Pivot Table Builder

Upload CSV or JSON data and build interactive pivot tables with drag-and-drop row, column, and value fields.

Tested tool guide Tested browser tools Checked August 16, 2026

What Pivot Table Builder does, with a checked example

Drop a CSV or JSON file on this page and its columns become draggable fields. Drag one column onto the row shelf, another onto the column shelf, and a numeric column onto the value shelf, and the tool regroups every record into a cross-tabulated grid of sums, counts, averages, or whichever aggregate you pick. No formulas, no SQL; the shelf layout is the query. The common misread: the grid is a summary, not the data. Individual rows disappear into cells, an empty cell means no records fell into that combination (not a recorded zero), and the same file yields very different tables depending on where each field is placed.

Worked example

A concrete input and expected output from the current implementation.

Input

region,product,sales
North,Widget,120
North,Gadget,80
South,Widget,90
South,Gadget,110
North,Widget,30

with rows: region, columns: product, values: sum of sales

Expected output

           Widget    Gadget    Total
North      150       80        230
South      90        110       200
Total      240       190       430

Each cell adds the sales of the records matching its region and product: North/Widget is 120 + 30 = 150, and the other three cells hold single records. The totals reconcile: 150 + 80 = 230 for North, 150 + 90 = 240 for Widgets, and 430 for the whole file (120 + 80 + 90 + 110 + 30).

How the result is produced

Parsing into fields

The file is read as plain text in the browser. The CSV header row, or the keys of a JSON array of objects, becomes the list of draggable fields, and every subsequent line or element becomes one record. Values that look numeric are kept as numbers so they can be summed or averaged; text values can only be counted or used for grouping.

Cell aggregation

The row shelf and column shelf each define a grouping. A cell at row group A and column group B holds every record whose values match both, and the chosen aggregate is applied to the value field: sum for numbers, count of records for text, or average, minimum, maximum where they make sense. Moving a field between shelves rebuilds the whole grid from the original records, so no subtotal ever goes stale.

Good uses

Summarize a month of sales rows into a region-by-product grid to see which combinations drive revenue.
Count survey responses by demographic and answer to find patterns without hand-building spreadsheet pivot formulas.
Cross-tabulate request logs by status code and endpoint to compare error rates across service areas.

Limits and checks

The grid shows aggregates only: twenty source records can collapse into one cell, so outliers and per-record detail are invisible until you change the layout or export the raw rows.
The number in a cell is what the chosen aggregate says, nothing more: SUM of a numeric column is common, but a text column dragged into values produces record counts, and an average is not a total. Check which aggregate is selected before quoting a figure.
An empty cell means no records matched that row and column combination, not a stored zero. If the tool shows blanks instead of zeros, the table understates presence unless you account for it.

Common questions

Can I count how many records fall into each cell instead of summing numbers?

Yes. Drag any field - a text field if you have one - onto the value shelf and select the count aggregate. If every field is numeric and the tool defaults to sum, it may quietly add instead of counting, so switch the aggregate explicitly.

Does uploading my data send it anywhere?

No. The file is read and processed entirely in your browser; nothing is transmitted to a server. That also means your data never leaves the device you opened the page on, and closing the tab discards it, so keep a copy of the source file if you want to return to it later.

References and verification

The example and behavioral notes were checked against the browser implementation. Standards and primary references below define the relevant format, formula, or platform behavior.