Organize 360° site photos by the place they were taken, not by the camera that took them: one site per physical location, one capture per visit, the job number in the file name, and every photo placed on a map by its geotag or a GPX track. Share each job as a single link that opens the photos in a browser as a walkthrough, instead of a folder of JPEGs that look warped the moment someone opens them.
That is the whole system. The rest of this post is how to get there from the usual starting point: a folder per job on a shared drive, thousands of files named IMG_20260303_101533_00_041.jpg, and one person who knows where the walk started.
Why do 360° photos get so messy?
A 360° camera makes mess faster than a normal camera, for three reasons.
Volume. A site walk on interval mode produces hundreds of photos per visit; a firm doing regular documentation ends up with thousands a month. Nobody renames those by hand.
The file name says nothing. Job number, site, client and location all live somewhere else: a folder name, someone’s head, or nowhere.
The photo only makes sense in place. A normal photo of a cracked lintel reads on its own. A 360° photo of a plant room is a distorted rectangle until it is in a sphere viewer, and even then you need to know where it was taken.
A folder per job with a subfolder per date works for the person who made it, until a second job lands on the same site, a client asks which photo shows the north boundary in March, or the person who did the walk leaves.
What naming convention works for 360° site photos?
Put the job number first and keep the rest short:
JOB_SITE_YYYYMMDD_SEQ.jpg
24-0187_Riverside-Depot_20260303_041.jpg
Job number first, so files sort by job anywhere and a stray file can still be traced. A short site slug, the same one every time. The date as YYYYMMDD so it sorts. The camera’s sequence number, so the walk order survives.
Batch-rename on import (ExifTool, Adobe Bridge or the camera’s desktop app), after stitching, with a tool that keeps the EXIF data. Stripping geotags and timestamps throws away the two things you will need most.
Two things stay out of the file name: the location, which should come from the geotag or GPX track (“NE-corner” means nothing to someone who has never visited), and the client, who belongs to the job rather than the site.
| What you need to find later | Where it should live | Where it usually ends up |
|---|---|---|
| Job number | The file name, and the capture it belongs to | A folder name only |
| Where the photo was taken | A geotag or GPX track, shown on a map | Nowhere, or in someone’s memory |
| When it was taken | The capture (one per visit) and the EXIF timestamp | A folder called “final_v2” |
| Which site | One site per physical place, reused across jobs | A new client folder per job |
| What was wrong | A note pinned on the photo | A separate spreadsheet of defects |
Organize by site first, then by visit
Jobs come and go; the site stays. The depot you documented for a dilapidation report (a pre-construction condition survey, in US terms) in March is the one you will walk for a progress check in June and a handover in December, possibly for three different clients.
So the top level is the physical site, and under it, one capture per visit in date order, each labeled with its job number. Every visit to the same place then accumulates in one history, and “what did the north boundary look like before works started” means opening the earlier capture. This is the idea behind a site record, and it keeps site knowledge in the firm when the people who did the walks move on.
How do you capture 360° photos so they can be mapped later?
Everything downstream depends on each photo knowing where it was taken. Two ways to get that:
Geotags. Many 360° cameras write a GPS position into the photo’s EXIF data, from a built-in receiver or the paired phone. It works well in the open, poorly under trees and beside tall buildings, and not at all indoors. Check before relying on it:
exiftool -gpslatitude -gpslongitude -datetimeoriginal photo.jpg
A GPX track. Record a track on a phone app, a handheld GPS or a watch while you walk, then pair each photo to the track by capture time. It usually beats the camera’s own fix and fills the gaps. The one thing that ruins it is the clock: GPX tracks are in UTC, and many cameras write local time with no time zone, so a wrong zone puts every photo hours off the track. Set the camera clock from the phone before each walk.
Two capture habits help. Walk a loop at a steady interval (every 5 to 10 meters outdoors, each doorway and the middle of each room indoors), same height, always forward, so consecutive photos link. And capture the whole site every visit, including what has not changed; the unchanged parts are the evidence in a dispute.
For the capture itself, see how to capture 360° panoramas with a DJI drone for aerial work and how to stitch 360° photos for ground cameras and panoramic heads.
How do you put 360° site photos on a map?
Once the photos have positions, the options are:
- GIS. QGIS can import geotagged photos as points. Fine for a one-off, awkward at thousands, and the client needs QGIS too.
