Switching Cabinet Design Software: What Comes With You and What You Rebuild
Nobody leaves a seat they have used for ten years because a feature list looked better. They leave when the cost of staying finally beats the cost of moving, and the cost of moving is the part nobody will give you a straight answer about. So here it is in three columns: what comes across, what you rebuild once, and what does not move at all.
The move at a glance
Your parts and your stock list come across as data you already have in a spreadsheet. Your cabinet library and your construction standard are rebuilt, and for a 32mm shop the standard is an afternoon rather than a season. Your old program's project file does not move at all, here or anywhere, and a vendor who says otherwise is selling you an importer that will break on somebody else's release. The question worth asking before any of it is how you get back out again.
| Moves | Parts, pasted from a spreadsheet. Materials, as a CSV of your real stock. |
|---|---|
| Rebuilt once | Your construction standard, your cabinet types, your hardware list. |
| Does not move | A native project file from any other program. There is no importer for one. |
| The honest cost | An afternoon on the standard, then one job at a time for the library. |
| The order | One live job first, cut it, compare it against your old list. Then decide. |
| Getting out | Three versioned export formats, plus your whole account as one JSON file. |
In this guide
What actually moves
Start with the part of the answer that is short, because short is what makes it believable. Two things come across, and both of them are data you can already open in a spreadsheet today.
Parts, pasted rather than retyped
If your current seat can produce a parts list as a spreadsheet, your parts move. Paste the rows straight in: tab separated out of a spreadsheet, or comma separated if you typed them. The columns are label, width, length, quantity, type and grain, the header row is skipped for you, and a number reads in whatever unit the page is set to unless the cell says otherwise. A cell reading 14 3/4 is read as a fraction, and one reading 18mm is read as millimetres even in an inch job. A blank quantity means one. A quantity with a typo in it rejects that row and names the line, rather than quietly becoming a one.
Material and edgebanding are assigned afterwards in the table rather than in the paste, which is the right way round: the material names in your old system are its names, and mapping them by hand once is faster than fixing a bad automatic guess forty times.
Materials, as a CSV of your real stock
Your stock list moves as a CSV, dropped in or pasted. Column order does not matter, because the headers are matched in any combination rather than by position. Any measurement can carry its own unit and that unit wins, so 3/4", 23/32 and 18mm all read correctly whatever the column happens to be called. A material whose name you already have is skipped rather than overwritten, so running the import twice does not double your library. A row that does not come out as a real board is left out and the message names the line, which is the behaviour you want on an import of six hundred rows: nothing arrives wrong and quiet.
What does not count as a migration path, even though it looks like one
There is an Import parts button that takes a JSON package, and it is worth being precise about what it is, because on a feature list it reads like an escape route and it is not one. That package is our own format, written by our own tools when a job moves from the Designer or from the photo-to-cut-list module into the optimizer. It is a handoff between rooms in the same house. It is not a bridge from another vendor's building, and no amount of hand editing will make it one.
What you rebuild once
Three things, and they are not equal. One is an afternoon, one is a job at a time, and one is a morning of typing that you will be glad you did.
| Your construction standard | An afternoon. It is a short list of decisions you already know by heart. The next section is that list. |
|---|---|
| Your cabinet types | One job at a time. You draw the boxes this kitchen needs, and by the third kitchen you have drawn most of what you build. |
| Your hardware list | A morning of typing. There is no hardware import here, and pretending otherwise would waste your afternoon rather than save it. |
The hardware one deserves the flat sentence rather than a soft one. The Hardware Library holds your hinges, slides and their real specs, and everything in it gets entered by hand, because there is no import for it today. For most shops that is a couple of dozen entries: the hinge you actually buy, the two slide families you actually stock, and the lengths you keep on the shelf. It is a morning. It is not nothing, and you should price it into the decision rather than discover it in week two.
The cabinet library is the one that sounds worst and behaves best. You are not sitting down to recreate two hundred cabinet types before you can draw anything. You draw the kitchen in front of you, and the types you actually build accumulate as you go. A shop that builds mostly frameless boxes on the 32mm system discovers that most of its library was never two hundred types. It was about a dozen, in different widths.
