November 2005

In the next few years, venture capital funds will find themselves squeezed from four directions.

They're already stuck with a seller's market, because of the huge amounts they raised at the end of the Bubble and still haven't invested.

This by itself is not the end of the world.

In fact, it's just a more extreme version of the norm in the VC business: too much money chasing too few deals.

Unfortunately, those few deals now want less and less money, because it's getting so cheap to start a startup.

The four causes: open source, which makes software free; Moore's law, which makes hardware geometrically closer to free; the Web, which makes promotion free if you're good; and better languages, which make development a lot cheaper.

VC funds will soon be squeezed from four directions. They're already stuck with a seller's market: too much Bubble-era money chasing too few deals.

Worse, those deals want less money, because it's getting so cheap to start a startup. Four causes: open source, Moore's law, the Web, and better languages.

VC funds will soon be squeezed from four directions. They're already stuck with a seller's market, and now the deals want less money because startups are getting cheap to start.

When we started our startup in 1995, the first three were our biggest expenses.

We had to pay $5000 for the Netscape Commerce Server, the only software that then supported secure http connections.

We paid $3000 for a server with a 90 MHz processor and 32 meg of memory.

And we paid a PR firm about $30,000 to promote our launch.

Now you could get all three for nothing.

You can get the software for free; people throw away computers more powerful than our first server; and if you make something good you can generate ten times as much traffic by word of mouth online than our first PR firm got through the print media.

And of course another big change for the average startup is that programming languages have improved-- or rather, the median language [blocked] has.

At most startups ten years ago, software development meant ten programmers writing code in C++.

Now the same work might be done by one or two using Python or Ruby.

During the Bubble, a lot of people predicted that startups would outsource their development to India.

I think a better model for the future is David Heinemeier Hansson, who outsourced his development to a more powerful language instead.

A lot of well-known applications are now, like BaseCamp, written by just one programmer.

And one guy is more than 10x cheaper than ten, because (a) he won't waste any time in meetings, and (b) since he's probably a founder, he can pay himself nothing.

In 1995 our biggest expenses were software, a server, and a PR firm—about $38,000 in all.

Now you could get all three for nothing: free software, cast-off computers, word of mouth beating print.

The median language [blocked] has improved too. Work that meant ten programmers writing C++ now takes one or two using Python or Ruby.

Better than outsourcing to India is to outsource, like David Heinemeier Hansson, to a more powerful language. BaseCamp was written by one programmer—more than 10x cheaper than ten: no meetings, and as a founder, no salary.

In 1995 software, hardware, and PR were our biggest expenses; now you get all three for nothing. And the median language has improved enough that one founder can do what ten C++ programmers once did.

Because starting a startup is so cheap, venture capitalists now often want to give startups more money than the startups want to take.

VCs like to invest several million at a time.

But as one VC told me after a startup he funded would only take about half a million, "I don't know what we're going to do.

Maybe we'll just have to give some of it back."

Meaning give some of the fund back to the institutional investors who supplied it, because it wasn't going to be possible to invest it all.

Because startups are so cheap, VCs now often want to give more than startups will take. One, after a startup took only half a million, told me: "Maybe we'll just have to give some of it back."

Because startups are so cheap, VCs now want to give them more than they'll take. One told me a startup he funded would take only half a million; maybe, he said, he'd have to give some of the fund back.

Into this already bad situation comes the third problem: Sarbanes-Oxley.

Sarbanes-Oxley is a law, passed after the Bubble, that drastically increases the regulatory burden on public companies.

And in addition to the cost of compliance, which is at least two million dollars a year, the law introduces frightening legal exposure for corporate officers.

An experienced CFO I know said flatly: "I would not want to be CFO of a public company now."

You might think that responsible corporate governance is an area where you can't go too far.

But you can go too far in any law, and this remark convinced me that Sarbanes-Oxley must have.

This CFO is both the smartest and the most upstanding money guy I know.

