Skip to content
Variable Data PDFA PDFIMPOSE FIELD GUIDE
The workflowField notesQuestions
Open PDFImpose
Variable Data PDF / FIELD NOTES
How-to

Variable Data PDF Inspection: How to Proof Before Printing

Proof a variable data PDF in two layers. First check every record by counting: data rows, merged pages and imposed sheets must reconcile, and sequence numbers must pass count, duplicate and sum checks. Then inspect a chosen sample: first, last, longest value, empty optional fields, unusual characters, pile boundaries and random records. Add checks for overset text, font substitution, missing images and printed barcode scans, re-check order after imposition, and record a named sign-off.
Distinct personalized print pieces arranged beside a merged mailing print sheet.
Workflow illustrationVariable Data PDF

A static job has one design to approve. A variable data job has as many designs as it has records, and the proof you were shown is only one of them. The problems that reach the press are almost never in the attractive sample record. They sit in record 3,417, where a surname is 34 characters long, an image link is missing or a ticket number repeats.

Inspection therefore has two halves: checks that cover every record by counting and comparing, and visual checks of a deliberately chosen sample. Do both on the merged PDF before it is imposed, then repeat the order checks on the imposed file. The data side starts earlier; the data file preparation guide covers encoding, delimiters and field lengths.

Do the record counts reconcile?

Write down four numbers and make sure they agree with the job's structure:

CheckpointExample: 2,500 two-sided postcardsHow to get it
Data rows (excluding header)2,500Row count in the spreadsheet or CSV
Pages per record2From the template: front and back
Merged PDF pages5,000Page count in the PDF viewer
Imposed sheets at 4-up duplexceiling(2,500 ÷ 4) = 625Page count of the imposed file ÷ 2

If the merged PDF has 4,998 pages, a record was dropped or a filter excluded it. If it has 5,002, a blank row or a stray header became a record. Rows filtered out deliberately (for example, suppressed addresses) should be counted and documented so the difference is explained rather than discovered.

Which records should you spot check?

Choose the sample on purpose. A useful minimum set:

  • First and last record. Off-by-one errors and dropped final rows show up here.
  • The longest value in each variable field. Sort a copy of the data by text length to find it, then find that record in the PDF.
  • Records with optional fields empty. A missing second address line should close up, not leave a gap or the word "null".
  • Unusual characters. Accents, apostrophes, ampersands and non-Latin scripts.
  • Records at pile boundaries. For cut and stack at 4-up, 1,000 items give piles of 250, so check 250, 251, 500, 501 and 750, 751.
  • A random sample. Pick 20 or so record numbers with a spreadsheet random function, so the sample is not just the ones you expected to look at.

For each sampled record, compare the PDF page against its data row field by field, not just by eye.

How do you check for duplicate or missing numbers?

Sequential numbers can be tested arithmetically. For a run from a to b:

  • Expected count = b − a + 1. Tickets 1001 to 3500 should be 2,500 rows.
  • Distinct count must equal the row count. A spreadsheet COUNTIF or a pivot shows any value that appears twice.
  • Sum check for 1 to n is n(n + 1) ÷ 2. For 1 to 2,500 that is 3,126,250. A gap combined with a duplicate can keep the count right but changes the sum.

Then check the PDF itself, since the data can be right and the merge wrong. Extract text from the merged PDF and search for the first, last and a few middle numbers, confirming each appears exactly as often as it should (twice if each ticket has a matching stub). The numbered raffle ticket guide covers prefixes, padding and stub matching.

How do you catch overset text, font fallbacks and missing images?

Overset text is the silent failure: a long value overflows its frame and simply disappears, so the page looks fine but the last word is missing. Use your merge tool's overset report if it has one (InDesign, for example, offers one when creating the merged document), and check the longest-value records visually.

Font fallbacks appear when a character in the data is not in the chosen font. Depending on the application you get a substituted font, an empty box or a missing glyph. In the PDF, open the document's font list and confirm only the expected fonts appear, all embedded. An unexpected extra font usually means a substitution somewhere in the data. The mail merge to print-ready PDF guide covers font embedding checks.

Variable images fail when a file name in the data does not match a file on disk. Count image references in the data, count distinct files in the image folder and check that every referenced name exists, matching case, before merging. Then sample pages for placeholder boxes or low-resolution previews.

How should you verify barcodes and QR codes?

Do not trust the on-screen image. Print a sample on the real stock at actual size and scan it with the kind of device that will read it in the field.

