Field Notes
Build vs. buy, with the arithmetic shown
A five-part series on what a residential site survey actually costs when you run your own crew, and where the crossover against outsourcing really sits. Every model on these pages is editable. Put your own numbers in and argue with the result.
The algorithm no one is running
Every residential installer knows what their survey team costs. Almost none of them know what a completed survey costs. Build the number line by line, then run it on your own inputs.
Utilization is the whole game
A real Tuesday, minute by minute: under a quarter of the paid day is spent capturing site data. Every fixed cost in the business divides by that number.
The break-even headcount
Demand is a dial; headcount is a switch. Why in-house cost per survey sawtooths instead of falling, and where the crossover actually sits.
The hybrid crew
Almost nobody should be all in-house or all outsourced. Size the crew to your worst month, buy the peak, and the staircase stops hurting.
Two technicians, two roofs
Same pay, same jobs a day, ten points of difference in rework rate. What that gap costs per year, and the capture spec that closes it.
Latest notes
New field notes, newest first. What each kind of site actually needs from the people who measure it.
How to get a solar permit in Houston
City application, engineering evidence, CenterPoint PTO and current incentive checks, with official sources.
Remote vs onsite solar site survey: what needs a visit
Which decisions remote evidence can support, what imagery cannot establish, and how to request the missing field data before committing a design.
Solar site survey tools: the field kit, by what it measures
Every tool on a survey truck, sorted by what it measures and the error it prevents, with real field captures and an interactive kit board.
Solar site survey software: a buyer's checklist
Seven tests for any solar site survey app, from offline capture to data ownership, with a trial step for each and a score-an-app gauge.
AI tools for solar site assessment: what they see, and what they can't
What imagery-based AI reliably sees versus what still needs someone on site, with a comparison table and an 8-point checklist for judging any AI survey tool.
What a residential solar site survey actually needs
The necessary minimum from a residential visit, sorted by who reads it: designer, AHJ, utility and structural engineer. With real field captures from our surveys.
What a commercial solar site survey needs
Roof, structure, electrical room, site and handoff. Most of it is invisible from above, and a miss costs an engineering cycle rather than a truck roll.
Utility-scale sites: prepare, monitor and attack the build
What to measure before civil design, every week of the build, and at commissioning. A three-phase playbook for moving crews down the critical path.
Reference notes
Standalone pieces, not part of the series above. The capture specs and permitting notes we get asked for most often.
The site survey checklist, built backwards from rework
Most checklists list what a survey contains. This one lists what the office calls you about three days later. Roof, shading, electrical, staging, and the on-site review step most crews skip.
What a battery adds to the site survey
Storage is not the panel survey plus one photo. A second electrical investigation, a dimensional rulebook, and often a second authority with an opinion.
How long does a solar site survey take?
Ninety minutes on the roof. Two to four weeks until engineering can start. Only one of those numbers is worth managing, and it is not the one people ask about.
Do solar panels need a permit?
Yes, and it is not the same approval as utility interconnection. What goes in the package sheet by sheet, and the five capture failures behind most rejections.
How to read this series
The models are not a quote and they are not benchmarks. Every default is a plausible starting assumption for a mid-size residential installer, chosen so the arithmetic is legible, and every one of them is wrong for your company in some direction. The point is the structure of the calculation, not our numbers.
Nothing you type is sent anywhere. The models run in your browser and keep nothing.
They are also deliberately narrow. Cost per survey is not the only thing that decides this question, and it is usually not the most important one. Turnaround, coverage, capture quality and what happens when your volume triples all matter more, and only the last part of the series touches any of them.
Run it against your own volume
One site, fixed fee, defined deliverable. We will put a real number next to yours before anybody talks about a contract.
Book a working session