It makes sense that the first firms to take this approach are themselves
consulting companies. They have the pressures of variable business and
cost issues, and they already need to track costs pretty closely. Making
this work for a product company seems harder to me, but I'm not sure why.
Possible reasons:
-- long-term product development needs people to see the firm's best
interests as their own
-- products require more coordination than contracted projects or
services, despite what coders think.
-- products aren't as well thought-out initially than tasks a company
plans to out-source
-- the difficulty of signaling that the same task is more critical or more
profitable on one product than another
-- the difficulty designing products to interface with other company
product and to scale appropriately (i.e. company needs that are separate
from market needs)
-- if a product has direct impact on a company's revenue, then upper
management needs to meddle or it would never get out the door ;-)
Am I off-base here? It could be that this is happening left and right,
it's just happening slowly.
Here's the one example I have. Chris Cole, the person I mentioned in my
earlier post, has what he calls the "Product Author Model". Senior coders
in his company pitched ideas, and when they are given a green light they
were responsible for design and implementation of that entire product.
They were funded for their salary and other costs of developing the
product, including salaries for other coders (though Chris believed in
small coding teams, much like Brooks in _The Mythical Man-Month_). Product
Authors also got a cut of the profit from that product, usually around 6%.
If people are interested, I can come up with some more details about this
model.