A WiFi survey London property owners actually find useful is not a man with a phone walking from room to room saying "yeah, the signal drops a bit here." It is a measured, mapped, documented diagnostic — a floor plan with coloured contours that show, in dBm and percentage, exactly where the radio signal is strong, where it is marginal, and where it has fallen off a cliff. I have been doing these surveys across London since 2011 and I can tell you, hand on heart, that the heat-map is the single most useful piece of paper a homeowner or office manager will ever receive about their network. It turns "the WiFi is rubbish upstairs" into evidence — and evidence turns evidence into a plan that lasts.

What a WiFi heat-map survey actually is

A heat-map survey is a structured walk-through of a property with a calibrated tool — typically NetSpot, Ekahau Sidekick, TamoGraph, or one of the equivalent enterprise platforms — running on a laptop or tablet. The floor plan is loaded into the software, scaled correctly to the property's real dimensions, and then I walk the building stopping at marked points roughly every two to three metres. At each stop the software records the radio environment in detail: every visible SSID, the received signal strength in dBm, the signal-to-noise ratio, the channel each access point is using, its 802.11 standard, its bandwidth (20, 40, 80 or 160 MHz), the BSSID, and where applicable the BSS Color value.

When the walk is finished the tool interpolates between those measurement points and overlays the result on the floor plan as a coloured map. Green is strong, yellow is workable, orange is marginal, red is dead. That is the layperson view. Underneath, the same data can be sliced a dozen other ways — 2.4 GHz coverage versus 5 GHz, AP-by-AP coverage, SNR maps, channel overlap maps, predicted data rate maps, and roaming-zone maps that show where a client device should hand off from one access point to another.

The kit, in plain English

For a serious survey I will be carrying one of two setups. For larger commercial work, an Ekahau Sidekick: a purpose-built dual-radio spectrum analyser that captures both bands simultaneously and is calibrated to a known reference, so the numbers it produces are repeatable rather than approximate. For domestic and small-business work, NetSpot or TamoGraph running on a Windows laptop with a known external WiFi adapter — still measured, still mapped, just without the spectrum-analyser side-channel data. Both give you a usable map. The enterprise kit gives you more granular non-WiFi interference detection — microwaves, baby monitors, dodgy Bluetooth audio, the lot.

The walkthrough — how I actually do it

People expect this to take ten minutes. It does not. A real WiFi survey London-wide averages between an hour and three hours of on-site time depending on the building, plus another hour off-site for the report. Here is the order I work in.

  1. Floor plan in, scaled correctly. Without a correctly scaled plan, every distance measurement and every contour is wrong. I will either load an existing plan, draw one quickly from a tape-measure walk, or use an estate agent's floor plan with a correction pass.
  2. Identify the existing radio environment. Before I move I capture a passive sweep — every SSID visible in both bands, signal levels, channels in use, and any obvious interference. In a London terrace this list will usually be twenty to forty neighbouring networks. In a Mayfair mansion block it can be over a hundred.
  3. Walk the floor. Slowly. Pausing at each measurement point for long enough to let the radio settle and the tool capture multiple frames. Through doorways, into bathrooms, into the cupboard under the stairs, onto balconies and into the garden — anywhere a phone, laptop or smart-home device actually lives.
  4. Re-walk for the second band. 2.4 GHz and 5 GHz behave very differently. 2.4 GHz travels further but suffers more interference. 5 GHz is faster, less congested, but stops dead at the second wall.
  5. Capture client-side data. A heat-map of the access point's transmit signal only tells half the story. I also check what a client device — typically a phone — actually negotiates from each position. Some "green" spots on the AP map are actually weak on the client return path because phones have tiny antennas.

What the resulting heat-map shows

A useful report has several maps, not one. Here is what each layer tells you.