If Sarbanes-Oxley deters people like him from being CFOs of public companies, that's proof enough that it's broken.

The third problem is Sarbanes-Oxley: compliance costs over two million a year plus frightening legal exposure for officers. One experienced CFO told me flatly: "I would not want to be CFO of a public company now."

You can go too far in any law. This CFO is the smartest, most upstanding money guy I know; if it deters people like him, that's proof it's broken.

The third problem is Sarbanes-Oxley, which loads public companies with compliance costs and legal exposure. You can go too far in any law, and an excellent CFO refusing the job is proof this one has.

Largely because of Sarbanes-Oxley, few startups go public now.

For all practical purposes, succeeding now equals getting bought.

Which means VCs are now in the business of finding promising little 2-3 man startups and pumping them up into companies that cost $100 million to acquire.

They didn't mean to be in this business; it's just what their business has evolved into.

Hence the fourth problem: the acquirers have begun to realize they can buy wholesale.

Why should they wait for VCs to make the startups they want more expensive?

Most of what the VCs add, acquirers don't want anyway.

The acquirers already have brand recognition and HR departments.

What they really want is the software and the developers, and that's what the startup is in the early phase: concentrated software and developers.

Google, typically, seems to have been the first to figure this out.

"Bring us your startups early," said Google's speaker at the Startup School.

They're quite explicit about it: they like to acquire startups at just the point where they would do a Series A round. (The Series A round is the first round of real VC funding; it usually happens in the first year.)

It is a brilliant strategy, and one that other big technology companies will no doubt try to duplicate.

Unless they want to have still more of their lunch eaten by Google.

Of course, Google has an advantage in buying startups: a lot of the people there are rich, or expect to be when their options vest. Ordinary employees find it very hard to recommend an acquisition; it's just too annoying to see a bunch of twenty year olds get rich when you're still working for salary.

Even if it's the right thing for your company to do.

With few startups going public, succeeding now equals getting bought. So VCs pump little 2-3 man startups into $100 million acquisitions—not by design, just what their business evolved into.

Hence the fourth problem: acquirers realize they can buy wholesale. Why wait for VCs to make startups more expensive? What they want is the early startup's concentrated software and developers—the rest they already have.

Google was first to see this. "Bring us your startups early," said its speaker at Startup School—they acquire right where a startup would do its Series A. Others will copy it, or lose still more of their lunch to Google.

Google also has an advantage: its people are rich, or will be when options vest. Ordinary employees find it hard to recommend an acquisition—too annoying to watch twenty year olds get rich while you work for salary.

With few startups going public, succeeding now means getting bought, so VCs fatten 2-3 man startups into $100M acquisitions. The fourth problem: acquirers realize they can buy wholesale, and Google figured it out first.

The Solution(s)

Bad as things look now, there is a way for VCs to save themselves.

They need to do two things, one of which won't surprise them, and another that will seem an anathema.

Let's start with the obvious one: lobby to get Sarbanes-Oxley loosened.

This law was created to prevent future Enrons, not to destroy the IPO market.

Since the IPO market was practically dead when it passed, few saw what bad effects it would have.

But now that technology has recovered from the last bust, we can see clearly what a bottleneck Sarbanes-Oxley has become.

Startups are fragile plants—seedlings, in fact.

These seedlings are worth protecting, because they grow into the trees of the economy.

Much of the economy's growth is their growth.

I think most politicians realize that.

But they don't realize just how fragile startups are, and how easily they can become collateral damage of laws meant to fix some other problem.

Still more dangerously, when you destroy startups, they make very little noise.

If you step on the toes of the coal industry, you'll hear about it.

But if you inadvertantly squash the startup industry, all that happens is that the founders of the next Google stay in grad school instead of starting a company.

There is a way for VCs to save themselves: two things, one unsurprising, the other an anathema.

The obvious one: lobby to loosen Sarbanes-Oxley. It was meant to prevent future Enrons, not destroy the IPO market—but few saw what a bottleneck it would become.

