Author Archives: pmarca

Public speaking versus blogging

Used to be, if you wanted to get a message out into the market, you would give a talk at a conference, a reporter would write down some of what you said and mangle the rest, and you’d call it a day.

Or, you could shortcut the process by simply giving an interview to the reporter and letting him mangle what you said directly.

These days, you have the option of staying home, blogging in your underwear, and not having your words mangled.

I think I like the direction things are headed.

Mid-year resolution #1: No more public speaking.

Mid-year resolution #2: More blogging.

Friend Connect, Open Social, Ning, and the web

First, I’m very happy to say that Ning will be rolling out our formal production support for Open Social in June.

In case you missed the news at the time, Open Social is a standard way, sponsored by Google, to build new features (“gadgets”) and/or plug those features into social networks all over the web, including social networks on Ning.

Ning has had beta versions of Open Social running for six months now so we’re excited to be able to now provide Open Social in production to all of our Network Creators and users. Open Social will be available within all 265,490 social networks already running on Ning, as well as the 10,000+ new social networks being created on Ning every week now.

Second, I’d like to discuss Google’s new Friend Connect initiative and how we plan to support it at Ning.

We at Ning think that Friend Connect is a great idea, and has huge potential to make Open Social even more functional and widely available for a broad swath of our users and Network Creators on Ning and throughout the web.

However, in the last couple days, there’s been some confusion around the idea that perhaps Friend Connect is somehow competitive with Ning — which is odd, because we don’t think so and because we think it’s obvious that it’s not. So let me start by first explaining what Friend Connect is, and then how Ning is going to implement it.

Friend Connect is a mechanism by which Open Social gadgets can be published and used not just within a social network but also beyond that social network. When an Open Social gadget shows up elsewhere on the web, via Friend Connect, the friend data and social context comes with the Open Social gadget from its origin social network — and that origin social network might be a network on Ning or a large walled garden network like MySpace or Orkut, and that Open Social gadget might be embedded on any page anywhere on the web.

In a sense, Friend Connect one-ups Flash widgets. Many social networks and other content hubs today publish Flash widgets like video players and music players that get embedded in pages all over the web. Friend Connect is a mechanism that provides the embedding capability for Open Social gadgets to be used all throughout the web — with the added benefit that with a Friend Connect-enabled Open Social gadget, the user gets her social context anywhere she goes, which isn’t the case with a typical Flash widget.

Now, as you are hopefully aware, Ning is a service for creating your own social network for anything — with your choice of features, your design, and your members, customized however you want it.

From a strategy standpoint, we want to enable maximum flow both into and out of Ning networks and the rest of the web. It should be as easy as possible for users to get from elsewhere on the web into a Ning network, and likewise as easy as possible to flow from a Ning network to anywhere else on the web — and ideally, while taking their social context with them. We think this makes strategic sense for two key reasons:

  • First, it’s good for users, and whatever is good for users is good for a service like Ning. We think that’s obvious.
  • Second, you don’t get lots of flow into anything on the web without having lots of flow out to the broader web. We think that’s also obvious — you are compromising your own product to your self-inflicted detriment if you’re not making it as easy as possible for activity to flow out as well as in. Google of course itself illustrates this — Google’s primary business, search, generates revenue purely by having people leave Google, by clicking on an ad — and that’s no accident, and there is no shortage of people who flow into Google as a result.

As a result, we have from the start published Flash widgets and more recently Facebook apps that any Ning user can embed elsewhere on the web or out to their Facebook page — to extend content and functions from a Ning network out onto the web. This has obviously been a good thing to do because users love it, and because those widgets and Facebook apps link right back to their origin Ning networks and drive traffic back in even while propagating content and functionality out.

For Ning, Friend Connect is simply a new and better way to do the same thing with Open Social gadgets — in both directions: out and in.

We will support Friend Connect in two ways:

  • Every network on Ning will be able to be an Open Social origin social network — pushing out Open Social gadgets to anywhere else on the web that carry with them the social context and friends data from their origin Ning network. So, for example, the members of a backpacking social network on Ning can still interact as friends on any third-party backpacking web site, by publishing an Open Social gadget out from their Ning network onto that third-party web site. In short, people will be able to flow more easily from Ning to many other web sites without losing the social context of their Ning networks.
  • Every network on Ning will of course be able to contain Open Social gadgets published out from other social networks on the Internet via Friend Connect. So, for example, a group of friends on MySpace who all enjoy cooking will be able to travel from MySpace to a cooking-specific social network on Ning, via any Friend Connect-enabled Open Social gadget published from MySpace into that Ning network. In short, people will be able to flow more easily from other social networks and walled gardens into Ning social networks without losing the social context from those other networks.

