How do you validate a predictive design without fooling yourself?

Wi-Fi Design & Planning
1

Lately, we’ve been doing more and more preliminary Wi-Fi planning because clients want guarantees even before we purchase the hardware.

And that would be fine, but I’m terrified of the classic situation: everything looks perfect on the map, but reality makes its own adjustments (to put it mildly).

For those who regularly do this, tell me: how do you verify your calculations after installing access points? I’m not interested in what’s written in brochures, but rather how you translate theory into practice so as not to screw up with the client.

2

For me, a preliminary estimate is just a working hypothesis, not the ultimate truth. A slight error in the wall model, a miscalculation of the ceiling height, or a misplaced point — and that’s it, the image doesn’t correspond to reality. Power and channel settings can also throw everything out of whack in reality. So my approach is simple: until you walk through the site after installation and measure everything yourself, you can’t trust the estimates.

3

Predictive is for placement and expectation-setting. Validation is a separate job.

4

My usual process:

  1. Predictive design to choose AP count/rough locations

  2. Install

  3. Passive survey to validate RF shape

  4. Active survey in problem workflows if the environment is sensitive

  5. Tune channels/power based on what actually happened

If the client only buys step 1, they’re buying a model, not proof.

5

Makes sense. What do you actually use for this kind of work?

6

NetSpot is worth a look for that. One of the nice things about it is that it basically covers all three parts in one app: Inspector for checking the Wi-Fi environment, Planning for predictive survey, and Survey for validating the real deployment afterward. That’s a pretty sensible workflow, honestly. Check the RF environment first, build the design, then walk the site and see how close reality is to the plan.