Tag Archives: vc

21st Mar, 2008

10 comments

Take the Next Step, Paul

In college I had a wonderful Humanities professor who insisted on making us write short essays so we could practice writing succinctly. After each essay she would personally sit down with us and critique our logic (and our grammar!). Her feedback to me was almost always the same, “your argument is logical and supports your conclusion but you need to take the next logical step. What does your argument imply?”

I was never able to take the next step, even when pressured. And she never took it for me. It would be fair to say that I hated her during these meetings.

Today I ended up quoting her while reading Paul Graham’s “You Weren’t Meant to Have a Boss.”

Paul’s thesis is that typical big business drains the life out of its employees because we weren’t meant to work in such large groups. It’s unnatural. To truly live, we need to be in groups small enough that we have room for creativity and freedom of action. That’s the way nature intended.

I agree. Jay and I talk all the time about how much more fun we’re having at CrowdVine than any of our other programming jobs. We’re free to build product. Programming isn’t just a job for us, it’s our hobby and passion. Being in a small group for the first time really is bliss. We’re not the only ones saying that either. Talk to people who’ve been much more successful than us like 37Signals or SmugMug. They’re not just successful, they’re happy.

So while I agree with everything Paul wrote, I found myself screaming, “take the next step, Paul!”

He’s a venture capitalist. He’s promoting programmers joining startups. Venture backed startups start as everything he describes–small companies that are great places to work and learn. But they only stay that way for a few years.

By definition the startups are either going to grow into an awful company with bosses or be acquired by an awful company with bosses (or fail). The startup founders are either going to turn into bosses (which Paul correctly points out isn’t very rewarding either) or they’re going to turn into employees with bosses.

The logical step that Paul couldn’t take is that he’s wrong for being in the venture business. The venture business depends on an ecosystem of bosses. Even if his founders feel like they’re getting a fair trade for a few unhappy years at a big company, they wouldn’t have the option of either growth or acquisition if other programmers couldn’t be pursuaded to work “unnaturally.”

The difference with companies like 37Signals and SmugMug (and CrowdVine) is that while they have the same natural working conditions, they’re structured so that those conditions don’t have to end. If Paul really wants to create good jobs he should turn YCombinator into a small business incubator.

Great discussion of this on Hacker News including responses from Paul. One commenter there made a big fuss that I was technically incorrect to call Paul a venture capitalist. True he’s a new un-labeled form of investor who’s using his own money and experience, and not using money from a venture fund. However I stand by my argument, which is based on the exit pressures which are very VC.

9th Oct, 2007

8 comments

Commodity Web Startups

Interesting posts from Paul Graham and Fred Wilson about the trend of decreasing software development costs leading to lots of people starting companies. They tend to focus on the impact on the VC world (because they’re VCs). I’m interested in the impact on founders.

From what I can tell by living and working in the bay area, the assumed life cycle of a startup goes like this:

You have an idea, you turn that idea into a compelling elevator pitch, then use the pitch to raise a seed round of investment so that you can build a prototype. Then you raise another round so you can build it to a point where it might actually attract and support customers. Then you raise another round to build up your infrastructure because you’re about to get heavy traction. Hopefully you’ve sold by this point. If not, you raise another round of funding so that you can build the company into a real business with actual revenue. Hopefully someone buys you soon because there’s no way your new revenue is going to cover your expenses. If you somehow ended up with a profitable business and no one has bought you, then you IPO.

My experience building CrowdVine is that the drop in software development costs and the increased availability of low-cost
infrastructure turn the above idea on its ear. Here’s how I’ve experienced it.

Seed Stage

When I started CrowdVine I avoided investment for three reasons. I felt that venture capitalists weren’t aligned with my goals as a programmer. I didn’t need money because I thought I could build everything without help and because I had a few small contract gigs that paid the rent without sucking up all my time. Also, nobody was offering me money.

Plenty of people have noted that the goals of VCs and entrepreneurs don’t always line up, but at least they draw from the same motivational well: financial gain. As startup costs drop you’re going to get more founders who aren’t primarily entrepreneurs, they’re primarily do-ers (programmers, designers, etc.) A lot of them are going to have different motivations. Mine are, in order, pay the rent, build something, make that something wonderful, and get positive feedback. My motivations never line up with investors in the seed stage. They only line up later if I build something wonderful that lots of people want and delivering it to lots of people requires upfront money.