  • Signal strength (RSSI). The headline map. -50 dBm is excellent, -65 dBm is good, -75 dBm is marginal, -85 dBm is essentially unusable.
  • Signal-to-noise ratio (SNR). Often more useful than raw signal. A strong signal in a noisy neighbourhood can still feel slow because the radio cannot distinguish your transmission from the background hash. SNR above 25 dB is healthy, below 15 dB is where things start crawling.
  • Channel overlap and co-channel interference. Every access point on the same channel — yours or your neighbour's — has to share airtime. The map shows where your own APs are stepping on each other (a common mistake when people self-install multiple routers) and where the neighbours are crowding you out.
  • Predicted data rate. Not the same as Mbps to your laptop. This is the maximum the radio could negotiate at that point, assuming the connection back to your router is healthy.
  • Roaming boundary. Where the device should hand off from one AP to the next. Get this wrong and your phone will cling to a distant AP at -78 dBm while standing four metres from a closer one at -45 dBm — the classic "WiFi is on but nothing loads" problem.

Pre-install survey vs post-install survey

These are two different exercises, both useful, often confused.

A pre-install survey is done before any kit is changed. It maps the existing network — usually a single router and maybe an extender or two — and identifies where the dead spots are, what is causing them (distance, building materials, channel congestion, interference), and where I would propose putting access points to fix the problem. The output is a placement plan: "AP1 at the hallway ceiling here, AP2 in the second-floor landing here, cabling route via the loft." That plan goes to whoever is pulling the Cat6 cabling — sometimes me, sometimes the client's electrician.

A post-install survey is the verification step. After the access points are mounted, cabled and powered up, I re-walk the property and produce a fresh heat-map showing the actual achieved coverage. This is where you confirm the install met the brief. It is also where, occasionally, you find one access point is mounted six inches from a metal junction box that is killing its rear lobe, and you move it. The post-install map is what the customer keeps in a file — it is the documented proof that the network performs as promised.

When a survey is essential, and when it is overkill

I will not pretend every flat in London needs a heat-map survey. They do not. Honesty matters here.

A survey is essential when:

  • The property is larger than around 120 square metres, or is on three or more floors.
  • There is more than one persistent dead spot that swapping the router has not fixed.
  • It is a business or a working-from-home setup where downtime costs money.
  • It is a listed building, a conversion flat, or anything with unusual construction — solid brick internal walls, lath-and-plaster ceilings, foil-backed insulation, steel beams, period stone.
  • The install will involve three or more access points. Once you are placing that much kit, you want the placement modelled rather than guessed.
  • The customer is replacing a network that has already failed once and they cannot afford to get it wrong twice.

A survey is overkill when:

  • It is a one-bedroom flat with one router and one dead spot, where the answer is obvious from a five-minute look — usually relocating the router or running a single ethernet drop to a sensible second AP location.
  • The fibre install is still pending and the router has not even arrived. Survey after the new line is in, not before.
  • The customer just wants better WiFi in the garden. Often a single outdoor AP solves that without a full survey.

If you ring me and the answer is "you don't need a survey, you need X," I will tell you. That conversation costs nothing.

What the customer actually gets at the end

This is where amateur "surveys" fall down. A proper output is a deliverable, not a verbal summary. The customer receives:

  1. The scaled floor plan with overlay maps — signal strength, SNR, channel coverage, predicted throughput. PDF format, one map per page.
  2. A written summary in plain English explaining what the maps show, what is causing the issues, and what I recommend.
  3. The proposed access point locations marked on the plan, with a cabling route between them and the router.
  4. A bill of materials — what hardware would be installed, what cable would be run, what mounting would be needed.
  5. A copy of the raw survey data, so any other engineer can re-run the analysis on the same numbers in future.

That document is the difference between "we think it will work" and a network design you can hold someone accountable to.

Three real scenarios from the field

Pre-install survey at a Marylebone consultancy

A consultancy with around twenty staff in two adjoining first-floor units near Marylebone High Street wanted reliable WiFi across both units before their next financial year. They had been struggling with one consumer router and a powerline kit that lost the link whenever the kettle was on. Brief: zero dead spots, fast roaming between the units, support for video calls everywhere, a guest network that was properly segregated.

I did a pre-install survey over about ninety minutes on a Saturday morning when the offices were empty. The 2.4 GHz spectrum in that part of W1 was a nightmare — about sixty visible networks on the survey, with channels 1, 6 and 11 all jammed. 5 GHz was much cleaner. The internal walls were solid Victorian brick and the survey showed clearly that one router would never cover both units; the rear unit was running at -78 to -82 dBm from the front-of-house position. The map identified four AP locations: front reception, the main open-plan area, the rear meeting suite, and the back-of-house corridor that served the kitchen and side offices. Cable routes were planned through the existing trunking. The install team came in the following week, ran the Cat6 back to a central switch, and fitted four PoE-fed access points using the proposed locations. The post-install walk confirmed -55 dBm or better in every working space. The consultancy filed both surveys with their facilities documentation.

