Predictive Wi-Fi surveys
Modelled from your drawings before the building exists. A predictive survey gives you an access point count, a layout and a cabling schedule early enough to design them into the fit-out rather than chase them afterwards.
When a predictive survey is the right answer
You cannot measure a building that hasn't been built. A predictive survey models it instead: we take your architectural drawings, apply a wall-material schedule, and simulate how radio will propagate through the space once it exists. The output is a design — access point positions, expected coverage, channel plan, cabling routes and a costed kit list.
It is the right approach in four situations:
- New build. The network needs designing alongside the containment and small power, not bolted on at practical completion when the ceilings are shut.
- Fit-out or refurbishment. Partition layouts are changing, so measuring the current building tells you about a building that is about to stop existing.
- Budgeting and tendering. You need a defensible number for a business case or a tender before committing to a site survey.
- Multi-site rollouts. Modelling a representative sample produces a standard design and a per-site cost that scales, instead of surveying forty branches individually.
What it does and doesn't give you
A predictive model is an estimate — a good one, made with calibrated material data and real design experience, but an estimate. It is honest work to say where its limits are.
Further reading: Predictive vs on-site surveys, honestly compared — when a model is enough and when it is not.
| Predictive gives you | Predictive cannot give you |
|---|---|
| An access point count and layout accurate enough to budget and procure from | The actual attenuation of walls that were built differently to the drawing — which happens more often than anyone admits |
| Coverage and capacity modelling against your intended device fleet | The neighbouring networks that will be sharing your spectrum on the day |
| Cabling and containment routes early enough to build into the fit-out | Non-Wi-Fi interference from plant, machinery or equipment nobody mentioned |
| A channel and power plan as a starting point for commissioning | The effect of stock, racking contents or fittings that arrive after handover |
Which is why, on any project of size, we treat predictive as the first half of the job. The model shapes the budget and the infrastructure; an on-site survey or a validation survey confirms it once the building is real. Anyone who tells you a predictive model alone is enough for a complex site is selling you the cheap half of the work.
How we run one
What we need from you
- Floorplans — CAD preferred, scaled PDFs are fine; scale matters more than format
- Ceiling heights, and any mezzanines, voids or double-height spaces
- Wall construction, or enough of the specification for us to infer it
- What the space will be used for, roughly how many people, and what devices they'll carry
- Any known constraints — listed building, exposed services, aesthetic restrictions on visible access points
What we do with it
We build a calibrated model of the building, assign material attenuation values, place access points to meet the design targets for your device fleet, then iterate: adjust positions, heights, transmit power and channel allocation until coverage, capacity and roaming all hold up. Then we lay the cabling and containment over it, and price the kit.
What you get
The same report as any other survey — coverage plots, AP schedule with mounting detail, channel and power plan, cabling and containment schedule, costed kit list with part numbers, and a plain-English summary. Here's what that looks like.
Common questions
How accurate is a predictive survey?
Accurate enough to procure and budget from, and good enough to build from on straightforward buildings. Its accuracy depends almost entirely on how good the information going in is — scaled drawings and a real wall-material schedule produce a very usable model; a sketch does not. On complex sites we'd always want to confirm it on site before final commissioning.
Can we skip the on-site survey to save money?
Sometimes, and we'll tell you when. A small single-floor office with straightforward construction can reasonably go from predictive design to install to validation. A warehouse, a listed building, a site with heavy machinery, or anywhere with dense neighbouring networks really shouldn't. Skipping the measurement on those sites is where the money gets wasted, not saved.
Do you need to visit at all?
Not for the predictive work itself — it's done from drawings. For an existing building we'd usually want to see it, even briefly, because drawings routinely disagree with reality. For a new build there's nothing to visit yet.
Can you work from our architect's drawings?
Yes, that's the normal starting point. CAD is ideal. If all you have is a PDF we can work from it provided it's to scale, or has a dimension we can scale from. If you have nothing at all, we can measure and draw the building, but that's a separate piece of work.
Will it tell us where the cabling has to go?
Yes — that's one of the main reasons to do it early. Getting containment and Cat6A routes into the fit-out design is far cheaper than retro-fitting them once the ceilings are closed, and it's the single biggest saving a predictive survey delivers on a new build.
Let's find out what your Wi-Fi is actually doing.
Book a survey and you get measured data, a design you can build to, and a number you can budget against — not an opinion.