If you’re founding a company but you’re not capable of building the product yourself, then you’re not taking advantage of the trend. Software development got cheaper but communication didn’t. Pure idea/sales/marketing founders are losing value against founders who can build their own product. The wave of new founders will have only one early expense: cost-of-living. If they’re not capable of supporting themselves with contract work then they’re either too young to have the right contacts or they’re not going to do very well building a company that can support themselves (so they may as well get Venture Capital involved right away). Paul Graham’s Y-Combinator program seems smart to be targeting young entrepreneurs who aren’t necessarily ruled out for having founder-type qualities but also don’t have other good options.

The reason nobody offered me money is because I didn’t go around asking for it. All the VCs expect you to put together a power point deck, drive to their office, and make a pitch. But none of that plays to my strengths. The problem is that I didn’t know very much about what the company was going to do and I’m not good at hyping something that doesn’t exist. I think a lot of programmers are essentially realists and don’t like making promises about vaporware. The Y-Combinator program has started to emphasize the qualities of the founders over the quality of the idea. That seems appropriate. At the seed stage, engineer founders need VCs to remove a lot of the funding friction in order to even get to the point where venture funding is a choice. The only way I would have had the option to take seed funding is if CRV (the Odeo backers) approached me and said, “Heard good things about you from the other Odeo folks. Looks like you’re going out on your own. I can give you $250k if you promise to come back and demo a product for us in six months.” If they needed me to be convincing about the idea then there would never be a deal.

Early Stage

CrowdVine now has enough revenue to pay the salaries for two people (plus some). We’ve identified a market that looks promising (conference social networks) and we’re making headway in that market. Also, Jay and I are happier than we’ve been at any other job because there are no barriers to building software for people that are going to give us positive feedback.

Now we’ve got to polish the product, create a repeatable business process, and get a consistent revenue stream that extends past the end of the year. We’re lucky that revenue is already well above costs, so we have spare cycles to build the business. If we had gone with an ad or subscription supported model we might have a ways to go until revenue caught up with costs. In that case I’d be hesitant to continue to support the business with unrelated consulting. It’s unfair to your customers who are expecting your full attention.

We had a brief period of this which we solved with CrowdVine related consulting. We built two customized social networks based on the CrowdVine platform. Jay seems to like this work, so we’ll probably continue to do more.

We could also ditch the consulting and look for a seed round, although I still don’t think my interests are aligned with
the interests of venture investors.

Growth Stage

Hopefully we manage to get our product out of beta, build a repeatable business process, and attract a steady stream of customers. Our margins are good so it’s easy to grow slowly. However, we have a pretty big gap between contacting a customer and receiving revenue (let’s say six months). So if we wanted to grow quickly we’d have to find a chunk of money in order to hire a bunch of staff. The company would start to need skills that I don’t possess (like managing a sales force) and would need those skills quicker than I could learn on the fly.

In the current venture capital structure, this is the first time where taking investment starts to make sense to me. Our interests are roughly aligned (we both want to grow) and they have things I want (money and advice). However, there’s two other good options.

One option is to leverage the money from our conference business to build fully automated subscription or ad supported businesses. 37 Signals has several profitable subscription sites that don’t seem to take a lot of their time. They get growth because each site continues to grow and because they have enough spare cycles to roll out new products without hiring a bunch of people. They also started out doing enough consulting to pay the rent.

The other option would be to reject massive growth in favor of running a small company that’s great to work at. I wouldn’t call this settling. Maybe this is the Adaptive Path model. Despite selling every single product they ever built (one, to Google), they seem to have refocused on building a highly respected design and usability firm. I’d be pretty lucky (and thankful) to have what they have.

Exit Strategy

There’s a common perception in the startup world that all great visionaries and developers have short attention spans. This is bullshit. Linus Torvalds doesn’t have a short attention span, or at least not one that makes him switch projects every two years. Venture capitalists do have short attention spans, maybe for personality reasons, but definitely for institutional reasons. My attention span is at least ten years (proven three times now) so I’m pretty sure that any venture backed competitors aren’t playing the same game I am.

If we take any investment we’re going to have to commit to an exit strategy that is either to be acquired or to go public. If we don’t take investment money then we don’t have to consider either option. In fact, sale changes from an exit strategy to a sustaining strategy. If we end up with something that’s valuable but we’re tired of running, then it can be sold. The Ebay market for startups seems to price these sales in the hundreds of thousands of dollars, not the hundreds of millions.

