I formerly worked at a college library where we had these devices. We went through three different units in the 9 years I was there, largely due to malfunctioning. There would be many days where we did not get a count at all, or the counter stopped working, perhaps, at 9 a.m. and we got no count the rest of the day until we were checking the numbers that night and saw, "Hmm, we only had 12 people in here today?" One time it broke and we shipped it to somewhere (Asia?) for repairs. We had no door counter for about 6 weeks then.
Our doors to the library proper is where the counter was, so anyone just visiting the lobby for the circulation department--book returns, etc,, did not get counted. Conversely, anyone in the real library would get counted on entering, on going to the circ area, on re-entering, on going to the washroom in the lobby area, on re-entering, on going outside to make a phone call (some people were polite), on re-entering... We had some summer Saturdays when I would check the number on opening, resetting it to zero, and an hour later, knowing only two people had entered the building, would look at the counter and see, maybe, 24 showing.
That was a nice number for making our reports look good, but if your goal is a useful count of people, you would be far better off to assign someone to go around and count heads every hour or so to tell you how many people are truly using the facility. We did this and used those numbers in planning staffing, etc., far more than the door counter numbers.
Even if your counter only counts people going one way, there will still be numerous re-entries that keep the number highly inaccurate in terms of telling how many folks actually used your library on any given day.
I truly don't think these things to be that useful.
Derek Barth
Bath Township Library