  1. Scan the first, last and several random codes. The decoded value must match the data row exactly, including leading zeros and prefixes.
  2. Check that each code matches the human-readable number printed next to it.
  3. Check the quiet zone. QR codes need a clear border of four modules; trims, perforations and nearby artwork often eat into it.
  4. Check size and contrast. A code reduced to fit a smaller ticket, or printed on dark or textured stock, may stop scanning reliably.
  5. For retail symbols, ask whether the buyer requires formal verification grading with a verifier, which is more than a successful phone scan.

A code that scans proves only that it carries data. Whether that data is valid, unique or already used depends on the system reading it.

What should you check after imposition?

Imposition is where a correct merged PDF can still be put out of order. Re-run order checks on the imposed file:

  • Grid order: for ordinary 4-up, sheet 1 holds records 1 to 4. For cut and stack, 1,000 records at 4-up give 250 sheets, sheet 1 holds 1, 251, 501, 751 and the last sheet holds 250, 500, 750, 1,000.
  • Front and back pairing: on duplex jobs, record 37's back must sit behind record 37's front. Hold a printed sheet to the light or check numbers on both sides.
  • Empty positions: 1,002 records at 4-up leave two empty positions on the final sheet. Note them on the cut plan.

In PDFImpose, a checked merged PDF can be laid out as a grid or as cut and stack. The app arranges the merged pages; it does not merge spreadsheets or generate barcodes, so those checks stay with the merge tool. The numbered tickets cut-and-stack guide walks through pile order in detail.

Will the file size slow the press?

Variable data files grow with the record count, and the way they are built matters more than the content. If the static background is stored once and reused on every page, each record adds little. If every page carries its own flattened copy of the background, the file grows by that image for every record: 10,000 records × 2 MB is about 20 GB, which can choke a RIP. Check the file size against the record count, and if it grows in proportion to a full image per page, ask the merge tool for reusable artwork or a PDF/VT export where the press supports it. Large files are also slower to impose in a browser.

Variable data proof sign-off checklist

  • Data rows, merged pages and imposed sheets reconcile, with any suppressed records documented.
  • Sequence numbers pass count, distinct and sum checks, in the data and in the PDF text.
  • First, last, longest, empty-field, unusual-character, pile-boundary and random records match their rows.
  • No overset text, unexpected fonts or missing images.
  • Printed codes scan, match the data and the human-readable text, and keep their quiet zone.
  • Imposed order, duplex pairing and empty positions match the cut plan.
  • File size is proportionate and the printer has accepted the format.
  • One named person signs off with the date and the file name of the approved version.

For the wider workflow, see the variable data printing guide or the Variable Data PDF overview.

Common questions

How many records should I check visually?

There is no fixed number. Cover the cases most likely to fail, such as the first, last, longest, empty-field and boundary records, then add a random sample of about 20. Counting checks cover the rest, since no one can read every page of a large run.

Can I proof a variable data job from one sample PDF?

No. A single sample shows the design works for one record. It cannot show long values, missing data, duplicates or dropped records. Ask for the full merged PDF, or at least the record count and a chosen set of difficult records.

What is the fastest way to find duplicate numbers?

Compare the row count with the count of distinct values in the number column. If they differ, a pivot table or COUNTIF greater than one shows which values repeat. Repeat the check on text extracted from the merged PDF.

Should I proof before or after imposition?

Both. Content checks belong on the merged PDF, where one page is one record and problems are easy to trace. Order, duplex pairing and empty positions can only be checked on the imposed file that goes to press.

Does PDFImpose check my data merge?

No. PDFImpose arranges an already merged PDF into a grid or cut-and-stack layout. Record counts, overset text, fonts and barcode content need to be checked in the merge tool and on printed samples before imposition.

BEFORE YOU CHOOSE

Does PDFImpose fit this job?

Bring a completed, checked data-merged PDF into PDFImpose for sheet layout and cutting order.

What to verify

Create the merge in a suitable personalization tool first. Check missing fields, long names, matching ticket stubs and record boundaries before imposition.

Prepare the data before imposition →

Account and export terms

Google sign-in is required. The free plan includes five lifetime exports, with eight hours between free exports. Pro is US$14.99/month or US$149.99/year, excluding tax.

View current PDFImpose plans →

This is a product-operated field guide in the PDFImpose network. Confirm the exported result and current terms before paying.