any public company financial statement you see, except that you really better read all the fine print.

Our new poster child:

Thanks to a relatively new accounting rule, firms like Morgan Stanley, Lehman Brothers and Goldman Sachs last quarter booked hundreds of millions of dollars in gains based on worsening perceptions of their own creditworthiness.

How does that work?

If the market decides a company is a bigger credit risk and starts demanding fatter risk premiums to buy its debt, the value of its existing debt falls. Under a rule being phased in throughout corporate America known as Financial Accounting Statement No. 159, that same logic applies to a company’s own debt. Companies that mark their liabilities to a market price, as Wall Street usually does, thus record as revenue a drop in the value of their own debt obligations.

In essence, they make money because they owe less. [Well, not really. They still have to pay off the debt according to its original terms. Unless they actually buy it back — which they apparently are not required to do, according to this accounting rule — and which, if they actually did buy it back, could move its price right back up to fair value. This situation is thoroughly screwy.]

Accounting experts [here it comes] said the exercise is perfectly legitimate, particularly if firms that mark liabilities to market do the same with their assets. At the same time, it highlights one of the ironies of so-called fair value accounting. “If you have a liability that declines in value because your credit worsens, you have a gain,” said Stephen Ryan, associate professor of accounting at New York University’s Stern School of Business.

But Moody’s Investors Service said buyers should beware of gains booked when brokers mark down their own debt liabilities. “Moody’s does not consider such gains to be high-quality, core earnings,” it said in a report issued Friday. [Cough.]

FAS 159, which brokers are adopting earlier than most companies, couldn’t have come at a better time for Wall Street. The firms are taking writedowns of billions of dollars to reflect the lower value of leveraged buyout loans and securities backed by mortgages and other assets that are stuck on their books. Concerns about those exposures weighed on perceptions of the big investment banks’ creditworthiness all quarter. The cost of protecting their bonds against default shot up, and the risk premiums on their debt widened as well.

Morgan Stanley said it booked about $390 million, or about 26% of its third-quarter profit, “from the widening of credit spreads on certain long-term debt” that it has issued. It isn’t alone. Goldman Sachs said Thursday it booked a gain of “a little bit under $300 million” because of adjustments in accounting for its structured notes. That’s a sliver of the $2.9 billion of net income it reported in the third quarter, but still helped offset some of its fixed-income and equities losses.

Bear Stearns Cos. added $225 million of equity revenue, “principally reflecting gains in our structured notes portfolio.” Without it, the 83-year-old firm – which boasts on its website that it has never had an unprofitable year — would have registered a third-quarter loss…

I’d just like everyone to whom I owe money to know that you’re never going to get it back. And I plan on booking the dollar value of your outrage as revenue in my current quarter.

From the Wall Street Journal.

Best customer service letter ever

Letter from online electronics vendor Woot as published on Woot’s web site:

I have received more than three emails from Zune buyers who are upset about Woot dropping the price of the Zune by $20 one month after it went on sale the first time. After reading every one of these emails, or at least scanning their subject lines, I have some observations and conclusions.

First, I need to make a better effort to hide my email address.

Second, I am sure that we are making the correct decision to lower the price of the 30GB Zune from $149.99 to $129.99. This confidence is based on more than the holy doctrine of corporate infallibility. The Zune is a breakthrough product, and we have the chance to “ride the lightning” and “shoot the curl” this holiday season, not to mention “kill the messenger” and “rock the vote”, further enabling us to “pay the rent” and “keep the lights on”. It benefits both Woot and every Zune user (but especially Woot) to drag as many new victims as possible into the Zune “dungeon”. We strongly believe that misery loves company this holiday season.

Third, being in technology for 1+ years, give or take a year, I can attest to the fact that the technology road is bumpy. There is always some idiot changing lanes without signaling, and the potholes never seem to get fixed. If you always wait for the next price cut or to buy the new improved model, you’ll never buy any technology product. I mean, why should you? Truth is, you don’t really need any of this junk. We’re afraid you’ll catch on to that fact and overpaid frauds like me will have to go back into fields like telemarketing and burrito construction. Fortunately, most of you continue to languish in a consumerist stupor, wallets spread wide for us to plunder as we please. The bad news for us is that if you buy products from companies that support them well, you will receive years of useful and satisfying service. But we’re hoping you’ll buy from Woot instead.

Third-and-a-half, even though we are making the right decision to lower the price of the Zune, and even though the technology road is, like, this total Deathrace 2000-type scene, we need to do a better job taking care of our early Zune customers, at least until we find a private security firm we can afford. For some reason, our early customers trusted us. We must live up to that trust with our actions in moments like these, lest you turn off the money spigot that maintains our decadent lifestyles. These peacock-egg omelets and mink-lined Jacuzzis don’t pay for themselves, you know.