Post-install survey at a Chelsea family home

A Chelsea family had already had a network installed — by someone else — and the children's bedrooms on the second floor were a problem. The installer claimed three APs would cover the house. They mostly did. But there was one corner of the top-floor master bedroom where Netflix would buffer and a teenage daughter's laptop kept dropping mid-homework. The family wanted to know whether the install had been done correctly or whether they had been short-changed.

I did a post-install survey. The result was nuanced and honest, which is what they deserved. The three APs that had been installed were genuinely well-placed; the heat-map showed strong 5 GHz coverage across roughly 92% of the floor area. The remaining 8% was the corner of the master bedroom and the en-suite — and the cause was clear from the survey: a large mirrored wardrobe lined with what was almost certainly foil-backed plasterboard insulation behind it. The mirrored wardrobe was effectively a Faraday cage for that corner. The fix was modest — one additional small wall-plate AP fed from the existing run in the landing, which I recommended. The survey report exonerated the original installer, identified the real culprit (the wardrobe, not the network), and gave the family a clear next step. They went away knowing exactly what they had and what one more drop would buy them.

Hampstead Heath edge mansion — three floors and a coach house

This one was a full survey job, the kind that really earns its keep. A detached house on the edge of Hampstead Heath, around 480 square metres across three floors, plus a converted coach house in the garden being used as a home office. The owner wanted one network ID seamlessly across the whole site, including the garden between the buildings.

The pre-install survey took most of an afternoon. The house had been through three previous extensions, each with different wall construction — the original Victorian core was solid brick; the kitchen extension was timber-frame; the loft conversion had steel binders running between the gables that absolutely murdered the 5 GHz signal in two diagonal stripes. The garden walk showed that the existing router in the main house gave nothing reliable in the coach house even with line of sight, because of distance and a glazed garden room in the way. We mapped seven AP positions in the main building, one in the coach house, and one outdoor-rated AP on the rear elevation to bridge the garden. The cabling route went via the loft, down the service riser, and through the existing garden ducting to the coach house. Post-install survey confirmed -60 dBm or better everywhere inside both buildings, -68 dBm in the centre of the lawn (good enough for a phone call on the bench), and a single roaming boundary at the side door so devices would hand off cleanly between coach house and main house. The owner filed the survey pack alongside the house deeds.

Why the survey changes the conversation

The reason I push surveys for larger jobs is not that they sell additional hardware. They usually save the customer kit. I have walked into properties where the previous owner had installed five access points and a heat-map showed two of them were actively interfering with each other on overlapping channels, and we removed two and improved coverage. Surveys produce truth. Truth is cheap in the long run.

If you have been told you need a mesh, or three extenders, or a "WiFi 6 upgrade" because that's the latest standard — and nobody has measured your actual building yet — you are buying a guess. A survey is what makes the recommendation accountable. If you want to dig further into the standards side, my piece on WiFi 6 (802.11ax) explained covers what OFDMA and BSS Color actually do for you, and the piece on WiFi signal loss and building materials covers exactly why those Victorian internal walls eat your signal.

Booking a survey — what to expect

If you ring and we agree a survey is the right next step, I will ask for the property size, the number of floors, what you already have installed, what the brief is (residential coverage, business reliability, garden access, smart-home density), and any access constraints (listed building, no drilling certain walls, AV team coordinating). I bring the kit. You don't need to do anything in advance other than have access to all the rooms — including the loft and any cupboards where the router and patch panel live. Most domestic surveys are a single visit. Larger properties or pre/post-install pairs are two visits, scheduled around the cabling work.

And if you want the bigger picture on how this connects to your fibre or copper line, my broadband engineer service page covers how the WAN side fits into the LAN design. Survey-driven networks are how the two sides actually meet.

Ring 020 3633 1131. Honest advice, freely given.