- Tour builders. Good walkthroughs, but another subscription, and the tour lives apart from the drawings and scans for the same site.
- A site platform. Upload the photos and the GPX track against the site; they land on the map and connect into a walkthrough beside everything else from that site.
At volume, two things matter: clustering, so a thousand markers collapse into a readable map until you zoom in, and color coding by visit, so the March and June walks can be told apart at a glance.
Photos in sequence become a street-view-style walk: click forward to the next photo, with a minimap showing where you stand. For a client who never visits, that is the difference between “here are 400 photos” and “walk the site from your desk”. The virtual tours page shows what this looks like.
Pair the photos with the drawings, the orthomosaic and the scan
A 360° photo is evidence; a drawing, an orthomosaic or a scan is the reference it is checked against. Keeping them apart is why a site walk so often ends in an email saying “see attached photos 12 to 19 and compare with sheet A-201”.
Put them on the same map instead. 360° markers over the orthomosaic show what the ground looked like at each point that day. A DWG on the same map lets someone step from the drawing to the photo of that spot. An E57 scan with embedded 360° photos can show each photo beside the point cloud. Measurements belong on the scan and the orthomosaic; a 360° photo is for looking and noting.
One survey firm moved its 360° site documentation off DroneDeploy this year, mainly over cost and flexibility, and now puts thousands of 360° photos a month on a map by job number next to its DWGs, one link per job. Nothing about the capture changed; a photo is now found by where it was taken rather than what the camera called it.
Add notes on the 360° photo itself
The last piece of context is what you saw. A defects spreadsheet or a captioned photo appendix detaches the observation from the photo within a week.
Pin the note on the photo instead. In a sphere viewer a note sits in a direction within the panorama (“cracking above the lintel, monitor”) with a color and icon, attachments, a teammate mentioned and the item assigned. For marking up, a sketch captures the view and lets you draw on it. Notes group into a list (a snag list, an inspection round), export as a PDF when someone insists on a report, or share as a walkthrough link that steps through each note.
A note on a 360 stays on that photo. If you also capture an orthomosaic or a scan of the site, notes pinned there follow the place rather than the file: a note made on the March capture shows on the June capture at the same spot, labeled with where and when it was made, so a defect raised at one visit gets checked at the next. The notes page covers how this works across file types.
Environmental consultancies use the same approach for walkover surveys: 360° photos and notes on a map of the site instead of a photo appendix at the back of the report.
How do you share 360° photos with clients?
Three ways not to do it. Emailing the JPEGs: the client opens a warped rectangle and asks what it is. A Dropbox or Drive link: same problem, no sphere viewer, plus a multi-gigabyte download. Google Street View: it renders, but the photos go to Google and can end up public.
One link per job is the shape that works. It opens in a browser with nothing to install, shows the walkthrough on the site map beside the drawings and the scan, and carries your logo and accent color. Protect it with a password or invited email addresses. Keep it editable after sending: add the follow-up visit and the final report, or let new uploads appear automatically. Invited people can reply to notes after verifying their address with a code, with no account to create and no seat to pay for. You can see when the link was viewed, and the client can approve the delivery or request changes from it, which is often how sign-off happens. The share links page has the details.
A simple system you can start this week
- Set the naming convention and rename on import: job number, site slug, date, sequence.
- Capture with a position: camera GPS on, GPX track recording, camera clock synced.
- Walk a loop at a steady interval, same height, always forward.
- One site per physical place, one capture per visit labeled with the job number.
- Upload the photos and the GPX track together, check the map, fix any stragglers.
- Add the reference files: DWG, orthomosaic, scan, report.
- Pin notes on the photos for anything worth coming back to.
- Share one link per job, and reuse it for every visit on that job.
Where Swyvl fits
Swyvl is built around this structure: sites on a map, captures in date order, every upload attached to a site. Geotagged 360° photos land on the site map automatically; upload a GPX track and they are placed along the route by capture time, with the track taking priority over a weak camera fix. Connected 360° photos become a street-view-style walkthrough with a synced minimap, markers cluster, and collections can be color-coded. Point clouds, orthomosaics, DWG, IFC, XGRIDS scans, video and PDFs from the same site open in the browser beside the photos.
Firms already upload thousands of 360° photos a month this way, numbered by job. The free plan covers 5 GB, 3 sites and 5 share links, enough to run one real job end to end. Paid plans start at $49 a month, and files are stored in the region you choose.