There are a number of problems with the way that current startups are organized. In order to solve these problems, we are developing a new way to organize startup companies which mixes many of the ideas from cooperatives, and also throws in some ideas from Hollywood.

One goal is to find a fairer way to compensate people for their effort put into ZAT. At the same time, we always need to make sure that the compensation package for ZAT help ensure that ZAT is successful. We came up with an idea that borrows from the way that Hollywood pays “residuals”. Wikipedia describes residuals: “In business, a residual payment is one of an ongoing stream of payments for the completion of past achievements. The entertainment industry use of the word refers to a payment made to the creator of performance art (or the performer in the work) for subsequent showings or screenings.”

So, our idea is that in ZAT, people earn a share of the residuals. Let me describe this with an example. This is just a straw man proposal — I’m sure this will evolve over time, but you’re welcome to comment on this one.

Initially, let’s say that the total residual amount is the total profits from ZAT. Like a co-op, the total residual amount is divided up among everyone that owns shares in the residuals. People (employees) earn shares of the residuals for work performed for ZAT, or for other contributions. We might pay people shares of residuals weekly, or monthly, for example. Then, when ZAT starts earning profits, we divide these profits up among the shareholders, in proportion to how many shares they hold.

People earn shares for time they spend working on ZAT.

If you stop working for ZAT (for any reason) you keep your residual shares (but you do not earn new shares). In the future, you will collect your share of any profits just like anyone else. This is different from how co-ops work, where you normally lose any shares when you leave. This will allow us to bring people in to work on a specific part of ZAT and reward them later, even if they are not still working for ZAT.

We allow different accrual rates, so that, for example, one person could earn twice the share of residuals for the same amount of time spent as another person. We call this their “accrual multiplier”, a number that is part of everyone’s job offer. This will allow the company to recruit high powered people and compensate them appropriately. It also allows us to give “raises” to people who do extraordinary good work.

We might also want to have an additional multiplier that the company (board?) sets at appropriate times (four times a year?), which generally decreases over time. This factor is meant to compensate early employees more than later employees, since early employees are taking a bigger risk that their work will never be worth anything. I would prefer if the determination of this multiplier could be automated — for example it could be inversely related to how much ZAT is currently paying out in residuals since the more profit we are earning, the less the risk.

Shares of residuals have no ownership rights. They do not, in themselves, give you a vote in the company. They cannot be sold or assigned. But upon death, they can be inherited.

We might possibly want to limit the lifetime of shares. This will make accounting much simpler (open ended rights cause myriad problems). For example, if you leave ZAT your shares may remain the same for 5 years after that, then over the following five years they will be reduced a month at a time until your shares are zero at the end of 10 years. This solves several issues: first of all, code written today probably is not a part of any code base after 10 years, so the value of work has a limited time limit; second, it encourages people to continue to work for ZAT, which is a good thing; and finally, it helps limit the number of non-employees who own shares over time.

We should have some way of calculating the value of a residual share. That way, when ZAT has the money to pay actual cash salaries, we can give employees the option of receiving a cash salary as part of their compensation and fewer residual shares. This will allow us to, for example, hire people who have a family and need some cash, even if it is not the same salary they could receive elsewhere. There are standard ways of valuing options (Black-Scholes) that can help here.

If we have a value for a residual share, then we can also sell residual shares in exchange for money, allowing us to raise capital. We might need to make it so that residual shares that are awarded for money do not have a limited lifetime, so that we can get investment from people who are not employees. Again, these shares would not inherently have voting rights.

We might have a way for people (especially ex-employees) to sell their shares back to the company at the current value. We might even allow current employees to purchase shares at that value.

We do need to figure out a way to govern the company, so we will need voting rights of some kind. This is something we need to figure out. We could give voting rights to shares of residuals, or we could have “one person one vote” like worker co-ops, or we could have a board with sole voting rights (like some non-profits). There are many possibilities.

In practice, this works similarly to the way a co-op like REI works. At the end of the year, REI calculates its profits and declares a dividend. Each member of REI receives a portion of this total dividend, in proportion to how much they have purchased from REI in the preceding year. In our case, people would receive their portion of the total ZAT dividend, in proportion to their ownership of shares.

The biggest question I have is how to make this legal. There are LOTS of labor laws in this country, and if we are paying employees this way, we have to make sure we don’t violate any laws. For example, are our employees exempt? Or, since we are paying them for time worked, are they hourly? Do we have to pay overtime? I would prefer it if they are salaried, exempt (like most programmers are) but we need to make sure we have all the legal issues covered. This could be very expensive.

Luckily, we can look at contracts that exist for paying residuals. How are those normally handled? We also can look at worker coops, and how they work.

First up was the O’Reilly “Where 2.0” conference in San Jose.

Right after that was the Google Developer Day.

Last, but not least was WhereCamp, held at the Yahoo! campus.

The first one (Where 2.0) was a normal conference, and fairly expensive, but the other two were free. The Google Developer Day was paid for by Google (of course) but WhereCamp was done as a BarCamp, with no paid speakers and free admission and (even) free food. Interestingly enough, I enjoyed the two free conferences more than the paid one. The BarCamp model might be the wave of the future!

I gave a short (lightning) talk at WhereCamp, and also held a session on ZAT. Met several interesting people, and (finally) got a chance to talk to the Platial people, which is funny since they are located here in Portland.

–wm

This is exactly the problem we are trying to solve at ZAT. I can’t think of any reason why using the web can’t be every bit as good as using a standard guidebook, and lots of reasons why it could be a lot better. What’s keeping us from realizing that dream?

I’m attending several conferences this month. Earlier this month was Innotech, then WebVisions. Tonight and tomorrow I’m attending BarCamp Portland (come, it’s free!). Then at the end of the month I’m traveling down to San Jose to attend Where 2.0 and the Google developer day. Hope you see you at one of these events.

Even though ZAT is still under development, you can play with it at http://dev.zat.com. We would love your feedback.

England is a special case (as usual). First of all, there is the whole problem of whether the “country” is the UK, or if you should treat England, Scotland, Wales, etc. as separate countries. I opt for the latter. But even considering only England, there are many cities with similar names. Wikipedia lists four cities named “Newcastle” but they usually disambiguate by specifying a location, like “Newcastle upon Tyne”. Unfortunately, that still leaves Newcastles in Herefordshire, Shropshire (sometimes called Newcastle on Clun), and Staffordshire (Newcastle-under-Lyme). But forcing people to specify a shire name to disambiguate a city name in England seems a bit harsh.

Are city names required to be unique in Canada? I can’t think of any non-unique examples, but that doesn’t mean there aren’t any. [when researching "Newcastle", I discovered that there used to be two places in Canada called Newcastle, one in New Brunswick that has since merged with the city of Chatham, and one in Ontario that still exists. But I still don't know the answer to the question of whether there are now any ambiguous city names in Canada.]

There are no less than 15 cities called Newcastle in the US, but luckily they are all in separate states.

Why is this important to ZAT? Well, to make a city in the US unique, you have to say what state it is in. In most other countries, just saying the city name and the country name make it unique.

I am wondering if it is possible to design a language that keeps the power and flexibility of PHP, but without the warts.

Jim Buckmaster, the CEO of Craiglist, gave a talk at the UBS global media conference in New York, in front of a crowd heavy with investment bankers. They could not understand. The meeting was described as “a culture clash of near-epic proportions”.

I posted a comment on the above article — number 110 if you can read that far down the page!

–wm