Exporting your data¶
Everything you send LumaTrack, and everything we compute from it, belongs to you. You can take all of it out at any time, on any plan including Free, without asking us and without paying for the privilege. This page lists every way out.
Two rules we hold ourselves to:
- No export is gated behind cancellation, a support ticket, or a fee. The downloads below are ordinary links in the app, live from the day you sign up until the day your account is deleted.
- The formats are ordinary CSV and JSON. Not a LumaTrack archive format. A spreadsheet, a warehouse loader, or a Python script opens them the day after you stop paying us, with no LumaTrack involved.
Self-serve downloads¶
Each of these is a link in the app. They stream, so a large organization does not wait on a buffered file.
| What | Where in the app | URL | Available on |
|---|---|---|---|
| The value ledger, one row per entry with its assumption fingerprint | Ledger, Export CSV | /ledger/export.csv |
Every plan |
| Run events, with status, timing, source, and failure or skip reason | Runs, Export CSV | /runs/export.csv |
Every plan |
| Automations and their baselines, rates, and cost components | Automations, Export CSV | /automations/export.csv |
Every plan |
| Realized value per automation: net, gross, cost, ROI, hours, success rate | Automations, Export value (CSV) | /automations/value.csv |
Every plan |
| Cost lines classed capitalizable vs operating | Ledger, Finance export | /ledger/finance.csv |
Business, Enterprise |
| The amortization schedule, with recognized-to-date and remaining balance | Ledger, Amortization schedule | /ledger/finance.csv?view=schedule |
Business, Enterprise |
| One client's full record as JSON, for MSP offboarding | Clients, client detail, Export data | /msp/clients/<id>/export/ |
MSP |
The ledger and runs exports honor whatever filters the page is showing, so a filtered screen exports exactly what you are looking at. Drop the filters to get everything.
One note on roles: the runs and automations exports need the builder role or higher, because they are the same data those pages already show. The ledger and value exports are open to every member of the organization.
One gap, stated plainly: the incident (loss) ledger has no CSV download of its own yet. Its rows, the raw events behind them, and the event pricing all come out in the organization bundle below.
Over the API¶
The same records read over the API with an organization-scoped key from Settings, API keys. See the API reference for parameters and pagination.
curl -H "Authorization: Bearer lmt_..." \
https://your-lumatrack-host/api/v1/runs?limit=200
GET /api/v1/automationsandGET /api/v1/automations/{slug}GET /api/v1/runsGET /api/v1/event-typesfor the pricing behind your incident ledgerGET /api/v1/summaryfor the headline figuresGET /api/v1/workspaces,/api/v1/webhooks,/api/v1/report-links
On Business and Enterprise, two bulk endpoints exist for warehouse loading and for a complete historical pull:
GET /api/v1/extracts/ledgerstreams every ledger row in stable orderGET /api/v1/extracts/adjustmentsstreams every closed-period adjustment
Both are cursor-paged with a watermark, so you can walk the entire history in one pass and re-run it later to pick up only what changed.
The whole organization, in one file¶
Email support@lumatrack.io and ask for your organization bundle. You get a single JSON file. It contains your organization record and members, workspaces, every automation with all of its runs, the value ledger, the loss and opportunity ledgers, your closed periods and every adjusting entry booked against them, event types and tracked events, initiatives, job roles and their rate cards, categories, shared costs, model price overrides, imported AI spend, saved views, report link history, API key metadata, webhook endpoints, your feedback, and the full audit log. We answer within 30 days, and in practice much faster.
The file joins to itself. Ledger, adjusting, loss, and opportunity rows carry
the id of the automation or event type they belong to, and those objects
are listed with the same id, so you can rebuild the relationships in a
spreadsheet or a load script without guessing.
Credentials are deliberately left out: report-link tokens, saved-view share tokens, and webhook signing secrets. A portability bundle should not double as a set of working keys, so you get the metadata for each and can read the live values in the app.
What survives, and for how long¶
This is the part worth reading before you leave, because the two halves of your history age differently.
The ledger is permanent. When a month closes, its value is frozen into ledger entries and stays there. Closed months are checksummed against the day they closed, so the figures you export in a year are the figures your CFO signed off on.
Raw run events age out with your plan's retention window: 90 days on Free, 13 months on Team and MSP, 3 years on Business, unlimited on Enterprise. A daily job closes and materializes any eligible month before pruning, so value is never lost when evidence ages out. But the run-level detail behind an old month is genuinely gone once it passes the window. If you move to a shorter window, a 30-day grace holds the old one first (see If you cancel).
If you want the run-level record and not just the priced result, export
/runs/export.csv while it is inside your window. That is the one export
that gets harder to take later.
If you cancel¶
Cancelling a paid plan moves you to Free at the end of the paid period. Your account stays open, your numbers stay visible, and every export on this page keeps working. The account does not expire and the ledger is permanent, so there is no deadline to come back and pull your data.
Run events are the one thing with a clock, and we give you 30 days on it. Retention normally follows your current plan, which would mean a Business org dropping to Free went from a 3-year window to a 90-day one overnight. It does not work that way: a downgrade opens a 30-day retention grace, during which you keep the window of the plan you left. Leaving never destroys evidence faster than staying would have.
So the sequence is:
- You cancel. Nothing is deleted.
- For 30 days your old retention window still applies. Settings, Plan and usage shows the held window and the date it ends.
- After that, retention follows the Free window and run events older than 90 days are removed on the nightly job.
The ledger keeps the priced result throughout. If you want the run-level
detail too, download /runs/export.csv inside those 30 days. Before you
cancel, Settings tells you exactly how many run events that is.
If you want the organization removed entirely, ask us. We delete your organization's data within 30 days and backups age out within 90, as the Privacy Policy says. Deletion is irreversible, so export first: the ledger CSV, the runs CSV, and the automations CSV together reproduce everything the app was showing you.