# Extract Data from PDF to Excel: Fields, Not Pages

- Author: FastBankConvert (https://fastbankconvert.com)
- Published: 2026-10-16
- Category: Tables and data from PDFs
- Canonical: https://fastbankconvert.com/blog/extract-data-from-pdf-to-excel

> Extract data from PDF to Excel when you need a few fields, not whole pages: summary values, form fields, and formulas that pull them from text.
To extract data from a PDF to Excel when you need only a few values, decide first whether you want a table or single fields. Fields such as a total, a date or a reference number are best pulled from converted text with a formula that finds the label and returns what follows it.

## Tables vs single fields when extracting PDF data

Tables and single fields are different jobs, and the right method depends on which one you have. A table is many rows that share the same columns. A single field is one value with a label beside it, such as "Closing balance" or "Invoice number". Tables need a converter. Single fields often need only a search.

### How to tell which you have

Ask what you would do with the result. If you will sort, filter or sum it, you have a table. If you will copy it into a cell on a summary sheet, a report or a form, you have a field. Many documents hold both. A statement has a transaction table in the middle and a summary block at the top, and you may want one, the other or both.

### What each job needs

| | Table | Single field |
| --- | --- | --- |
| Example | A month of transactions | Opening balance, period end date |
| Shape of the result | Many rows, fixed columns | One cell per value |
| Typical tool | A converter, or a script | A search formula, or typing it |
| Main risk | Lost or merged columns | Picking the wrong number with the same label |
| Check | Row count, column total | The value against the page |

### The wrong-number risk

A label may appear more than once. "Balance" can appear in the summary, on each transaction line and in the footer. A search for the first match returns the summary figure on one statement and a different one on another, if the layout changes. Name the whole label ("Closing balance", not "Balance") and check the first few results by eye.

If your real need is a whole table, [converting just the table in a PDF to Excel](/blog/convert-pdf-table-to-excel) is the better starting point. The rest of this post is about fields.

## Pulling summary fields from a statement PDF

Summary fields sit in a block at the top of a statement or in a box on its first page. The opening balance, closing balance, statement period and account number are the usual four. To get them into Excel, convert the first page to text, then find each label and take the value beside it.

### Which fields a statement summary holds

Most statements print some of these near the top:

- The statement period, a start date and an end date.
- The opening balance, which is the balance brought forward.
- The closing balance, which is the balance at the end of the period.
- Totals for deposits and for withdrawals.
- The account number, usually with all but the last four digits hidden.

The exact labels vary by bank. One bank prints "Beginning balance" where another prints "Previous balance". Collect the labels from the statements you handle and keep them in a small list, one per row, so a formula can use them.

### Why summary fields are worth pulling

The opening balance, the movements and the closing balance have to agree. A transaction conversion with the wrong total has dropped or doubled a line, and the summary values are what let you see it. Grabbing these three numbers separately from the table gives you a check that does not depend on the table being read correctly.

### Getting the text into cells

Convert the PDF with any tool that returns the page as cells or as text, and place the result on a scratch sheet. Treat that sheet as raw material. Do not edit it; point formulas at it, so you can paste in next month's conversion and every field updates.

## Can Excel formulas extract fields from converted PDF text?

Excel formulas can extract fields from converted PDF text when the label is stable. TEXTAFTER and TEXTBEFORE find a label and return the text after or before it, and a wrapper turns the result into a number. [Microsoft's TEXTAFTER page](https://support.microsoft.com/en-us/office/textafter-function-c8db2546-5b51-416a-9690-c7e6722e90b4) lists Excel for Microsoft 365 and Excel 2024, on Windows and Mac.

### A worked example: the closing balance

This example uses made-up values. A converted statement summary block sits in column A of a scratch sheet, one printed line per cell:

| Row | A (converted text) |
| --- | --- |
| 1 | Statement period 03/01/2026 to 03/31/2026 |
| 2 | Opening balance $10,000.00 |
| 3 | Total deposits $2,500.00 |
| 4 | Total withdrawals $4,090.00 |
| 5 | Closing balance $8,410.00 |

The formula for the closing balance, with the text in A5:

`=VALUE(SUBSTITUTE(SUBSTITUTE(TEXTAFTER(A5,"Closing balance "),"$",""),",",""))`

It works from the inside out:

1. `TEXTAFTER(A5,"Closing balance ")` returns everything after the label: `$8,410.00`.
2. The first `SUBSTITUTE` removes the dollar sign.
3. The second removes the thousands comma.
4. `VALUE` converts `8410.00` into a number Excel can add.

