John’s Blog

Since I’m leaving Reactivity, I thought it’d be appropriate to post the note that started it off back more than 7 years ago, for posterity’s sake. I’ll have more to say about leaving probably next week.

—

Last Friday, Adam, Tom Kosnik (a biz school prof who teaches at Harvard and Stanford) and I were hanging out and talking–mostly I was trying to figure out what I do next. To make a long story short, it looks for sure like ATG is going to get demolished this month, which means that Apple isn’t a place that I want to be any more. So we were talking about startups that Tom knows that I could go to, then it occurred to me that I could start doing freelance work–there seems to be an awful lot of work to go around in the valley right now. Then Tom asked both of us what our ideal jobs would be. And, bottom line is that I really want to spend my days with people I both care about and respect, and spend my time helping to *create* something. And that my skill set seems to be helping people to connect with each other and to do cool things together.

One of the things that I’m starting to realize is that nobody is going to *give* me my perfect job–and so it seems like the best way to do something fun is to just create the perfect job and figure out how to make it work.

Tom noted that the biggest problem in the valley right now is that nobody can find good, talented engineering or creative people–Sun has something like 1000 job openings, Intel has the same, Cisco and Oracle as well. Microsoft is going to be hiring something like 3500 people this year. But for us–everyone on the distribution list–finding talented people is about the easiest thing we can do–I know *tons* who are looking for interesting places to be. And almost everyone that I know says that they’re “planning to start up in a year or so…” but just looking for the right opportunity.

And so Tom suggested that we just take a look at what the best biz schools and engineering schools in the country are (Stanford Engineering, Stanford GSB, MIT Engineering, MIT Biz, HBS, CMU Engineering, Cornell Engineering, Cornell Biz, Princeton Engineering) and try to figure out how to build a business built around collecting the right groups of people together.

Which brings us to the actual idea:

A metastartup (this is a temporary name only–trying to think of a better one) is a company who’s mission is to foster a community of talented engineers and business people, with the goal of spinning off startup companies from that community, as well as to build a loosely coupled network of those companies.

I believe that there will be very many types of flavors to a metastartup–a couple have already emerged:

IdeaLab (http://www.idealab.com/): This is a company in Pasadena founded by a guy named Bill Gross. The basic idea is that he’s got lots and lots of ideas for startups, as well as a whole bag of money, and is recruiting anybody he can get his hands on to come work on one of his ideas to start up. I think that they’ve got 18 companies in their portfolio after a couple of years. He basically sets people up with offices, HR functions, ideas, a team, and VC. Check out http://www.herring.com/mag/issue44/gross.html, which is an article from The Red Herring.

Interval Research (http://www.interval.com/): Interval is a research tank of about 100 people, created by Paul Allen, and funded from 1992 to 2002 by $100M, I think. It’s mostly composed of relatively famous designers who have made their marks in the computing industry, although there are quite a few artsy types there as well. Their business model is to be really free and loose and let their researchers work on whatever they want, and to spin off ideas whenever it seems appropriate. They’ve got 3 spinoffs to date: Carnelian (software for traditional publishers), Purple Moon (software for girls) and Electric Planet (also software for girls). Not a lot of public information about them; they’re extremely inwardly focused.

What I’m thinking of, though, is less patriarchal, less funded, and more egalitarian than IdeaLab, and much more urgent, much less funded, and much less academic than Interval.

The model would be to have a core set of people in the metastartup, who focus on continuity, mentoring, cross-fertilization of ideas, and recruiting, and hiring mostly young people, mostly right out of school or after their first or second high tech job, for a time period between one year and two years, with the expectation that sometime in the second year of being at the metastartup, they would find a team to join that spins off, with the metastartup retaining some equity portion of the spinoff, and the spinoff retaining some reciprocal equity stake in the metastartup (and, transitively, in every other spinoff in the community).

Which all sounds sort of neato–but the big question is obviously how to pay everyone’s salaries. The revenue model that I’ve been working on so far has a few different possibilities. The obvious ideal is to have a few of the startups go public, which would presumably fund metastartup development for some reasonable length of time, and that’s my expectation for the long term. But how to get it started? Seems like there are two possibilities (along with some versions that are mixtures of the two): venture capital funding (Tom Kosnik seems to think that this would be relatively easy to get) or self-funding, through consulting work.

VC has lots of different possibilities–the major advantages would be (1) a close connection with a VC’s network and (2) cash. I think that there are a couple of disadvantages: (1) you lose quite a bit of your equity when VC’s fund and (2) in funded ventures, people seem to have a problem with “keeping their eye on the ball.” That is, when your salary doesn’t depend on execution, it’s easy to work on something that’s not quite critical to your business. One of the guises of a VC-connection that Adam and I just talked about is that there would be the possibility of a VC deciding that this is a really good way to develop engineering and creative talent as a resource for startups, and absorb the whole metastartup as some sort of development arm.

The consulting route is clearly harder, at least at first, because of the obvious problem of splitting time between what you want to be doing, developing the startups, and what pays your bills. I think that we can mitigate this problem a bit by being sort of careful about the consulting jobs we pick (there seems to be way more consulting work in the valley than anyone can do)–and trying to position the company as a sort of IDEO, that is, an advanced research company. One possible example is that a friend of mine is the CEO of an internet startup (about 100 people, I guess), and they’re really looking to build some sort of research lab to do advanced prototyping–the metastartup could contract that work out, doing market research and prototype and technology design. I have the feeling that we could find a lot of work like that.

There are a bunch of other ideas in this, but it’s really late, and I need to go to bed–one more piece, though, is one borrowed from Interval, and that’s the idea of people in the company having a certain number of “project points” that they have to spend. So, for instance, you might assign everyone 20 points wehn they start–and their main project, the one that nothing takes priority over, is worth 12 points, a consulting project is worth 5 points, and smaller roles on other people’s project might be worth 2 or 3 points each. That’s a neat idea for at least 3 reasons—it helps cross-fertilization of talent and ideas, it gives us a way to measure % of time spent bringing in dollars, and it would give us a way to involve people at some level less than a full employee–you could imagine hiring “associates” or something that don’t quit their day jobs, but get involved at the level of 5 points or so–this would be a risk reduction way to get people involved and engaged. Better than that, though, is that it would provide a way to get students still in school involved–to have associate groups at Stanford and in Boston, for instance, of students spending some % of their time working on metastartup stuff–which could be anywhere from contributing ideas from their scholarly work to doing actual consulting work. Seems like that would be a really strong hook for recruiting.

The last thing that I’ll say tonight is that I believe that the metadata idea could provide a way to get away from the “startups-that-make-the-three-founders-and-only-them-rich” category, by focusing on building a very egalitarian group, with very open policies on equity and salary.

Anyway, more tomorrow night, hopefully, but I wanted to get this started at least, and hopefully get reactions to it.

Thanks–

John
lilly@research.apple.com