What does not move at all
Your old program's project file does not come across. There is no importer for one here, and this is the sentence that should decide how much you trust the rest of this page, because it is the sentence every vendor is tempted to fudge.
The reason is worth understanding, because it also tells you what to expect from anybody who claims otherwise. A native project file is undocumented on purpose, changes on the vendor's own release schedule, and holds meaning that only their code knows how to interpret. An importer written against one is correct until the week they ship an update, and then it is silently wrong, which is worse than never having existed. A drawing that opens and is subtly incorrect costs a shop far more than a drawing that does not open at all.
So the honest framing is not that your work is trapped. It is that a drawing is not a portable thing, and a parts list is. Move the parts list, rebuild the drawing when that cabinet next comes up, and you will find you have moved the part that had money in it.
One more thing does not move, and it never will, because it is a decision rather than a gap: we do not write g-code, toolpaths or a post for a named machine. A post maintained against the machine dialects in the field never finishes and never gets good. What leaves here is a neutral drawing of the nested sheets that the software you already trust reads.
The shop standard is the real work, and it is smaller than you think
This is the part a switcher underestimates in both directions. It is more important than the cabinet library, and it takes far less time.
Every dimension in a cabinet here comes down a four-step chain. There is a compiled-in floor you cannot edit, your shop standard sits on top of it, a project can differ from your shop standard, and a single cabinet can differ from its project. Set a number once at the shop level and every new job starts there without anybody remembering to. That is what makes the rebuild finite: you are writing down what your shop already does, once, in the place that everything else inherits from.
Here is the whole list, which is shorter than the anxiety about it:
| The box | Frameless or face frame, carcass thickness, door thickness |
|---|---|
| Reveals | Top, bottom, left, right, between adjacent fronts, and a separate side gap on a finished end if you use one |
| The back | Thickness, whether it sits in a dado, how deep, and how far in from the back edge |
| Cleats | How many, how tall, how thick, inside or outside the case, and how far off the reference face |
| Shelves | Setback from the front edge, and total side clearance |
| Drawers | Slide type and clearance per side, box length and height reductions, box material thickness, and the joinery |
| Joinery | How the case parts go together, and for a loose tenon the size and the machine |
| Drawer bottoms | Thickness, captured in a dado or not, depth, and distance up from the bottom edge |
| Fronts | Overlay style, and the per-side tolerance an inset door gets |
| Edgebanding | Which material goes on sides, shelves, doors, applied ends and drawer boxes |
Where these liveYour shop-wide answers are set on the Shop Details page, along with the name and contact that print on your drawings, your pricing defaults, how sizes are rounded, and your label template. A project or a single cabinet can still differ where a job needs it to.
Read that table and count how many of those numbers you had to think about. For most shops it is two or three. The rest you have been cutting to for years, and writing them down is the whole migration for the part of the work that matters most, which is that a cabinet drawn here comes out the way your shop builds cabinets.
Getting back out again
Ask this question of every vendor you talk to, us included, and ask it before you sign rather than after. A shop that has been charged to get its own data out already knows why.
Every cut list leaves here in three shapes. A spreadsheet CSV with every part as a row. A plain parts list for the cut station, which the app calls the Cut station parts list, one row per distinct cut, sorted by material then longest first. It suits RazorGage-style saw stations and any other stop system that takes a label and a length. And a DXF of the nested sheets, which any CAD or DXF viewer opens.
What matters more than the list of formats is that they hold still. Each format is frozen: a change ships as a new format rather than a quiet edit to this one, so the mapping you set up once keeps working. The bytes are deterministic, and a golden file in our build gate fails the build if they move. That is a promise you can check rather than a reassurance you have to accept.
And the account itself downloads whole. One button on your account settings writes a single JSON file holding your projects and documents, your materials, your hardware, your offcuts, your saved settings, your layouts, and what you chose about marketing email. It is the same set that deleting your account erases, which is the only definition of a complete export worth trusting. Nobody has to approve it and nothing has to be requested.
