A flattened PDF has its form fields, comments and layers merged into the page, so nothing can be edited separately any more. Whether a flattened PDF goes to Excel cleanly depends on how it was flattened. If the text survived as text, it converts like any digital file. If flattening turned each page into a picture, it needs OCR first.
What does flattening do to a PDF file?
Flattening merges the separate parts of a PDF page into one fixed layer. Typed form entries stop being editable fields, comments and stamps become part of the page drawing, and hidden or optional layers are either merged or dropped. The page looks the same afterwards, but its pieces can no longer be changed or pulled apart.
People flatten files for good reasons. A filled-in form is flattened so the answers cannot be altered after signing. A marked-up document is flattened so the reviewer's notes stay visible in every viewer. A designer flattens a file so it prints the same everywhere. A statement or report may be flattened before it is sent, so the recipient receives a fixed record.
What flattening does not do, at least in principle, is destroy text. Merging layers and fields is a change to how the page is organized. The words and numbers can stay as real characters, now drawn as part of the page instead of sitting in a field on top of it. The trouble starts when the tool doing the flattening takes a shortcut, which the next section covers.
From a spreadsheet point of view, flattening changes three things:
- Field values join the page text. A value typed in a form field becomes ordinary text, or ordinary pixels, in the position where the field sat.
- Notes and stamps join the page. A sticky note or a "PAID" stamp is now part of the page content and may come through into your converted rows.
- Hidden content is settled. Layers that were switched off either disappear or become permanent, depending on the tool.
Two kinds of flattened PDF: vector text or page images
A flattened PDF comes in two kinds. In the first, text stays as text: characters with fonts, which you can select and search. In the second, each page was rendered to a single image, so every letter is now a pattern of pixels. Both kinds look the same on screen. Only the first kind converts directly.
The image kind usually comes from one of these routes:
- Print to an image-based printer. Some print drivers and "save as PDF" options render each page as an image before writing the new file.
- A flatten option that rasterizes. Some editors flatten by drawing the page into a picture, because that guarantees nothing can be edited.
- Scan or photo workflows. A file printed on paper and scanned back in is flattened in the most literal sense.
The vector text kind usually comes from a reader or editor that flattens form fields and annotations into the page while keeping the fonts. Typed form answers become real text in a real font, and the original printed text stays untouched.
The difference matters most for conversion speed and accuracy. A text page is read character for character, exactly as written. An image page has to be recognized optically, which takes longer and can misread digits. The trade-off between speed and accuracy in PDF conversion explains why the image route costs both time and certainty.
How can you tell whether a flattened PDF still holds text?
You can tell whether a flattened PDF still holds text with five quick tests in any PDF reader: select a line, search for a known amount, zoom in hard, check the fonts list, and compare the file size per page. Text that selects word by word and searches correctly settles it on the first two tests.
Run the tests in this order and stop when the answer is clear.
| Test | How to do it | Holds text if | Image only if |
|---|---|---|---|
| 1. Select | Drag across one transaction line | Single words highlight | The whole page highlights, or nothing does |
| 2. Search | Search for an amount you can see, such as 1,250.00 | The amount is found and highlighted | No match anywhere |
| 3. Zoom | Zoom to 400 percent or more on small print | Edges stay sharp | Edges blur or show square pixels |
| 4. Fonts | Open the document properties and look at the fonts list | Fonts are listed | The list is empty |
| 5. Size per page | Divide the file size by the page count | Small, often well under 100 KB a page | Large, often several hundred KB a page |
The example values in the table are rough guides, not rules. A text page with a large logo can be heavy, and a tiny low-resolution image page can be light. That is why tests 1 and 2 come first: they ask the file directly whether it has text.
Watch for one trap at test 2. A page can be an image with a hidden OCR text layer underneath, added by a scanner or an earlier tool. Search works, but the text came from recognition, not from the original. If zoom shows pixels while search finds the amount, you have that kind of file, and its accuracy is only as good as the OCR that made it.
Form fields and comments in a flattened statement PDF
Form fields and comments in a flattened statement PDF become part of the page, so a converter reads them as if the bank had printed them. A note written in the margin, a highlight, a stamp or a typed value can all appear as extra text in your rows. Most of the time this is noise you clean out afterwards.
Common examples on financial files:
- Reviewer notes. An accountant's comment such as "query this" beside a transaction can end up in the description column of that row.
- Stamps. A "RECEIVED" or "PAID" stamp drawn over a column can drop letters into the cells it crosses.
- Highlights. A highlight is normally just colour and adds no text, so it usually does no harm.
- Typed form values. On a form, the filled-in values are exactly the data you want, now part of the page text.
If you have both versions of the file, convert the one from before flattening. The comments then stay separate and do not pollute your cells. If you only have the flattened copy, look at the first converted page for stray words and remove them before you rely on totals.
Converting a flattened PDF that kept its text layer
A flattened PDF that kept its text layer converts like any digital PDF. The words and numbers are real characters, so a converter reads them exactly, with no recognition step. What changes is that any merged notes or stamps now come along as extra text, which you check for after conversion.
Before converting, check two things the flattening might have changed:
- Page size and margins. Some flattening routes, especially printing to a new PDF, rescale the page. A column at the right edge can be cut off. Printing a bank statement to PDF so the text survives covers the scaling settings that keep every column on the page.
- Page count. A print-based flatten can add a blank page or split a long page in two. Compare with the original.
After conversion, look at the first and last page of output. Merged stamps and notes tend to show as short words in odd cells: a single "PAID" in a date column, or a reviewer's initials after a description. A filter for text in the amount column finds most of them quickly.
FastBankConvert is built for digital statements like this: it reads each page and puts every transaction in its own row with date, description, amount and balance. You can get started with FastBankConvert and be told when trial conversions open.
Flattened PDF pages that need OCR before Excel
A flattened PDF page that became an image needs OCR before it can reach Excel. OCR, optical character recognition, looks at the shapes on the page and decides which character each one is. Only after that step does the page have text a converter can put into columns.
OCR on a flattened page is often easier than OCR on a scan. The image came straight from a digital page, so it is not skewed, smudged or shadowed. The quality depends on the resolution the flattening tool used. A page rendered at a low resolution has small, blocky characters, and those are where digits get misread.
When you OCR a flattened statement, check the results more closely than you would for a digital file:
- Compare the closing balance with a sum of the opening balance and the transactions. A misread digit usually breaks the arithmetic.
- Check dates against the statement period. A misread 8 for 3 can produce a date that looks valid.
- Search descriptions for zeros in place of letters. "0FFICE" for "OFFICE" is a typical OCR swap.
These checks are worth it because an image page gives no guarantee. A text page is exact. An image page is a careful guess.
When a flattened PDF is the wrong file to convert
A flattened PDF is the wrong file to convert when a better source is easy to get. If the flattening produced images, the original file, a fresh download, or the transaction list on the bank's website will all give more accurate data than OCR. Going back one step is often faster than cleaning up recognition errors.
Consider going back when:
- You own the account. Your bank's portal can give you the statement again, unflattened, and often a CSV export too.
- The images are low resolution. If zooming in shows blocky characters, OCR will struggle with the digits.
- You need only recent transactions. For the last few weeks, the activity page in online banking may be quicker. Copying an online banking page straight to Excel explains what that route gives you and what it leaves out.
- The file must stay as issued. If the flattened copy is the official record, keep it untouched, and use another source for the spreadsheet.
Keep the flattened file either way. It shows what the document looked like when it was sent, which is useful if a figure in your spreadsheet is ever questioned.