The result is 8410, a real number. The opening balance works the same way with the label changed, and the three numbers can then be tested: opening balance plus deposits minus withdrawals should equal the closing balance (10,000 + 2,500 - 4,090 = 8,410).

### Finding the line without knowing its row

If the layout shifts and the closing balance is not always in row 5, search the whole column for the label:

`=INDEX(A:A, MATCH("Closing balance*", A:A, 0))`

The asterisk matches whatever follows the label. Wrap the result in the TEXTAFTER formula above. If the label is missing, MATCH returns an error, and that is useful: the error tells you the statement did not have the label you expected, instead of returning a wrong number.

### Dates

A period reads best by splitting on " to ". `=TEXTBEFORE(TEXTAFTER(A1,"Statement period "), " to ")` returns the start text, and wrapping it in `DATEVALUE` turns it into a date, if your regional settings read the order month-day-year. A date format your settings do not match yields an error or, worse, a date in the wrong month. Check one date against the PDF whenever a new bank shows up.

### When a formula is the wrong tool

If the label is not on the same line as the value, or the value sits in a box that the converter reads in a different order, formulas get fragile. Use the position of the cell then, or take the three numbers by hand. Fragile formulas that return a plausible but wrong number are worse than none.

## Fillable PDF forms and their data fields

A fillable PDF form stores what people typed into it as form fields, which are separate from the printed text on the page. Those entries can often be exported as data without any conversion, which makes them the easiest PDF content to get into Excel.

### How to tell a form from a flat PDF

Open the file in a PDF reader. A fillable form usually highlights its fields when you open it, and clicking a box lets you type. A flat PDF, which is a printout or a scan of a form, has no fields. Typing in it does nothing unless the reader offers a "fill and sign" overlay, and that overlay is not a field.

### Exporting form entries

Many PDF readers and editors can export the entries of a fillable form. The command usually sits in the form tools, with a name that varies by program, and the output is often a data file that Excel can open. Check your own reader's help for the exact route, because menus change between versions. For a stack of completed forms, some programs can merge the entries from several files into one table, with one row per form.

The key point is that the export reads the field values, not the page. A check box returns a value such as "Yes" or "Off", a drop-down returns its chosen option, and a date field returns the text the person typed. No reading of the layout takes place, so no layout errors can happen.

### What cannot be exported

A flat form that someone filled in and then printed to PDF has no fields. Its entries are printed text, like everything else. For those, treat each entry as a single field and use the label-based approach from the previous section, or convert the page.

## Is extracting PDF data worth it for one file?

Extracting data from one file is usually not worth building anything for. If you need three numbers from one PDF, typing them takes about a minute and builds in a check, since you read each value while you enter it. Set up an extraction only when you will repeat it.

### A rough rule

Compare two costs. The time to build the extraction is the time to set up the conversion, write the formulas, test them on two or three documents, and fix the first surprise. The time to type is the number of values, times the number of files, times a few seconds each. If you type 3 values from 1 file, build nothing. Three values from 40 files a month is 120 entries every month, each a chance to mistype, and that repeat is what a build pays back.

### What else tips the balance

- **Repeat.** A monthly job pays back a setup that a one-off will not.
- **Consistency.** Fields that sit in the same place on every document are cheap to automate. Fields that move around cost more.
- **Stakes.** When a typo is expensive, a repeatable extraction with a check is safer than retyping.
- **Volume of values.** Dozens of values per file are tedious to type and easy to mistype.

### When the job gets bigger

If the fields come from many files and you are comfortable with a little code, [pulling data out of PDFs with a Python script](/blog/extract-data-from-pdf-to-excel-python) is the next step. The details are in the section on [Python libraries for PDF table extraction](/blog/extract-data-from-pdf-to-excel-python#python-libraries-for-pdf-table-extraction). If you stay in Excel, [cleaning up converted PDF text with Power Query](/blog/transform-pdf-to-excel-power-query-cleanup) turns these formulas into steps that run again on the next file.

Where the file is a bank statement, FastBankConvert is built to return the dates, descriptions, amounts and balances as columns, so the summary check can use the same sheet. It is not open yet, and the [get-started page](/get-started) collects your email for the launch.