Most site visit reports are a photo dump with a timestamp. Forty images in a WhatsApp group, a few voice notes, and an implicit assumption that everyone will work out what needs doing.
Nobody works it out. The photos scroll away, the decisions live in the head of whoever attended, and three weeks later something is wrong that was visible in image nineteen.
A visit produces three things, or it was a walk
Progress against plan. Not "slab work going on" — which activities, at what percentage, against what was scheduled. The useful output is the variance.
Decisions taken on site. A carpenter asked whether the shutter should be 18mm or 25mm, someone answered, and that answer changed the drawing. If it is not recorded, it is not real, and the drawing is now wrong.
Items raised. Snags, obstructions, material shortfalls, coordination clashes — each with an owner and a date.
If a visit produces none of those, it was a walk. Walks are sometimes worth doing. They are not worth reporting.
The structure
Keep it to one screen. A report longer than that gets skimmed, and the important item is always in the part that was skimmed.
- Date, project, who attended — including the contractor's representative
- Progress — activity, planned %, actual %. Three to six lines
- Decisions taken on site — what was asked, what was decided, what drawing it changes
- Items raised — description, photo, owner, due date, severity
- Next visit date
That is the whole thing. The discipline is in decisions taken on site, which is the section that does not exist in most reports and is the one that prevents the expensive failure.
Photograph deliberately
Forty photographs is not thoroughness; it is deferral. It moves the work of understanding from the person who was there to the people who were not.
Take fewer, purposefully:
- Wide shot per room, same position every visit. Comparable across weeks — progress becomes visible rather than asserted
- A close-up per item raised. Attached to the item, not floating in a stream
- Anything that differs from drawing. With something for scale
A photo attached to an item is evidence. A photo in a group chat is a memory that expires.
Log deviations against the drawing they change
This is the highest-value habit on site and the one most consistently skipped.
An actual wall is 40mm off. The carpenter adjusts. The site supervisor knows. The drawing does not, and neither does whoever specifies the appliance three weeks later.
Every deviation should be recorded against the specific drawing or element it affects, with the measurement and a photo. It takes ninety seconds on site and saves a replacement order later. This is the same failure the execution phase section covers in the five phases of an interior design project.
Capture it on site, not that evening
A report written from memory in the evening is materially worse. Detail is gone, measurements are approximate, and the item that seemed minor at 3pm has been dropped by 9pm.
The practical consequence is that your capture method has to work on a phone, in a half-finished flat, with bad signal and dusty hands. If it requires a laptop and connectivity, you will get evening reports, and evening reports are the lossy kind.
Close the loop
Every item raised needs an owner and a due date, and the next visit starts by reviewing the previous visit's open items.
Without that, reports become a write-only archive — diligently produced, never consulted. The review at the top of the next visit is what makes the whole exercise worth doing. It takes four minutes and it is the reason anyone bothers.
Related: snag lists for interior projects.