Why PDF to Word conversions mess up formatting, and how to fix it

Why a PDF fights back

A PDF is built to look the same everywhere, not to be edited. Microsoft explains it plainly: a PDF stores where text, pictures and graphics sit on the page, but not necessarily how they relate to each other. It doesn't record that two blocks are columns, that a set of lines is a table, or that small text at the bottom is a footnote.

So when you open a PDF in Word, Word applies a set of rules to guess which parts are headings, lists, tables and paragraphs. When a guess is wrong, the formatting breaks. Knowing which guess went wrong tells you which fix to use.

The five usual causes

1. The PDF has little or no structure (tagged vs untagged)

Some PDFs carry an extra, hidden layer called tags. Tags mark content as a heading, a paragraph, a list or a table and set a reading order. The W3C's accessibility guidance describes headings marked with heading tags in the PDF's structure tree, usually added by the program that made the file (for example, Word's heading styles exported to PDF).

Many PDFs have no tags at all, and every scan starts without them. Even when tags exist, Microsoft lists "bookmarks, tags and comments" among the things that don't convert well into Word, so don't expect a tagged file to convert perfectly. What helps far more is real text: a PDF exported from a word processor usually converts more cleanly than a scanned page. If yours is a scan, start with how to turn a scanned PDF into an editable Word document.

2. Multi-column layouts

On a two-column page, a converter has to decide whether to read across the page or down each column. When it guesses wrong, lines from the left column get mixed with lines from the right. A milder version: the columns survive, but as text boxes or frames that move around when you type. Microsoft lists frames among elements that convert poorly.

3. Borderless tables

A table with visible gridlines is easy to spot. A price list laid out with spaces and no borders is not. It often comes through as separate lines with spaces or tabs between the values, or as a table with the wrong number of columns. Microsoft also notes that tables with cell spacing are a known problem.

4. Font substitution

If the PDF used a font that isn't on your computer, Word shows a different font in its place. A substitute that is slightly wider or narrower changes where lines break. That pushes paragraphs onto the next page, squeezes table cells and makes a one-page form spill onto two pages. Special font effects, such as glow or shadow, may turn into graphics instead of editable text.

5. Headers, footers and footnotes

In a PDF, a page header is just text near the top of each page. If Word doesn't recognize it as a header, the same company name or page number repeats in the body every few paragraphs. Microsoft notes that an undetected footnote can become regular text, and that footnotes spanning more than one page and endnotes convert poorly.

Fixes inside Word

Work on a copy, keep the original PDF open beside it, and fix structure before you fix appearance.

Reset messy formatting

1. Select the affected text (or the whole document with Ctrl+A or Command+A).

2. On the Home tab, select Clear All Formatting in the Font group.

Microsoft notes this only removes formatting applied by hand. Once text is plain again, you can rebuild it cleanly.

Rebuild headings with styles

Select each heading and choose Heading 1, Heading 2 and so on from the Styles gallery on the Home tab. Styles keep headings consistent, and changing a style later updates every heading that uses it. Styled headings also make the document easier to find your way around and more accessible.

Turn loose lines back into a table

When a borderless table arrives as separate lines:

1. On the Home tab, turn on Show/Hide so you can see tabs and paragraph marks.

2. Make sure each row is one paragraph and the values are separated by a single tab (or a comma, if the values contain no commas).

3. Select the lines, then go to Insert, then Table, then Convert Text to Table.

4. Set the number of columns, choose the separator you used and select OK.

Then check every number against the original. A missed tab shifts a value into the wrong column.

Sample (synthetic, from "Northgate Supplies" price list): "Toner cartridge 4 89.00 356.00" became one cell. After adding tabs between the four values, Convert Text to Table produced a correct four-column row.

Fix columns

If the converter jumbled two columns into one, it is often quicker to paste the text as one column, put it in the right reading order, then reapply columns. On the Layout tab, select Columns and choose two (or More Columns for widths and spacing). Select text first to apply columns to just that part.

Replace the substitute font

Select all the text and apply one font that you have installed, then adjust size and spacing to match the original. Applying fonts through styles (for example, changing the Normal style) is faster than reformatting paragraph by paragraph.

Move headers and footers back where they belong

1. Delete the repeated header or footer text from the body.

2. Double-click the top or bottom margin, or go to Insert, choose Header or Footer, then Edit Header or Edit Footer.

3. Type the text once, add page numbers there, and select Close Header and Footer (or press Esc).

Footnotes

If footnotes landed in the body, cut each one, click where its number belongs and reinsert it with References, then Insert Footnote, so numbering updates automatically.

When fixing by hand isn't worth it

A few broken lines are quick to repair. Rebuilding every table and column on a twenty-page scan is not. If the layout matters and the page count is high, a tool that reconstructs layout before writing the Word file can save time.

A final check before you send it

• Page count and page breaks roughly match the original.

• Every table has the right number of columns and rows.

• Numbers, dates, totals and names match the PDF exactly.

• Headings use styles, not just bold text.

• Headers, footers and footnotes are in their proper places, not in the body.

Skip the rebuild

EasyToDecode is launching soon with a scan-to-Word converter for scanned PDFs and photos. It will produce a .docx that keeps the same columns, tables, headings and page breaks, with real text you can edit. See how it works or join the waitlist to hear when it opens.