Startups are fragile plants—seedlings, in fact—worth protecting because they grow into the trees of the economy. Politicians realize that, but not how easily startups become collateral damage of laws meant to fix something else.

And when you destroy startups they make very little noise. Step on the coal industry's toes and you'll hear about it; squash the startup industry and the founders of the next Google just stay in grad school.

VCs can still save themselves with two moves. The obvious one: lobby to loosen Sarbanes-Oxley. Startups are fragile seedlings that grow into the economy's trees, and when you squash them they make almost no noise.

My second suggestion will seem shocking to VCs: let founders cash out partially in the Series A round.

At the moment, when VCs invest in a startup, all the stock they get is newly issued and all the money goes to the company.

They could buy some stock directly from the founders as well.

Most VCs have an almost religious rule against doing this.

They don't want founders to get a penny till the company is sold or goes public.

VCs are obsessed with control, and they worry that they'll have less leverage over the founders if the founders have any money.

This is a dumb plan.

In fact, letting the founders sell a little stock early would generally be better for the company, because it would cause the founders' attitudes toward risk to be aligned with the VCs'.

As things currently work, their attitudes toward risk tend to be diametrically opposed: the founders, who have nothing, would prefer a 100% chance of $1 million to a 20% chance of $10 million, while the VCs can afford to be "rational" and prefer the latter.

Whatever they say, the reason founders are selling their companies early instead of doing Series A rounds is that they get paid up front.

That first million is just worth so much more than the subsequent ones.

If founders could sell a little stock early, they'd be happy to take VC money and bet the rest on a bigger outcome.

So why not let the founders have that first million, or at least half million?

The VCs would get same number of shares for the money.

So what if some of the money would go to the founders instead of the company?

Some VCs will say this is unthinkable—that they want all their money to be put to work growing the company.

But the fact is, the huge size of current VC investments is dictated by the structure [blocked] of VC funds, not the needs of startups.

Often as not these large investments go to work destroying the company rather than growing it.

The angel investors who funded our startup let the founders sell some stock directly to them, and it was a good deal for everyone.

The angels made a huge return on that investment, so they're happy.

And for us founders it blunted the terrifying all-or-nothingness of a startup, which in its raw form is more a distraction than a motivator.

If VCs are frightened at the idea of letting founders partially cash out, let me tell them something still more frightening: you are now competing directly with Google.

My second suggestion will seem shocking: let founders cash out partially in the Series A round. Instead of all the money going to the company, VCs could buy some stock from founders directly.

Most VCs have an almost religious rule against this. Obsessed with control, they fear losing leverage once founders have any money.

This is a dumb plan.

Selling a little early aligns founders' attitude toward risk with the VCs'. Otherwise they're opposed: founders with nothing prefer a 100% chance of $1 million to a 20% chance of $10 million; VCs can afford the latter.

The real reason founders sell early is that they get paid up front. That first million is worth so much more than the rest; let them take it and they'd happily bet the rest on a bigger outcome.

So why not let founders have that first million, or half? The VCs get the same shares either way.

Some will call this unthinkable. But the size of VC investments is dictated by the structure [blocked] of VC funds, not startups' needs—and as often as not it destroys the company rather than grows it.

The angels who funded our startup let us sell them some stock—good for everyone. It blunted the terrifying all-or-nothingness of a startup, in raw form more a distraction than a motivator.

If VCs are frightened at letting founders partially cash out, let me tell them something still more frightening: you are now competing directly with Google.

The shocking second move: let founders cash out a little in the Series A round. Most VCs religiously refuse, fearing lost leverage. But it aligns founder and VC attitudes toward risk—and that first million is worth far more than the rest.

Thanks to Trevor Blackwell, Sarah Harlin, Jessica Livingston, and Robert Morris for reading drafts of this.

Thanks to Trevor Blackwell, Sarah Harlin, Jessica Livingston, and Robert Morris for reading drafts.

Thanks to the friends who read drafts.