The numbers fit our internal research pretty well. In aggregate we run 16 doors per managment software sale. I would guess (but have no way of confirming) that about 5% of our jobs are 100 doors and more, with 1-2% 500 doors or more. Systems often continue to grow over time and because our channel strategy includes distribution, we often are surprised to find out the size of some installations.
Across the board, the average size of an access job is 12 - 14 doors. However, the bimodal distribution here reveals the standard deviation is broad...
Thanks, Brian! A couple of thoughts...
Might it be more correct to attribute the bimodality to the preference or ability of integrators to procure certain size jobs, and not necessarily the distribution of the jobs themselves?
Consider the fact that if true bimodality existed at the job size level that was independent of integrators preferences/abilities, you would have detail data that would look like two humps, a major one and a minor one like so:
Integrator A - Big hump 4-16 Smaller hump: 32-64
Integrator B,C Somewhat similar distribution...
And that is what all your anecdotes are saying above, agreed?
But since you asking all the respondents to give you a single number, to in effect average their humps, you should actually see a single hump right in the middle, where the two humps average together.
Possibly related, I think you should make it clearer that you are not providing the average access control job size. That number would be determined by adding up all the doors done in all jobs and dividing by the total number of jobs.
Since you are in effect, averaging averages, you lose any weighting associated with the distribution of jobs per integrator.