Here is an uncomfortable pattern: most field project failures are not execution failures. They are discovery failures that executed on schedule. The install crew did exactly what the plan said, at a site the plan had never truly met.
Drawings lie. Buildings do not.
Floor plans age. Remodels happen off the record. The "identical" store format has four regional variants nobody documented. Ceiling types, conduit paths, power availability, closet space, mounting surfaces: every one of these is a line item in your budget, and every one of them is a guess until someone qualified stands in the building.
The economics: a survey visit is the cheapest visit in the entire program. Every unknown it retires would have cost multiples more as a mid-install surprise: a second truck roll, an emergency parts order, a blown maintenance window, a wave rescheduled.
What a real survey captures
A useful survey is not a walk-through with photos. It works a structured instrument aligned to the install runbook: exact locations and conditions, distances and pathways, power and connectivity verification, environmental constraints, access and safety rules, and deltas against the assumed scope. The output is not a report that gets filed. It is data that changes the plan: per-site bills of materials, corrected labor estimates, and a readiness picture wave by wave.
The survey is leverage
Surveys convert unknowns into decisions while decisions are still cheap. They let you negotiate scope with facts, sequence waves around real conditions, and set budgets that survive contact with the field. Skipping them does not save the money; it moves the spend downstream and multiplies it.
A simple rule
If the cost of being wrong at a site exceeds the cost of surveying it, survey it. For most multi-site technology work, that math is not close.