Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: industrial research compensation

At 01:20 PM 6/12/01 -0400, Ludwik Kowalski wrote:
I never worked in industry but would like to know who
makes decisions about how much different people are
worth to a large company. Do rewards fluctuate from
year to year, according to assignments? Is there a
tenure equivalence to protect narrowly specialized
experts (those whose expertise gained over many
years would have little value outside the company)?

The following observations apply to big industrial research operations like
AT&T Labs and Bell Labs. It leaves out a whole lot of details. These are
just observations and should not be taken as an expression of anyone's
official policy.

1) It's partly subjective.

2) The research operation is divided in to "Labs" or "Centers" each of
which has a director, half-a-dozen department heads, and 60 or so
individual researchers. Each year the corporate Human Resources people
decide on how big the "raise pool" and "bonus pool" will be. (This
reflects competitive pressures among other things.) Then the director gets
together with the department heads and decides how the raises and bonuses
will be divvied up within the center.

3) If you produce something nifty every year, you will get a good raise
every year. If you have a bad year, you won't get fired, but you'll get a
crummy raise that year.

4) There is a fundamental dilemma: Is your performance reflected in your
raise, or in your salary? The answer is a little of both. If you are a
really strong performer and you are perceived as underpaid, you are going
to get a much much bigger raise than if you performed equally well but were
already well paid. As a result,
-- Over the long term (many years) salary eventually reflects long-term
-- Over the short term there are never wild fluctuations in salary.

5) If you do something that is of great value this year, but of
questionable long-term value, it might be recognized by a bonus rather than
a raise. Because of bonuses, total compensation can fluctuate from year to
year, but bonuses are usually small compared to base salary so this isn't a
big deal.

6) If you are just starting out, you shouldn't take on a project that will
have no results for five years. You want to start out with something where
there will be significant partial results and milestones along the way.
After you've been around a while, you get more rope to play with, but even
then it is wise to have a "portfolio" containing some bread-and-butter
stuff along with the long-term and/or high-risk stuff. Being narrowly
specialized is almost always a bad idea.

7) There is no such thing as tenure.... but there is a lot of institutional
memory of past contributions. It's hard to do something super-brilliant
every year, and management knows that.

If you turn in a long string of below-par performances, you will start
getting a lot more _advice_ from management. If advice doesn't turn things
around, eventually you will be _told_ what to work on, rather than being
left to decide for yourself.

If you have expertise in an area that is going out of fashion, you will be
encouraged to come up to speed in a new area, and given some time to do that.

8) There are rewards other than salary. One of the big "perks" at an
industrial lab is _leverage_. One type of leverage is having the company
commit millions or maybe billions of dollars to commercialize and deploy
your idea. (Some people find such leverage rewarding; some people don't.)
Obviously this can fluctuate completely from year to year. You might come
up with a really great idea, but still lose out because the company is
committed to going in another direction.

9) On a smaller scale, there is the leverage you get by collecting
collaborators to help you with your favorite project. This is subject to
fluctuation also, depending on how fashionable your project is.