Move one job, not the shop
The mistake is treating this as a migration project. It is not. It is one job, run twice, and the second run tells you everything a three month evaluation would have.
| 1 | Set your shop standard. The table two sections up is the list. An afternoon. |
|---|---|
| 2 | Bring your stock list in as a CSV, so the sheets are the ones you actually buy. |
| 3 | Draw one real kitchen you have already built. You know what the parts should say. |
| 4 | Put both parts lists side by side, yours and ours, and read every row that differs. |
| 5 | Cut it. A list you have not cut from is a hypothesis. |
Step four is the one that earns the decision. Every row that differs is either a setting you have not written down yet, or a real disagreement about how a cabinet is built, and both of those are worth finding on a kitchen you already know rather than on a customer's. Jobs already in production stay where they are until they ship, because there is nothing to gain by moving a job whose parts are on the rack.
Start with the part that moves
Bring your stock list across first
Drop your materials CSV in and see your own sheets, thicknesses and prices come up. It is the fastest honest test of whether any of the rest of this is true.
Frequently asked questions
What actually transfers when you switch cabinet design software?
Two things, reliably: your parts and your materials. Parts move as rows out of a spreadsheet, pasted in rather than retyped. Materials move as a CSV of your real stock, with thicknesses and sheet sizes. Everything else is a rebuild rather than a transfer, and any vendor who tells you otherwise is describing a project file format they do not have. Ask for the import list by name before you believe it.
Can new cabinet software open my old program's project files?
Ours cannot, and we would rather say so than have you find out on a Tuesday. There is no importer here for another program's native project file. Native formats are undocumented, versioned to the vendor's own release schedule and often deliberately closed, so an importer for one is a promise that breaks on somebody else's update. What we take instead is parts as pasted rows and materials as a CSV.
How long does it take to rebuild a cabinet library in new software?
For a 32mm shop it is closer to an afternoon than a quarter, because the thing you are rebuilding is a short list of decisions you already know by heart: carcass and door thickness, your reveals, how the back is held, how many cleats and how tall, shelf setback, which slides and how much clearance, and your drawer box reductions. You set those once as your shop standard and every new project starts from them. What takes longer is the cabinet types you have drawn over the years, and those come back one job at a time rather than in one sitting.
Can I get my work back out if I decide to leave?
Yes, and it is the question to ask any vendor before you sign, including us. Every cut list leaves here as a spreadsheet CSV, a plain parts list for the cut station or a DXF of the nested sheets, and each of those formats is frozen: a change ships as a new format rather than a quiet edit to this one, so the mapping you set up once keeps working. Your whole account also downloads as a single JSON file: projects, materials, hardware, offcuts and settings. Nothing about leaving requires asking us for it.
Should I move every job at once when changing cabinet software?
No. Move one live job, cut it, and compare the parts against the list your old seat produced for the same cabinets. That gives you a real answer in a day instead of an opinion in a month, and it fails cheaply if the answer is no. Jobs already in production stay where they are until they ship. Nothing is gained by moving a job whose parts are already on the rack.
Keep reading
Sources & references
There is no published standard for moving a shop between design programs, so the shape of the move here is trade practice rather than a citation. What is citable is the file formats on both ends of it, and those are below. Everything said about what CabDesign does was read out of the shipped code, with the file and line for each claim in an HTML comment beside it in this page's source. No competitor is named and no competitor's price appears, because we hold no verified published figure for the vendors this reader is most likely leaving.
- Comma-separated values (the plain tabular format both the parts paste and the materials import are built on, and why its lack of a single specification matters on an import)
- AutoCAD DXF (the neutral drawing interchange format the nested sheets leave as)
- Vendor lock-in (the general shape of the problem, and why switching cost is designed in rather than accidental)
- Data portability (the principle behind asking every vendor how you leave before you ask how you join)
- Blum, the 32mm system (the hole grid a frameless shop's construction standard is written against; manufacturer source)
A CabDesign guide. Built to help you plan and build better cabinets. Check every allowance against your own tooling and a scrap before cutting a run. Browse all guides.