The Biggest APoS Survey Mistake: Surveying Half a Floor at a Time

The mistake: surveying half a floor now, the rest "in a later phase"

An AP-on-a-Stick (APoS) survey is only as good as the area it covers. The single most damaging mistake we see — and one we have made ourselves on a handful of early engagements — is allowing a site to be surveyed in fragments: the client gives access to one half of a floor today, with the other half occupied, under fit-out, or "scheduled for phase 2." It feels reasonable. It is not. A WiFi floor is one continuous RF system, and a design built on half the picture is almost always wrong.

This guide explains why partial-floor APoS surveys fail, using lessons from real projects, and sets out exactly what to do when full access genuinely isn't available on day one.

What happened — a field case

On a multi-tenant office fit-out, we were given access to roughly half a floor; the remaining half was still a bare shell awaiting a second-phase tenant. We placed APs on the stick across the accessible half, measured clean coverage, and produced a design. Weeks later, when the second half was built and we returned to survey it, the AP positions and channel plan from phase 1 no longer held: the new partition walls, racking and glass changed propagation at the boundary, two phase-1 APs now sat too close to phase-2 APs on the same channel, and roaming across the middle of the floor had never been measured at all. The result was rework, a second mobilisation, and a redesign that should have been done once.

Why a partial-floor APoS survey produces a wrong design

• RF doesn't respect phase boundaries. Coverage, co-channel interference and cell overlap behave as one continuous system across the whole floor. APs chosen for the surveyed half collide with APs added later in the other half — interference you never measured because both halves were never live together.
• An empty half reads optimistically. A bare or under-fit-out area has no furniture, racking, partitions or people. Attenuation is far lower than the finished, occupied space, so signal looks better than it ever will in production.
• The AP count and channel plan can't be finalised. Bill of materials, channel and power plans derived from half the floor are provisional at best — you cannot commit procurement to them.
• Roaming across the boundary is untested. The busiest walking paths often cross the exact line between phases, and roaming handoff there is precisely what a partial survey can't validate.
• Stick positions may conflict with phase-2 build. Mounting points chosen against today's empty shell can clash with walls, ducting or furniture that don't exist yet.
• It costs more. A second mobilisation, re-surveying overlap areas and redesigning is far more expensive than one full-floor survey done correctly.

How to do it right

1. Insist on full-floor access in one mobilisation before finalising any bill of materials. Make this a condition in the survey scope.
2. If access must be phased, treat the design as provisional. Say so explicitly in the report, and don't let procurement lock in AP counts from partial data.
3. Plan channels and power holistically for the entire floor — never per-fragment — so phase-1 and phase-2 APs are reconciled on paper before either is installed.
4. Place boundary APs conservatively and leave channel headroom for the APs that the unsurveyed area will need.
5. Document the access constraint and every assumption made about the unsurveyed area's construction and occupancy.
6. Schedule a full validation survey (active and passive) once the whole floor is built out and occupied — and confirm roaming across the former phase boundary.

Other common APoS mistakes worth avoiding

• Surveying an empty building — no people or furniture means optimistic, unrealistic readings.
• Wrong AP on the stick — using a survey AP that doesn't match the antenna pattern and band behaviour of the AP you'll actually deploy.
• Wrong mounting height or orientation — measuring at a convenient height rather than the real ceiling/mount height.
• Skipping spectrum analysis — missing non-WiFi interference that a predictive model and a coverage-only walk can't see.
• Skipping post-deployment validation — never confirming the installed network meets the design targets.

Getting the survey right the first time is far cheaper than redesigning after install. We deliver APoS (AP-on-a-Stick) surveys and Ekahau wireless site surveys across India — and internationally — with full-floor methodology, genuine Cisco WiFi 6 APs and ECSE-certified engineers. For a tailored proposal or to discuss a phased site, use the contact options below.