I gave a talk on my dissertation work.  Several of the more senior scientists were amazed at the bonafide science it contained.  They'd had me pinned as someone who did 'service work'-- programming, project management, all the stuff that enables science for others (them). This is the kiss of death to scientific collaboration.  Once you're marked with the taint of 'service', everyone assumes you're a scientific moron. 

You can tell the difference in how they explain something.  If you're a service person, they give you analogies ("imagine if...").  If they see you as a fellow scientist, they go right to authors and citations so you can read the original paper.

Neither approach is particularly efficient.