VKT Toolkit · August 28, 2026 · 5 min read

How to Handle Merged Cells When Exporting Web Tables (rowspan/colspan Guide)

Government reports, financial statements, school timetables — the most useful data tables on the web are also the most complex. They use merged cells to group categories, span headers across columns, or create multi-level layouts. And when you try to copy them to Excel, everything falls apart.

This guide explains why merged cells break, what rowspan and colspan actually are, and how to export them correctly with zero manual cleanup.

What Are rowspan and colspan?

In HTML, tables are built with <tr> (rows), <td> (data cells), and <th> (header cells). Two attributes control cell merging:

• rowspan="3" — this cell spans 3 rows vertically;
• colspan="2" — this cell spans 2 columns horizontally.

Here's a typical example from a government budget table:

The header "Department" spans 2 rows; "Budget (USD)" spans 2 columns. This creates a clean hierarchy — but it's exactly what breaks on copy-paste.

Why Merged Cells Break on Copy-Paste

When you select this table in a browser and Ctrl+C, the browser sends raw HTML to the clipboard. Excel receives it and tries to map HTML cells to spreadsheet cells. The problem:

• rowspan cells — Excel may only paste the value in the first row, leaving the next row's first cell empty. Your "Education" label disappears from the second data row;
• colspan cells — Excel may push the merged header into one cell but leave the adjacent cells blank, shifting all subsequent columns to the right;
• Multi-level headers — tables with 2-3 header rows using both rowspan and colspan produce completely garbled output;
• Nested tables — some sites put tables inside table cells. Copy-paste treats them as one flat table, destroying the structure.

Result: 30 minutes of manual cleanup for a table that took 2 seconds to display.

The Fix: Read the HTML Structure, Not the Rendered Copy

The fundamental problem is that copy-paste operates on the rendered output. To get merged cells right, you need a tool that reads the underlying HTML attributes — the actual rowspan and colspan values — and reconstructs the grid mathematically before exporting.

That's exactly what VKT TableGrab does:

1. It scans the page for all <table> elements;
2. For each cell, it reads rowspan and colspan attributes;
3. It builds a 2D grid, filling merged values across all spanned cells;
4. It exports the grid as a properly structured CSV.

The result: "Education" appears in every row it should, "Budget (USD)" spans correctly as a header, and all columns align.

How to Use TableGrab for Merged-Cell Tables

1. Install from the Chrome Web Store or Edge Add-ons;
2. Open the page with the complex table;
3. Click the TableGrab icon — it auto-detects all tables;
4. Hover over a table in the list to highlight it on the page;
5. Click Export — the CSV downloads with merged cells properly reconstructed.

No DevTools, no scripts, no manual fixing. The free tier handles up to 5 tables per page — more than enough for most reports.

FAQ

More guides in the VKT blog; questions to [email protected].