What a WiFi survey actually involves

Someone has told you that a survey comes before a quote, and you’d like to know whether that is a real step or a polite way of getting through your front door. Fair enough. From the outside it looks like an hour of a stranger walking around your flat holding a phone, after which a number appears.

That hour decides whether the finished job works. It is worth knowing what is in it.

I measure where you actually use the internet

Coverage problems do not show up in the middle of a room. They show up in the corner of the bedroom where the headboard is, at the desk pushed against the outside wall, on the terrace. Those are the points I take readings at. The middle of the living room is the easiest place in the house to get a good number, and it is the place nobody sits.

So the survey starts with questions rather than equipment. Where do you work from. Which room has the TV. Where do the school calls happen. Which corner does your phone give up in, every single time. Then I measure at those points, standing where you would stand, at about the height a laptop or a phone actually sits. A body in the way changes the reading; so does a metre of height.

The level I am looking for is −67 dBm. That is the standard WLAN design threshold for voice and video. Above it, calls hold and streams behave. Below it, you are relying on a patient device and a quiet moment.

Signal on its own tells you almost nothing

Your phone shows you how loud the signal is. It does not show you how much else is shouting at the same time. What decides whether a video plays is not the strength of the signal but how far it sits above the noise underneath it. That gap is the part nobody looks at.

Two flats can show identical bars and behave nothing alike. One sits on a quiet floor. The other is stacked above and below with neighbouring networks, so the same strength of signal lands on top of a racket and every frame waits its turn behind the rest of the building. One works. One stalls at nine in the evening when the whole building is streaming.

The fast numbers are the fragile ones. The headline speed on the box comes from the widest channel, and a wide channel needs a much better margin above the noise before it will run at that rate. Two rooms away the link drops quietly to a narrower channel and a lower rate, with your phone still showing full bars. So I measure interference room by room alongside signal. That is the half of the survey most people have never seen done.

The building decides more than the floor plan does

An indoor router in India is allowed 100 mW EIRP. That leaves roughly 87 dB of link budget, and five metres of air plus one reinforced concrete wall eats most of it. So I look at what the walls are made of and which one has a wardrobe or a mirror backing onto it. In a duplex I look at the staircase, because signal doesn't turn corners through a slab.

I also check where the line enters. In most homes the router sits wherever the cable happened to arrive — a utility corner, a shoe cupboard by the front door. Nobody chose that spot for coverage. It causes more dead rooms than any hardware fault I see.

One router covers about 950 sq ft in a concrete Indian home. That is a rule of thumb, not a measurement, and your walls and your layout will move it a long way either way. The survey is what turns that rule of thumb into a number for your home. Terraces, gardens and outdoor staircases get treated as their own problem, because an indoor unit leaking through a door is not a plan.

If the walls are not closed yet, this happens on a drawing instead — see the pages for architects and interior designers.

Your existing hardware, and your neighbours’

I look at what you already own first: the router your provider supplied, where it sits, and whether it is quietly doing something unhelpful. If there is a booster, a range extender or a plug-in repeater in the chain, I will say so plainly. It re-broadcasts a signal that was already weak, and because it receives and repeats every frame on the same radio it spends the airtime twice, which is what halves your bandwidth. It also leaves a second network your phone refuses to let go of. Plenty of jobs turn out to be one or two well-placed access points behind a router that is perfectly adequate.

Then I scan what is on the air around you. In a Mumbai tower the 2.4 GHz band is usually shared with the whole building, and it is worth knowing which channels are already stacked before anyone picks one for you. The channel plan gets built from that scan.

Last, I ask how many things connect. Phones, laptops, TVs, cameras, speakers, the robot vacuum, every smart bulb. Device count changes how many access points you need and how they are tuned. It explains a lot of homes that measure reasonably and still behave badly.

What comes out of it

A written plan. How many access points, exactly where each one goes — which ceiling, which wall, which corner — and how each one is fed. Wherever possible that means wired backhaul, which is the single biggest difference between real coverage and a chain of repeaters fighting each other. It includes the cable routes those runs take, and a hardware list: TP-Link Omada EAP ceiling units, TP-Link Deco, Ruijie Reyee or Ubiquiti UniFi, chosen for the home rather than for the invoice.

With it comes one fixed price covering hardware and labour. Nothing is bought until you say yes. What the installation involves follows from the plan, but all of it is decided in that first hour.

How long it takes

About an hour in a flat. Longer in a duplex or a villa, with more levels, more walls and an outdoor area to cover. It is not a walk-through: measurements at named points, and a scan of what is on the air.

Sometimes the answer is that you do not need me to install anything

This happens often enough to be worth saying out loud. Some homes come out of the survey needing the existing router moved two metres and retuned, and that is the whole job. Some need a channel changed and the transmit power brought down, because a router shouting at full power is still deaf to a phone answering from the far end of the flat.

And sometimes the line really is the bottleneck. Speed arrives at your router; WiFi is what carries it to the room you’re actually sitting in, and no amount of hardware invents bandwidth that was never delivered. Adding access points to a slow line gives you a slow connection, evenly distributed. If that is what the measurements say, I'll tell you at the survey instead of selling you equipment you don't need.

If you would rather know what your home measures than what it ought to measure, that’s the hour. Send me the size and layout, and the rooms where it stops working.