Why your Drake import put the cost in the wrong field and never said so

The import ran. No error, no warning, no rejected rows. Drake reported the assets as imported and
you moved on.

Three weeks later the depreciation on one activity is nonsense, and when you open the asset detail
the cost is blank and the business-use percentage reads 12,500%.

Nothing went wrong during the import. Drake did exactly what it was asked.

The 4562 import is positional, and that is the entire contract

Drake's "Import Assets from a Spreadsheet" reads a workbook of 104 columns, running from A to
CZ, and it identifies each field by position. Column C is the description. Column D is the
acquisition date. Column E is the cost. Column F is the business-use percentage.

The header row is decorative. Drake begins reading at row 2 and does not consult row 1 to work
out what anything is — the labels are there so a human can eyeball the file, and nothing more.

That design is not unusual for a bulk import, and it is fast and unambiguous when the file is
right. It has one consequence that catches people, and the consequence is not in the documentation.

What a single inserted column does

Suppose the spreadsheet you built started life as an export from somewhere else, and at some point
a column was inserted — a client code, a location, an "old asset number" someone wanted to keep.
Insert it left of column D and everything from D onward slides one place to the right.

Now:

• Drake reads column E expecting cost, and finds your acquisition date
• Drake reads column F expecting business-use percentage, and finds your cost

Insert it one place further right, between D and E, and the pairing changes but the outcome does
not: Drake reads E and finds the client code you inserted, and F still receives the cost.

A cost of 12,500 arrives as a business-use percentage of 12,500. That is the 12,500% above.

Drake does not complain, and there is nothing for it to complain about. It asked for a number in
column F and it received a number. No validation rule that reads one cell at a time can tell the
difference between a business-use percentage of 100 and a cost of 100. The information that would
reveal the error — that column F contains costs across every row, that column E contains dates —
only exists when you look down the column instead of across the row.

This is why the failure is silent rather than loud, and why it surfaces weeks later in a number
nobody was auditing.

Dates are text, and the format is MMDDYYYY

The second failure has the same shape: valid input, wrong meaning.

Drake documents its date columns — acquisition date in D, in-service date in Y, sale date in CC —
as text in MMDDYYYY. Not 1/15/2024. Not 2024-01-15. 01152024, formatted as Text.

A real Excel date cell will often import. The problem is that an Excel date is stored as a serial
number and rendered according to the machine's regional settings, so the same file can read one way
on the desk that built it and another way on the desk that imports it. Day and month swap silently
between locales, and 01/06/2024 becomes a different date without a single character changing.

Writing the digits as text removes the question entirely. There is nothing left to interpret.

Row 1 is never read

Because the import starts at row 2, a file with no header row loses its first asset. Not an error
— the asset simply is not there, and a schedule of 84 assets imports 83.

The fix is a header row, even a nonsense one. What matters is that row 1 is not an asset.

Two things that also bite, more visibly

Currency formatting in money columns. A cell holding $12,500.00 as text may import as zero.
Drake wants a plain number; strip the symbol and the thousands separator.

Columns past CZ. The template ends at column 104. Anything beyond it is ignored, which is
usually harmless — unless a column was inserted mid-row and pushed real data off the end.

Check the file before Drake does

We built a free checker for exactly this, because the failure is invisible by design and the fix is
trivial once you know which column moved.

Check a Drake import file — pick Drake, upload the workbook, and it
reports whether description, date and cost sit where Drake will look for them, whether the date
cells are text or Excel dates, and whether the money columns carry symbols. It works out the
direction and size of a shift when it finds one, so you get "columns look shifted 1 to the right"
rather than "something is wrong".

No account, and the file is read in memory and discarded — nothing is stored, nothing is written to
disk, nothing goes to an AI model. It checks the shape of the file and makes no statement about
whether any row's tax treatment is right; that judgment is yours and the tool does not go near it.

If you would rather not build the spreadsheet at all, converting a prior preparer's schedule
straight into the import file is what this site does — see
Drake 4562 import from a PDF.

What "it imported fine" actually means

It means Drake found a value of the expected type in each position it looked at. It does not mean
the values were the ones you intended, because Drake has no way to know what you intended.

A positional format moves that responsibility upstream. The only place the error is visible is in
the file, before the import — which is the one moment nobody thinks to look.

FAQ

Why doesn't Drake warn me when a column is shifted?

Because a shift produces valid input. Drake asks for a number in column F and receives a number; a
per-cell validation rule cannot distinguish a business-use percentage of 100 from a cost of 100.
The pattern is only visible by reading down the column across many rows.

What date format does Drake's 4562 import expect?

Text in MMDDYYYY with no separators — 01152024. An Excel date cell often imports, but it is
stored as a serial number and rendered by the machine's regional settings, so the same file can be
read differently on a different desk.

Does the header row matter for Drake's import?

Not for identifying fields — the format is positional and the labels are ignored. It matters
because the import begins at row 2, so a file whose first row is an asset loses that asset.

How many columns does the Drake 4562 template have?

104, running A to CZ. Unused columns are left empty rather than removed: taking one out shifts
every field after it.

Can I check a file without uploading it anywhere?

The checker on this site reads the file in memory and discards it — it is never written to disk or
stored. Short of that, the manual check is to open the workbook, confirm your description sits in
column C, your acquisition date in D and your cost in E, and confirm the date cells are formatted
as Text rather than Date.