In the 37 Signals model, it looks hard to sell any one product because there doesn’t seem to be any person who owns any one of the products. However, if you setup the products so that each one has a clear GM/CEO-type then you can sell that product without giving away the entire company. Obvious seems to have gone down this road with Twitter. Ev is the CEO of Obvious and he spun Twitter out as it’s own company with Jack as the head. Jack is the man that makes Twitter run and if
they ever sell that company then he’s the one who’s going to join the acquiring company. I’m sure he also has ownership incentives that would make him agreeable to sale. I think of this as the Buffet model for structuring a company: find managers you trust and then give them incentives and authority to run.

I don’t have any exit strategy. I like working and having my own company has been an excellent excuse to work more. I think the most valuable thing I could build is a company that I’d want to work at for the next fifty years. I recognize though that not everyone is going to want to spend fifty years working for me, so I’m looking for people who I trust and who have the initiative to some day run their own businesses. My hope is that this company can give people room to grow so that they can eventually run businesses within it. Cheaper software development means that a lot of “businesses” could be run by a single person. Maybe Jay will be running our consulting division next year.

Going out of business

There’s a pretty big risk that the company doesn’t grow at all. If we were venture backed we’d have to start flailing until we ran out of money. If I was purely a businessman, I’d have to walk away or start over. However, I’m an engineer and running a two (or even one) person company still fulfills many of my motivations.

Since I built the company without debt, I consider the risk of going out of business to be roughly equivalent to the risk of having an extended streak of unemployment. I don’t have a fortune 5000 company that’s committed to pay my salary, but I do have a much better skill set than I did a year ago (I did a lot of programming) and I have a product that I can leverage for consulting work. I predict a lot more of these startups turn into small businesses that stick around.

30th Jul, 2007

2 comments

All My Friends Go WIth Union Square

Twitter and Wesabe both took funding from Union Square Ventures. Intriguing. So I did some research and found out two interesting things. One, they are located in NYC, not on Union Square in SF.

Two, they write really excellent posts about the companies they invest in.

Here’s Fred Wilson’s take on Twitter

There is something really powerful about public, asynchronous text communications where a reply is not expected. A great example is blogging. You blog something and it’s out there on the Internet for public consumption. Others read it and they either comment or create their own blog post in reaction. Collectively, we engage in a discussion.

…

Twitter provides a platform for banter that blogging doesn’t and it’s available in so many places via IM, mobile text messaging, or the Web that it induces a different sort of behavior. Twitter encourages people to adapt and invent behavior to suit their needs.

Synchronous communication wasn’t working for me, not so much that it failed to function but that I failed to use it. Twitter is now the only online way that I communicate socially. No emails. No IM.

Here’s what Brad Burnham said about Wesabe.

If you manage your expenses on a web based service you have the opportunity to contribute to community and to take advantage of its collective wisdom. Allowing your service provider to aggregate transaction data anonymously makes it possible for that provider to deliver a service that is better than desktop software in a number of important ways.

1) Providing very useful analytics, that compare your behavior to others like you. Do you spend more or less than most folks in your community for cable television, or lawn care?

2) More information about the vendors you use every day. Is it going to cost you more to bounce a check at Wells Fargo or at Wachovia? The answer turns out to be less than obvious.

3) Information about how others feel about service providers in your world. It turns out that many folks are willing to say how they feel about the places they spend their money. Would it help you to know that of the three dry cleaners in your neighborhood, one had a 100% satisfaction rate?

4) Peer produced data categorization and cleansing. I have given up using my annual gold card statement from American Express, because half of the vendors are listed as an unrecognizable string of characters, and even when they get the vendor right, they often do not put that vendor in the right category. Once I contribute my data to a co-op, a lot of these things are fixed much more easily. If anyone participating in the community recognizes an incomprehensible string of characters as “Whole Foods” and makes the change in their account, everyone in the community benefits from their contribution. After three or four people do it, the service provider can begin making the change. If most people categorize expenses in certain ways, the service provider can usefully suggest categories, and auto-fill entries to speed you on your way.

Wesabe is the only service to ever give me a useful view of my data. It’s not a competition over features, the other competitors flat out fail.