Ning’s role remains the same — to be the easiest, most powerful, and most widely used way for anyone to create your own social network for anything. Friend Connect then makes social networks on Ning more powerful and more relevant for a larger base of users to do more on the web, to flow out and to flow in — a good thing for those users and a good thing for us.

In praise of dual-class stock structures for public companies

A dual-class stock structure means that a company has two different classes of common stock. Each class of stock has the same economic ownership of the company, yet different voting rights.

In a typical scenario, Class A shares have a single vote per share, whereas Class B shares have 10 votes per share, for any shareholder vote.

Using this mechanism, for example, the Class B shareholders might only own 20% of the company in economic terms but have a clear majority voting position relative to the Class A shareholders.

In short, Class A shareholders have shares labeled with the earlier letter in the alphabet, but Class B shareholders control the company — in stark contrast to the more normal single-class stock structure which is more classically democratic: “one share, one vote”. Since Class B shareholders will typically be some set of founding management or founding investors in the company, in practice the presence of a dual-class stock structure means that the founders control the company and can overrule all other shareholders on a wide range of issues, including if and when to sell the company.

Both public and private companies can have dual-class stock structures, but the controversy around dual-class stock structures is usually confined to public companies, due to the presence of public shareholders. And so I will focus purely on public companies.

I used to be an absolutist against dual-class stock structures — I used to believe that dual-class stock structures were obviously a bad idea, that the democratic single-class approach of “one share, one vote” was more fair to public investors and more likely to lead to a healthy company in the long run, since total founder control of a public company can allow the founders to overrule normal market forces and the interests of their public shareholders.

And in fact, practically all investor advocates and shareholder activists agree with that stance — dual-class stock structures are at the top of the list of techniques that entrenched managers can use to foil the normal market discipline of a public stock, and to frustrate outside public shareholders who can easily become disenfranchised even when they have majority ownership of a company… with a long-run outcome similar to the kind of insularity and inbreeding you find in royal families. These days, the New York Times Company has of course become the poster child for entrenched bad management operating against the interests of their public shareholders due to its dual-class stock structure — how could anyone possibly be in favor of that?

And on the face of it, a dual-class stock structure simply seems unfair — how can someone own part of something but have a tenth of the rights of someone else who owns the same amount?

After 15 years in the technology industry, though, I have done a complete 180-degree turn on the topic — with some caveats.

I come not to bury dual-class stock structures, but to praise them.

I now believe that dual-class stock structures are a great idea for a technology company that is in the process of going public, under the following conditions:

  • The key leaders of the company — typically the founders — who will own the controlling Class B shares, are also major economic shareholders in the company. They own a significant portion of the company and are therefore highly incented to maximize the value of the company over time.
  • The key leaders of the company who own the controlling Class B shares have a long-term goal of building a major franchise, and the commitment required to execute against that goal.
  • The controlling Class B shareholders have a commitment to treat Class A shareholders fairly and equally in all respects other than voting power.
  • All public shareholders understand what they are getting into up front — no bait and switch.

The key to the whole thing is shared goals — particularly the shared goal of long-term value creation, particularly the creation of a long-term franchise, the kind of franchise that can require 10 years or longer to build.

With such goals, I now believe the interests of public shareholders will often be better served by ceding voting control to the founders and key leaders of the company.

This is a provocative statement, so let me back it up.

In practice, the world at large, the markets in which companies operate, and Wall Street in particular, throw up all kinds of short- and medium-term noise in the face of every public company, all the time.

And in fact, my sense is that the level of such noise is steadily increasing for about a dozen different reasons, including but not limited to the proliferation of hedge funds, buyout funds, arbitrage funds, corporate raiders, shareholder activists, shareholder representation firms like ISS, sell-side analysts, cable television financial news, financial web sites, Internet message boards, online stock trading, increased consumer interest in stocks and markets, and visibly shortening investor time horizons across the entire landscape.

