Comparison · WiFi Surveys (Ekahau)

Predictive vs on-site WiFi survey: which do you need?

WiFi surveys answer different questions at different points in a project. A predictive survey asks how many access points you need and where they should go. An on-site survey checks that answer against the real building. A validation survey checks that the installed network meets its targets.

Ask which fits WiFi Surveys (Ekahau)

Published October 27, 2026

Quick answer

Which should you choose?

Use a predictive survey when the building is not finished or you need an access point count and budget early. Use an on-site survey when construction is difficult or plan data is unreliable. Add a validation survey after installation to prove the result against your targets. Many projects use two, and some use all three.

This page compares the three so you can pick the one that matches your stage, your site and your risk. They are not rivals in most projects. They are steps, and which steps you need depends on how much uncertainty you can accept.

Side by side

Predictive vs on-site WiFi survey: which do you need?
Factor	Predictive survey	On-site survey	Validation survey
Suits	New builds, fit-outs and early budgeting, where the building or the network does not exist yet.	Existing sites with difficult construction, such as warehouses, heritage buildings or factories.	Handover, audits and coverage complaints, once the network is installed.
Best for	New builds, fit-outs and early budgeting	Existing offices, warehouses and sites with difficult construction	Handover, audits and coverage complaints
Needs the building to exist	No	Yes	Yes, with the network installed
Main input	Floor plans and material data	A walk of every area with survey equipment	The installed network and the design targets
What you receive	Estimated access point count, positions and channels	Measured heatmaps, interference findings and confirmed positions	Coverage checked against targets, with a list of fixes
Accuracy depends on	The quality of plans and wall material data	Access to every area and representative conditions on the day	Furniture, people and settings matching normal use
Point in the project	Before purchase and installation	Before installation, or when planned positions are in doubt	After installation and configuration
Time on site	None	Most of the three	Moderate, scales with area walked
Cost direction	Usually the lowest, because there is no site work	Usually the highest, because it takes the most site time	Between the two, scaled by area and testing depth
Main limit	A model, so it can miss what the plans do not show	Needs access everywhere coverage is required	Shows what is wrong but is not a new design

WHEN TO CHOOSE

When to choose each.

Choose Predictive survey when

The building is a plan, a fit-out or a refurbishment that does not exist yet.
You need an access point count and bill of materials to set a budget.
The construction is standard and you have accurate floor plans and material details.
You expect to run a validation survey after installation to confirm the result.

Choose On-site survey when

The site has steel racking, thick masonry, metal surfaces or machinery that make a model unreliable.
Wall materials are unknown and cannot be checked from plans.
A wrong access point count would be expensive to fix, for example in a large warehouse.
You need to find existing interference before you design around it.

Choose Validation survey when

The network is installed and you need evidence that it meets the design.
Users report dead spots, drops or poor roaming.
A client, auditor or regulation asks for documented coverage evidence.
You are handing over a site and want a clear list of any remaining fixes.

What is the difference between a predictive and an on-site WiFi survey?

A predictive survey models coverage in software, and an on-site survey measures it in the real building. The predictive model places virtual access points on a floor plan using wall and material data. It is only as accurate as the data you put in. An on-site survey, often done with a temporary access point on a pole or tripod at the planned position and height, captures how signal really travels through the building and what interference is present.

The trade-off is effort. On-site work is slower and needs access to every area. Better simulation tools have made on-site design surveys less necessary in standard offices, but they still matter where construction is unusual.

Do you need a validation survey as well?

Yes, if the network has to meet a stated target, because only a validation survey proves it. It is done after installation and configuration, and compares measured coverage and performance against the design. It can also reveal installation mistakes, wrong settings or faulty hardware.

If a budget allows only one survey type, validation is often the one that gives the most useful evidence for an existing network. If the network does not exist yet, a predictive survey comes first.

How do you decide which combination to use?

Match the survey to the largest uncertainty in the project. If the uncertainty is quantity and cost, start with predictive. If it is the building, add on-site measurement. If it is the result, finish with validation.

New office with standard construction: predictive, then validation after installation.
Warehouse or factory with metal and racking: predictive for a first budget, on-site to confirm positions, then validation.
Existing network with complaints: validation first, with a redesign only if the findings call for it.
Small site with a few access points: ask whether a light predictive design and a quick check is enough.

The verdict

For most new or standard buildings, a predictive survey followed by a validation survey gives the best balance of cost and certainty. Choose an on-site survey when the building makes a model unreliable or a wrong design would be costly to correct. Skip validation only when nothing depends on proving the result, which is rare.

Sources and further reading

QUESTIONS

Common questions.

Is a predictive WiFi survey accurate?

It can be, when the floor plans and wall material data are correct. It is a model, so it can miss what the plans do not show. A validation survey after installation confirms how the real network performs.

When is an on-site survey worth the extra cost?

When the building has unusual or unknown construction, such as steel racking, thick masonry or heavy machinery, or when a wrong access point count would be costly to fix after installation.

What is an AP-on-a-stick survey?

It is an on-site survey where a powered access point is mounted on a pole or tripod at a planned location and height. The engineer then walks the area and measures how the signal behaves.

Can I do a validation survey without a design survey?

Yes. A validation survey checks an installed network against whatever targets you set, and it is the usual first step when users report coverage problems on an existing network.

Do I need all three surveys?

Not usually. Many projects use predictive and validation. Add on-site measurement only where the building or the cost of a wrong design justifies it.

KEEP READING

Get started

Need engineers on site? Tell us where.

Tell us the sites, the scope and the timeline. We reply with a written plan.

Request Services Browse services