Floor plan PDFs are the most complete record of who's at a show and where: every stand, its number and often its size. But they're built for printing, not for spreadsheets. Here's how to get the data out.
Why copy-paste doesn't work
Text in a floor plan PDF is placed stand by stand, not in rows, so copying it produces a jumble of names and numbers. Company names are often split across lines, rotated to fit narrow stands, or abbreviated. Many plans are scans, with no text at all.
Doing it by hand takes hours per hall and mistakes are easy: a wrong booth number sends you to the wrong stand on show day.
Option 1: OCR and cleanup
Optical character recognition pulls the text out, but you still have to work out which name belongs to which booth and fix split names. It's workable for a small hall, painful for a big show.
Option 2: AI floor plan extraction
An AI floor plan reader like ExhibitionBase understands the layout: it links each company to its booth number and hall, reads stand sizes where they're shown, and scores every row so you only review the uncertain ones. Large halls are read region by region so small stands aren't missed.
- Upload the PDF (or a PNG/JPG image) of the plan
- Let the AI read every stand
- Review flagged rows, save to the exhibition and export to CSV for Excel
Tips for better results
Use the original PDF from the organiser rather than a screenshot. Vector PDFs are read more accurately than scans.
- Upload the highest-resolution version available
- If the show also has an exhibitor directory, extract it too so names are cross-checked
- Re-run it closer to the show: plans change as stands are sold