Therefore, we have decided to offer every Woot customer who purchased a Zune from us on August 22, 2007… a $10 Woot credit towards any Woot order of $40 or more… We make this decision with every confidence that most of you will never want any of the crap we sell anyway.

We want to convincingly pretend to do the right thing for our valued Zune customers. We’d apologize for disappointing some of you, but we long ago lost the capacity for sincere remorse. We will continue to do our best to trick you into having high expectations of Woot.

As an actual Zune customer, I think he makes some excellent points.

Department of astonishing chutzpah, Richard Brodhead edition

In the wake of the now-debunked rape case against three lacrosse players, Duke University will establish a center devoted to justice and training lawyers to fight wrongful convictions, president Richard Brodhead said Wednesday.

Duke will invest $1.25 million over the next five years for the project at the law school, which will also expand its Wrongful Convictions Clinic and Innocence Project. The clinic and the Innocence Project investigate claims of innocence by the state’s convicted felons and raise awareness of problems in the criminal justice system.

Via the Associated Press.

Despite the case being such a travesty from the start — the prosecutor, Michael Nifong, was ultimately disbarred and ejected from office — Brodhead has become famous for stating the need for the accused students to be, quote, “proved innocent” by the criminal justice system, turning the whole American concept of “innocent until proven guilty” on its head.

Presents! Presents, falling from the sky!

The class-action legal juggernaut that was the law firm [and organized criminal conspiracy] now known as Milberg Weiss suffered more body blows yesterday as its co-founder, Melvyn I. Weiss, was indicted…

[P]rosecutors in Los Angeles charged Mr. Weiss, one of the architects of class-action securities lawsuits, with conspiracy, racketeering, obstruction of justice and making false statements to a grand jury… could face a sentence of as much as 40 years…

The indictment also broadened existing charges against Milberg Weiss, the firm Mr. Weiss co-founded in 1972, contending that the firm received some $250 million in legal fees over the last 25 years from class-action cases in which it paid kickbacks [bribes] to individuals who had served as named, or lead, plaintiffs…

“The indictment outlines a decades-long kickback scheme that was deliberately concealed from courts that were overseeing significant class-action cases,” said George S. Cardona, the United States attorney in Los Angeles, in announcing the charges against Mr. Weiss…

The charges against Mr. Weiss contend that he knew about and participated in the plaintiff kickback scheme since its inception in the late 1970s, and that he continued to participate in it in even after federal prosecutors began their investigation.

But there’s more!

In a related development, Steven Schulman, a former named partner at Milberg Weiss, agreed to plead guilty to a conspiracy charge in connection with the plaintiff kickback scheme.

Mr. Schulman, who will cooperate with prosecutors, also agreed to disgorge $1.85 million in profits, pay a $250,000 fine and accept a prison sentence that is likely to be 27 to 33 months, according to court papers.

All the gory details about this organized criminal conspiracy were outlined in detail in a sworn statement from their partner and now convicted felon David Bershad from earlier this year. Fun reading!

News clippings from the New York Times.

Go boomers!

For Americans ages 35 to 54:


  • 18,249 deaths from overdoses of illicit drugs in 2004, up 550 percent per capita since 1975…


  • 46,925 fatal accidents and suicides in 2004, leaving today’s middle-agers 30 percent more at risk for such deaths than people aged 15 to 19…


  • More than four million arrests in 2005, including one million for violent crimes, 500,000 for drugs and 650,000 for drinking-related offenses… [representing] a 200 percent leap per capita in major index felonies since 1975… [and this doesn’t even include OJ!]


  • 630,000 middle-agers in prison in 2005, up 600 percent since 1977…


  • 21 million binge drinkers (those downing five or more drinks on one occasion in the previous month), double the number among teenagers and college students combined…


  • 370,000 people treated in hospital emergency rooms for abusing illegal drugs in 2005, with overdose rates for heroin, cocaine, pharmaceuticals and drugs mixed with alcohol far higher than among teenagers…


  • More than half of all new H.I.V./AIDS diagnoses in 2005 were given to middle-aged Americans, up from less than one-third a decade ago…

    What experts label “adolescent risk taking” is really baby boomer risk taking. It’s true that 30 years ago, the riskiest age group for violent death was 15 to 24. But those same boomers continue to suffer high rates of addiction and other ills throughout middle age, while later generations of teenagers are better behaved. Today, the age group most at risk for violent death is 40 to 49, including illegal-drug death rates five times higher than for teenagers.

    From Mike Males writing in the New York Times.

    p.s. Think anyone’s ever going to notice that these people are the children of the quote-unquote Greatest Generation? Paging Tom Brokaw…