I do not know how to conduct a site-survey or "Cut, Copy & Paste Specifications"

If you believe the technical specifications you read in majority of the large CCTV Tenders released over the past few years then you can be sure of the following -

1. Every area - is identical in every aspect
2. Lighting conditions - are all the same 
3. Image Requirements for Video Analytics - are all the same
4. Differences between "Detection, Observation, Recognition and Identification" do not exist

(If you answered yes to all the four points above then I don't think you should read further) 

Lets start with a basics - There are a number of different CCTV form-factors (Dome, Box) available with different lenses (Standard, Telephoto, Wide-Angle) for different applications (Detect, Identify). 

The first step in deploying an 'effective' security solution is understanding your operational or security requirements then converting them into the design requirements.

I am referring to RFPs that come out with specs such as - 

Dome PTZ Camera (Day/Night) with 20x Zoom - 129 Nos.
Bullet Fixed IP (Day/Night) with with 4.7mm to 84.6mm Zoom Lens - 237 Nos.
Vandal Fixed Dome (Day/Night) 3mm to 9mm Lens - 198 Nos. 

The above are common in RFPs for Airports, Railway Stations, Multi-Location/Pan-India sites, Hospitals, Malls, Refineries, etc. This usually shows that whoever designed the specs has not put any thought into effectively use the different available products for the right applications.

Take an airport for example - If you plan to use the same lens for all the areas of operation then I can guarantee that the camera count will increase if you want 100% coverage. Different areas have different requirements, I would mix-and-match cameras with different field-of-views so I can reduce the camera count while maintaining 100% coverage 

- At the Entry/Exit areas I would use a Wide Angle Lens to get an overview of the passenger flow. Back that up with a PTZ to zoom in where i suspect there could be a issue

- On the gates and check-in counters I would use a Lens focused specifically towards a narrow area to get a clearer image of the passengers face for recognition

- On the boarding gates I would want to see the entire area (to detect possible congestion) while ensuring I have a clear image of the passengers face

- At the immigration counter I will have a lens suitable for identification purposes

The same has to be interpreted clearly in the RFP so as a vendor it is easy for me to relate to the operational requirements while proposing the system. I have heard of instances whereby the RFP mentions the type and count of cameras required while saying "100% Coverage Required" and during the hand-over some poor vendor will have to forcibly supply a few extra cameras to meet the compliance or the case where the Installer has to constantly move the cameras around because the Operators were not happy with their positions (strangely operators or the actual users are never consulted during the RFP design) 

Its a good idea to have a top down approach from "whats my operational or security requirement", "what do I want to achieve when I pull up my recorded video or image", "can I reduce the camera count by replacing the lens for an area". 

Your FIRST line of defence starts with a good design.