When you are running a public company, here are some of the things that get routinely thrown at you that have practically nothing to do with building a long-term franchise:

  • Stock market booms and busts — the stock market is bipolar; it doesn’t matter what finance academics say about efficient markets, everyone knows greed and fear whipsaw the market around all the time.
  • Economic booms and busts — e.g. this ridiculous credit crisis and real estate fiasco. Modern economies are apparently characterized by one self-inflicted crisis after another. And even when you’re not directly affected by a particular crisis, if you’re running a public company, you’re probably going to be indirectly affected, often for no good reason aside from the universe’s desire to inflict collateral damage.
  • Hedge funds aggressively short-term buying and shorting stocks for the quick pop, and often spreading malicious and untrue rumors along the way. I’m no Patrick Byrne, but every CEO of a public company regularly contends with just silly rumors all the time that are obviously being spread by someone talking their book, or rather lying their book — and SEC oversight of such market manipulation is almost completely absent.
  • Leveraged buyout funds that make apparently attractive buyout offers financed with massive amounts of debt, and then strip-mine the company for fees and dividends before sending it back out into the public markets weaker than before. These guys are wonderfully skilled at paying themselves; on average, their franchise-building skills are questionable at best.
  • Corporate raiders of various stripes. I’ll certainly grant that corporate raiders as a category have probably been good for capitalism as compared to the clubby Fortune 500 status quo of the 1970’s, but when a raider gets his hooks into your public company, he’s only in it for the quick pop, and he’ll agitate to get it acquired as fast as possible.
  • Hostile takeovers — which may provide a quick payoff to current investors but which definitively bring to an end any opportunity to build a long-term franchise.
  • The intense quarterly earnings guessing game that you end up playing even if you don’t want to — even if you refuse to issue guidance, and perhaps especially if you refuse to issue guidance, in which case Wall Street just goes ahead and sets expectations for you without consulting you. The vertiginous stock drop that follows “missing your numbers” can actually damage your company — you wouldn’t believe how many customers check Yahoo Finance before each sales call.
  • Financial journalists — who can be outstanding writers with journalism degrees from the best schools, and in many cases know almost nothing about the companies they are covering or the products those companies make, which does not keep them from writing all kinds of nonsense. High-quality business journalism is distinctly the exception, not the rule; every CEO knows it, and the noise from inaccurate bad press can again actually damage your company.
  • And then, finally, pure good old fashioned company-specific fluctuations — sometimes things are going well, sometimes they’re going poorly. If you’re building a franchise, that’s OK, and even to be expected; you just need to power through the rough patch. However, if you’re subject to short-term market demands, a rough patch can kill your dreams amazingly quickly.

All of these things can meaningfully interfere with long-term value creation, particularly if you are trying to build a long-term franchise.

The huge advantage of a dual-class stock structure is that it lets the company’s core management simply ignore most of this stuff and stay focused on the long-term goal.

What’s the ideal situation for a public shareholder with a long-term time horizon? To invest in a company whose leaders are highly motivated to build long-term value — to grow the value of the company 10x or 100x or 1,000x — not flip it to the first interested acquirer for a quick pop, or even the fifth, or the tenth. And therefore to invest in a company whose leaders have the ability to pursue building for the long term, versus getting constantly compromised by short-term market noise.

At this point, if you listen closely, you can hear the howls of outrage. They are saying, how can shareholders expect to countenance being effectively powerless?

And of course the answer is alignment of goals.

Investors that have short-term goals, or even medium-term goals, shouldn’t invest in public companies with dual-class stock structures. Remember, I’m presuming no bait and switch. You are always free to not invest in any company for any reason, including this reason.

But, if you’re an investor with a long-term time horizon, you will, I believe, be best served investing in companies with a similar long-term time horizon.

The best part of taking this position is that I get to roll out the big gun: Warren Buffett similarly advocates dual-class stock structures for precisely this reason, and puts his money where his mouth is — he famously has been a long-term investor in the Washington Post Company, for example, which has a dual-class stock structure that gives the Graham family total control, and which has been a stellar long-term investment for Berkshire Hathaway.

Now, dual-class stock structures have been customary in the media industry for a long time, but are relatively new to the technology industry. The most prominent example of a public technology company with a dual-class stock structure is of course Google, whose structure puts total voting control of the company in the hands of Larry Page, Sergey Brin, and Eric Schmidt. Corresponding to that, Larry, Sergey, and Eric have made a 20-year commitment to Google, and are clearly pursuing the goal of building a long-term franchise.

As far as I can tell, shareholders haven’t exactly been scared off from investing in Google as a result.

If anything — and I haven’t done a statistical analysis of this, but just look at the charts and the stock prices of this decade — Google may actually be getting a premium in the market due to its dual-class share structure, as investors are able to make a clean bet on long-term value creation, and they know that the core team can just put their heads down and power through any short-term nonsense.

I think Google has changed the rules on this topic — I think many technology companies, certainly the ones with high potential, that go public over the next decade will have dual-class stock structures, due in part to the Google precedent.

On to some practical questions:

Don’t companies with dual-class stock structures risk limiting their access to capital if they are only attracting long-term investors?

Perhaps, but again, Google is a clear counterexample. You can hardly say it’s been starved for capital.

