Is it worth paying for a Wifi site survey for a small deployment?

Question

The first thing many people say when deploying a wireless network in a new location is "get a site survey".

For a large deployment involving hundreds of AP's, I can see the wisdom in paying for a site survey as it gives a good baseline for the deployment and saving 10 nodes can more than pay for the survey (especially since they are likely to be expensive "enterprise" quality nodes).

But in a small deployment with < 10 nodes, is a formal site survey worth the money?

I've been quoted prices between $300 - $500 for a small office site survey and for that price, I could just add 3 - 5 additional nodes. (assuming something like the "cheap" Ubiquiti $100 nodes that would be typical in a small office deployment, not $1000+ Cisco or Meraki nodes).

I've overseen 100+ node deployments where I felt the survey was well worth the money, but I've also spent $500 on site survey for 3 node deployments where I felt that the site survey was overkill and I could have added 3 more nodes to cover areas where signal proved to be inadequate (and indeed have added nodes beyond what the site survey said was "optimal"). And that "site survey" consisted of some diagrams showing theoretical (ideal) node coverage along with a walkthrough with a high gain antenna showing potential wifi coverage at a (relative) few points.

Is there some rule of thumb to determine when a site survey is worth the money?

Answer 1

It really depends what your objective is. If you're looking to run VoWifi with its sensitivity to coverage issues, or if you have something very different from a garden-variety office (say an industrial plant with metallic structures every which way) you should probably go for the survey, and find someone competent to do it.

If the site is both small (under 5 to 10 APs) and fairly "standard" I tend to eyeball it based on a good set of plans of the site and some indications as to where people are, and it usually turns out pretty well. We leave some slack in the wiring so we can adjust things if necessary.

I find that the biggest problem tends to be people on site telling me "we need 3 APs" because they are thinking only of signal propagation, but they are forgetting entirely about user density in our world of multi-device always-connected users... so getting the AP count right, and making the site manager agree to an increased AP count and cost, tends to be hardest.

Answer 2

A wireless site survey can tell you about "invisible" problems like multipath or external interference. You need to discover if any neighbors subscribe to the (incorrect) philosophy of cranking radio power to the highest possible level. You could be directly under or over one of these (radio is three dimensional). You will get a good idea of WAP quantity, placement, frequencies, radio power, etc. If you have a large deployment, it is a good idea to have a followup wireless site survey after implementation, under working conditions, in order to tweak these things.