Wireless Site Surveys: What They Are and Why Guessing AP Placement Costs You

Most bad Wi-Fi wasn’t installed badly. It was planned badly — usually by not being planned at all. Someone counted the access points at the old office, bought the same number for the new one, and mounted them wherever the ceiling tiles were easiest to reach. Six months later there are dead spots in the conference rooms, dropped calls in the warehouse, and a vague sense that the network “just isn’t good here.”

A wireless site survey is how you avoid that outcome. It’s a structured process for figuring out where access points actually need to go in your building — with your walls, your racking, your neighbors’ networks — before anyone runs cable or drills a hole. Here’s what surveys involve, what skipping one really costs, and when your business needs which kind.

What a wireless site survey actually is

At its core, a site survey answers three questions: where do you need coverage, what’s in the way, and where should the access points go to deliver it. Surveyors use software and a measurement kit to map radio frequency (RF) behavior throughout the space — signal strength, noise, interference from other networks and equipment — and turn that into a placement plan with real data behind it.

There are three common types, and they’re used at different stages of a project:

Predictive surveys

Done in software before anyone visits the site. The engineer imports your floor plan, marks wall materials (drywall, concrete, glass, metal), and the modeling tool estimates coverage for candidate AP locations. Predictive surveys are fast and inexpensive, and they’re a good starting point for new construction where the building may not be walkable yet. Their weakness: the model is only as good as the drawings, and drawings rarely show the stainless-steel shelving, coolers, or machinery that eat Wi-Fi signals in real life.

On-site (AP-on-a-stick) surveys

An engineer brings a real access point on a tripod, places it where the plan says an AP should go, and walks the floor measuring what the signal actually does. This is the gold standard for tricky environments — warehouses with high racking, manufacturing floors, buildings with old plaster-and-lath or concrete walls. It validates the predictive model against reality before you commit to cabling.

Validation surveys

Done after installation. The engineer walks the finished space and confirms the deployed network hits the coverage and performance targets — and produces a report showing it. This is your proof that you got what you paid for, and your baseline for troubleshooting later.

What guessing actually costs

The instinct to skip the survey is understandable — it looks like a line item you can cut. But the failure modes of guessed AP placement all cost more than the survey would have:

• Re-cabling. Every access point needs a cable run back to the switch. If an AP is in the wrong spot, moving it means pulling a new run — lift rental, labor, ceiling work — often at a higher price than doing it right during the original install.
• Buying more APs than you need. The common panic response to bad coverage is adding more access points. But too many APs on overlapping channels interfere with each other, so you spend more money to make the network worse. This is one of the most frequent problems we see in offices that grew their Wi-Fi piecemeal.
• Dead zones where work happens. Coverage maps drawn by guesswork tend to be strongest in hallways and open areas — and weakest in the corner conference room, the shipping office behind the racking, or the far end of the production floor. Exactly where people need it.
• Roaming problems that look like random glitches. Voice calls that drop when someone walks down a hallway, scanners that lose connection between aisles, laptops that cling to a distant AP — these are placement and power problems, and they’re hard to diagnose after the fact without survey data.
• Troubleshooting without a baseline. When Wi-Fi problems show up later, the first question is “what did the RF environment look like when it worked?” Without a validation report, nobody knows, and every fix is another guess.

What actually blocks Wi-Fi in commercial buildings

Part of why guessing fails is that signal behavior is genuinely unintuitive. A few things that routinely surprise building owners:

• Metal is the big one. Steel racking, shelving, ductwork, coolers and machinery reflect and block signal. A warehouse aisle can have great coverage while the next aisle over has almost none. (We covered this in depth in our post on warehouse Wi-Fi and racking.)
• Inventory changes the map. A racking bay full of palletized product behaves differently from an empty one. Surveys of warehouses should happen when stock levels look like normal operations.
• Older Chicago-area buildings punch above their weight. Brick, concrete block, plaster with metal lath — common in older industrial and commercial stock around Chicagoland — attenuate signal far more than modern drywall. A layout that works in a new tilt-up won’t behave the same in a 1940s masonry building.
• Your neighbors’ networks count. In multi-tenant buildings and dense industrial parks, adjacent businesses’ Wi-Fi occupies the same airspace. A survey measures that interference; a floor plan doesn’t show it.
• Glass and open atriums. Signal travels through open vertical spaces in ways that create weird coverage overlaps between floors — sometimes your third-floor devices connect to a second-floor AP.

When you need a survey (and which kind)

Not every project needs the full treatment. A rough guide from how we approach it:

• Small office, simple layout, standard construction: a careful predictive design is usually enough, validated after install.
• New office buildout or relocation: predictive survey during planning — ideally while cabling is being designed, so AP drops go in with the rest of the structured cabling instead of as an afterthought.
• Warehouse, manufacturing, or any space with heavy metal and racking: on-site survey. This is the environment where predictive models miss the most.
• Voice over Wi-Fi, barcode scanners, or wireless equipment on the move: on-site survey with roaming requirements defined up front. These applications fail on placement plans that look fine for laptops.
• Existing network with chronic complaints: a diagnostic survey of the current environment usually finds the cause faster — and cheaper — than another round of new hardware.

What you should get out of a survey

A survey isn’t just advice — it’s deliverables. When you pay for one, expect to receive heat maps showing predicted or measured coverage across your floor plan, a specific AP placement plan with mounting locations and channel/power recommendations, a cable drop list your installer can price and pull from, and — for post-install validation — a report demonstrating the network meets the targets. Those documents keep paying off: they become the baseline for troubleshooting, the plan for future expansion, and part of the documentation package for the building.

One more practical note: survey and cabling belong together. The survey determines where APs go; every AP location is a cable run and a switch port, often with PoE power requirements. When one contractor handles both the wireless design and the network installation, the plan and the pathways get built as one project instead of two that have to be reconciled.

The bottom line

Wireless site surveys exist because buildings lie. Floor plans don’t show the racking, the coolers, the plaster walls or the neighbor’s access points — and Wi-Fi lives or dies on exactly those details. For a small office, an hour of predictive design may be all the insurance you need. For a warehouse or plant, an on-site survey is routinely the difference between a network that works on day one and a year of chasing complaints. Either way, the survey is the cheap part. Guessing is what’s expensive.