More generally, I would say that this question reflects the fact that companies with dual-class stock structures are still subject to market discipline. If the market overall doesn’t like the dual-stock structure, it can refuse to provide capital to those companies, and instead provide that capital to companies with shorter-term objectives and more outside shareholder control. But I don’t predict that will happen, and I would be willing to bet my own company on that.

Viewed systemically, dual-class stock structures are an alternative governance model that can compete in the open market for capital with other governance models. Capital will flow appropriately. All is good.

How would you apply this to the drama unfolding around Microsoft and Yahoo?

Well, clearly, if Jerry Yang and David Filo had dual-class-powered voting control of Yahoo, the whole situation there would be playing out very differently.

Microsoft would have been forced to negotiate a purely friendly deal from the very start, and at a price that would have caught Jerry and David’s attention from the start. Hostile threats would have been meaningless. Had a deal gone down, it would have been on Jerry and David’s terms, and the premium might have been even higher than a normal process would generate, since Jerry and David would have had the perfect walkaway option: we’re not selling, and there is no appeal — pay up or shut up. And the company could have been entirely focused on current operations the whole time — no distraction.

But what about the outside shareholders? Several of Yahoo’s largest outside shareholders — including one who owns 16% of the company, four times the amount Jerry himself owns — are in the national press tonight saying, we are furious Yahoo didn’t sell for $34/share when Microsoft was already offering $33/share.

The obvious answer: in the alternate scenario with a dual-stock structure and founder control, those outside shareholders would not have invested in Yahoo in the first place, unless they believed Yahoo had a valid plan for long-term value creation and building a franchise.

If, in that alternate scenario, investors didn’t believe Yahoo had a valid plan for long-term value creation, then that’s another matter. There would be no excuse for that. But that would be a very different problem that would apply regardless of stock structure.

In point of fact, many of Yahoo’s largest shareholders today are also, or have been in the past, major Google shareholders. Clearly Google’s dual-class structure didn’t scare them or their peers off at any point I could see — instead, Google shareholders seem delighted to be aligned with a core team that has the control to execute a long-term plan.

And here’s the kicker: it’s not like outside shareholders in Yahoo, even though they own over 70% economic and voting control in the company, can just make the company do whatever they want — as you can see from all their frustration in the press tonight. Sure, the outside shareholders as a group will ultimately get whatever they want, up to and including a sale to Microsoft, but they may have to put up a significant fight to do so, and that fight may require a significant amount of time and effort, with significant opportunity cost. And along the way, the process has been and will be characterized by confusion, ambiguity, and uncertainty. The whole thing is clearly a massive distraction to any form of long-term value creation to the point where even Microsoft believes Yahoo is risking damaging its long-term prospects by reacting to Microsoft’s hostile public bid. So it’s hard for me to see how a single-class share structure is nirvana for anyone in a situation like this.

So what about the New York Times Company?

Well, it is true that the New York Times Company and similar failing newspaper companies — most of them, except for the Washington Post Company with its superior diversification — with dual-class stock structures are not exactly good investments today, since their entrenched management teams can fight off shareholder activists and hostile takeovers indefinitely while riding their declining franchises straight into the ground.

But on the other hand, it’s not like you couldn’t have seen it coming. Every investor in any declining dual-stock media company today knew they were buying into that stock structure and did it with their eyes open. And any investor still holding stock in such a company has been aware of the Internet for 15 years and has been able to track the performance of the company’s management team in dealing with the Internet over that entire time. Certainly it’s possible to be delusional about your investment and think that recovery is right around the corner, but you can’t blame the stock structure for that delusion.

And remember, the New York Times Company had its dual-class stock structure for decades, and for much of that time, ordinary investors would have done very well to own its shares. It just so happens that the wrenching technology shift that is causing so much trouble coincided with a generation of managers who are unprepared to deal with it. That’s life.

So I think the fate of the dual-stock media companies has a lot more to do with the “media company” part of it and a lot less to do with the “dual-stock” part of it. The only investors mad about the dual-stock part — well, the non-delusional ones — are the ones who want the short-term stock pop from a sale to Rupert Murdoch. And those are not the investors you want if you are trying to build a franchise.

What’s your recommendation to technology companies that are going public?

Strongly consider implementing a dual-class stock structure, but only under the following conditions:

  • The founders are committed to run the company for the long term and want to build a real franchise.
  • The founders are also major economic owners of the company.
  • The founders have an absolute commitment to treat all other shareholders fairly, and to consider themselves entirely in the same boat economically.
  • All public shareholders starting with the IPO know exactly what they are getting into — no bait and switch.

Under these conditions, a dual-class stock structure is not only an outstanding idea — I think, for our industry, it may be the future.