Digitplus
Infrastructure

A Pre-Deployment Site Survey Framework for Rolling Out Equipment Across Multiple Branches

A pre-deployment site survey for multi-branch IT rollouts: what to measure at each branch, how to make twenty surveys comparable, and the rule that decides when a branch gets its equipment.

Digitplus Editorial Team9 min read
Abstract blue network mesh of glowing nodes and connecting lines on a dark background

Equipment for a branch rollout is usually specified correctly. The branches are where it goes wrong.

A switch arrives at a site where the only free socket shares a circuit with the air conditioner. The cabinet does not fit up the stairs. The "comms room" is a store with no ventilation. Each of these is cheap to find with a tape measure and a meter, and expensive to find with the delivery van outside. A pre-deployment site survey is the visit that finds them first.

This piece is about running that survey across many branches, so the results line up and decide what ships where. If you are fitting out one new office, our network installation checklist for a new office is the better starting point. The problem here is different. It is twenty branches, surveyed by different people, which all have to be compared on one sheet.

Why a multi-branch survey needs one form

A survey of one site can live in an engineer's notebook. Across many branches, that fails.

Two surveyors visit two branches. One writes "power OK". The other logs 198 V at 11am and 231 V at 4pm. Nobody can compare those two records, and only one of them is useful.

So the survey is a fixed form. The same questions, in the same units, with the same photos, at every branch. Any field a surveyor can answer with "fine" is badly written. Rewrite it as a measurement or a yes or no.

The time a survey takes depends on the size of the branch and how far it is from the last one. A small branch with one comms point can be done in a few hours on site. Set that against a wasted installation trip, which costs a second mobilisation of the installers and a go-live date the branch was already told about.

Power and earthing: measure, do not ask

Ask a branch manager whether the power is fine and the answer will be yes. The lights are on.

Measure instead. At each branch, record:

  • Voltage over a working day, from a logging meter left in place. One reading at a convenient moment hides the sag when the building's load peaks and the spike at changeover.
  • The circuit the IT load will sit on, and what else shares it. An air conditioner or a kitchen on the same circuit as the network equipment is a finding.
  • The phase arrangement at the distribution board, and which phase feeds the IT circuit.
  • Generator and changeover. The generator's rating, whether changeover is automatic or someone walks out to switch it, and how long the gap lasts. Time it with a stopwatch.
  • Existing UPS units. Rating, battery age and what is already plugged in.
  • Neutral-to-earth voltage at the socket the equipment will use, and an earth resistance reading taken by an electrician.

Earthing gets its own line because a bad earth fails quietly. The damage shows up months later as faults nobody traces back to the building. Some equipment makers state a maximum neutral-to-earth voltage in their installation guides, often around 2 V. Check the figure for the equipment you are shipping and record the reading against it. Where the earth needs work, that is an electrician's job, and it finishes before the equipment arrives.

These readings size the power protection. A UPS chosen from the survey is sized to the load and the changeover gap that were measured. A UPS chosen without one is sized to a guess. The sizing method is in our guide to power protection and UPS planning. We treat cabling, network and power as one system in our infrastructure work, and the survey is where that starts.

Space, environment and the way in

Record exactly where each device will be mounted. "Back office" is not a location. A named wall, a rack position and a photo with a tape measure in the frame are.

Then the conditions that decide whether that spot will work:

  • Free rack or cabinet space in U, with clearance front and back for airflow and for an engineer to stand.
  • Room temperature, and whether any cooling runs at night and at weekends when the branch is closed.
  • Dust and water risk. A store room under a washroom is a finding.
  • Whether the room locks, and who holds the key.

The route in matters as much as the room. Note road access for the delivery vehicle, whether there is a lift, and the narrowest doorway on the route, measured against the largest box. Note the hours when installers are allowed to work.

Most of all, note who opens the door. At a leased or shared building that is often the landlord's caretaker or building security, not the branch. An installation team that arrives without the person holding the key has lost a day.

Network readiness at each branch

Record what cabling exists, where the drops are and whether it has ever been tested. Untested cabling of unknown age is a remediation line, not an asset. Note the length of the longest planned run as well. A horizontal copper run longer than 90 metres needs an intermediate distribution point, and that changes the design.

For the link out of the branch, record the provider, the link type and whether a second path exists. Then confirm the branch can reach head office systems from the spot where the equipment will sit.

Agree the IP range and the VLANs for each site before anything ships. The installer then configures to a plan rather than inventing one on the day. The design side of this is in our piece on LAN and WAN design for multi-branch businesses.

Turning surveys into a release decision

This is where the survey earns its cost.

Put every branch on one sheet, with a status for each section of the form: power, earthing, space, environment, network and access. Each status takes one of two values. Ready. Or blocked, with a named fix and a named owner.

Equipment is released to a branch when every section reads ready. Not before.

That rule is what makes the survey a control rather than paperwork. It also sorts the rollout for you. Ready branches go first. Blocked branches are scheduled once their fix is signed off, and the fix is usually electrical or cabling work that takes weeks rather than days. Better to learn that from a spreadsheet than from an installer standing in a branch with no earth.

Someone accountable signs each survey before release. On a public-sector job that signature may need to sit alongside the LPO and the procurement approvals. Keep the signed survey in the project file. It becomes the before picture for acceptance testing, and the record if a unit later fails on site.

Sequencing the rollout itself, with a pilot site first and a rollback planned, is covered in how to structure a multi-site IT refresh. The survey sheet is the input to that plan.

Who should carry it out

Two options cover most rollouts.

Your own IT team knows the branches and the business. It often lacks the instruments, and the time to visit every site.

The deploying vendor has the instruments and knows what its equipment needs. It is also checking conditions for its own installation. That conflict is manageable if the form is yours and the readings are recorded raw, with photos, so you can check them yourself.

An independent surveyor is a third route for a very large estate. For most rollouts it is more than you need.

At Digitplus, a site survey is the first step of any infrastructure job, because cabling, power and rack space cannot be specified from a phone call. We document the existing conditions before we design anything. The stages that follow are set out in how an IT engagement with Digitplus works.

Frequently asked questions

What should a pre-deployment site survey record?

Anything about the branch that would change whether the equipment can be installed and run safely on arrival. In practice that means measured power and changeover, earthing, mounting position and rack space, temperature and cooling, existing cabling and the outbound link, and the route in with the person who holds the key. Record measurements and photos, not opinions.

How long does a branch site survey take?

It depends on the branch. A small branch with one comms point takes a few hours on site, plus a logging meter left in place for a working day. On a spread-out estate, travel between branches usually takes more time than the surveys do.

Can branch staff fill in the survey themselves?

Part of it. Photos, room dimensions, opening hours and the name of the keyholder can be recorded by branch staff from a short guide. Power, earthing and cabling need instruments and someone who can read them. Split the form that way and the engineer's visit gets shorter.

What happens when a branch fails the survey?

It does not get equipment yet. The blocker goes on the release sheet with a fix and an owner, and the branch moves later in the sequence. Shipping to a blocked branch anyway puts boxes in a store room. Many warranties start from the invoice or delivery date, so the cover runs down while the unit sits unopened.

Next step

If you are planning a multi-branch rollout, send us the branch list and what is going to each site. We will survey the first branch with you and turn what we find into the form for the rest. See how we run deployment and implementation, or contact us to start.

Related to this: Deployment & Implementation.

Have a project that needs this thinking?

Tell us what you’re planning. We’ll come back with practical next steps and a clear, line-itemised proposal, no obligation.