Migration file formats
Export CSVs from your old system and we read them. The parser is tolerant about column NAMES — it recognises the common synonyms — so in most cases you can upload what your old system gives you without editing anything. This page is the exact list, for when you want to check before exporting, or when a column was not picked up. Before you start CSV only. Export spreadsheets to CSV first — no XLSX, no ZIP. Up to 20 files per migration. Files up to 4 MB go inline; larger ones upload straight to storage, up to 50 MB each. Same parser either way. Any subset. Dentists only, pricing only, or all four — whatever you have. Headers are matched case-insensitively, spaced names fold to their underscore form (`Doctor Name` is read as `doctor_name`), and a column we do not recognise is ignored rather than fatal. The diff page opens with What we read from your files — one line per upload saying what it was read as and how many rows it held. Check that first: an unrecognised header does not stop a file being read, it just gets read as something else, and that line is where you see it. dentists.csv — your doctors and practices `dentist_name` — required. Also accepted: `doctor`, `doctor_name`, `name`, `contact_name`, `customer`. This is the key we match on. `practice_name` — also `practice`, `office`, `office_name`, `company_name`. This becomes the practice the doctor bills to: two doctors at one practice join one account rather than becoming two, and every imported case is addressed to it, so terms, a rate card and an invoice have somewhere to land on day one. Leave it out and each doctor becomes their own practice — right for a solo office, and the board says so when it is not. `email` — also `e-mail`, `contact_email`. Lowercased on import. `phone` — also `telephone` `address_line1` — also `address` `city`, `state` (also `st`), `postal_code` (also `zip`, `zipcode`) `qbo_customer_id` — also `quickbooks_id`, if you already use QuickBooks Doctors are matched on their name, so re-uploading the same export changes nothing. Blank fields on a doctor you already have get filled in; values you have already are never overwritten. materials.csv — your working materials `material_name` — required. Also `material`, `name`. `description` Often you can skip this file: any material named in your pricing CSV is created automatically. pricing.csv — your price book `restoration_type` — required. Also `restoration`, `service`, `product`, `item`, `product/service`. For example crown, bridge, denture. `material_name` — required. Also `material`. `unit_price` — required. Also `price`, `fee`, `amount`, `rate`. `$249`, `249.00` and `1,250` all read correctly. `dentist_name` — also `doctor`. Leave it empty for your standard price; fill it in for a doctor-specific rate. Re-importing a price updates it, which is how you fix a typo in your export and simply run the migration again. cases.csv — your case history `case_number` — required. Also `case#`, `case #`, `case no`, `case no.`, `caseno`, `case id`, `rx#`, `rx #` — `Case #` with the space is what several legacy exports write. This is the key we match on, so a re-run never duplicates a case. `dentist_name` — required. Also `doctor`, `doctor_name`. A doctor who is not in dentists.csv is created from this. `patient_ref` — also `patient`, `patient_name` `restoration_type` (also `restoration`), `material` (also `material_name`), `shade`, `notes` (also `note`) `status` — also `current_stage`. See below. `arch` — also `jaw`. `upper` / `lower` for a per-arch job such as a denture or a night guard. `maxillary` / `mandibular`, `upper arch` / `lower arch`, `upper jaw` / `lower jaw`, `maxilla` / `mandible` and `max` / `mand` all read too; anything else leaves the unit unlabelled rather than guessing a jaw. A job reference that appears twice is one case with two units, not two cases — which is how a two-arch job usually comes out of an old billing export, one line for the upper and one for the lower. Each line becomes a unit on the same case, so the case page shows 2 units with the arch named on each, and the invoice carries one line per unit. `due_date` — also `due`, `date_due`, `ship_date`, `want_date`. Read in whatever format your export writes (1/16/2026, 2026-01-16, “Jan 5, 2026”). US month-first when the date is ambiguous, since 05/06/2026 has to resolve one way; a day above 12 makes the US reading impossible and day-first wins. A blank cell is not an error, it is a case with no due date — but on OPEN cases include it, or you import work nobody knows is late. `received_at` — when the work came in, and the one to include if you are bringing over years of history. Also `date_in`, `date in`, `date_received`, `date received`, `received`, `order_date`, `order date`, `date_entered`, `entry_date`, `in_date`. Without it every imported case is dated the day you import, so your archive has no chronology — nothing can be read or ordered by when it happened. A date in the future is ignored. A bare `date` heading is deliberately NOT read: in a combined export it is as likely to be the invoice date or the appointment, and a wrong date on a decade of history reads as fact. Imported cases land as shipped — history, not work in progress — so you do not need to include every stage change. We translate the common status words: complete, completed, delivered, closed and shipped all become shipped; open, in progress and received become received; anything we do not recognise becomes shipped. If a column was not picked up Rename it to the canonical name in the left-hand column above and upload again. Bad rows are listed on the diff page rather than stopping the import, so you always see a useful preview — and nothing is written until you press Apply migration. Not sure? Send us your export and we will tell you exactly what comes across before you commit to a date. White-glove migration is included.
This page uses JavaScript to render its full interactive content. Enable JavaScript, or continue reading at
https://casetop.io/docs/migration-formats.