Everything you need to know, in ten seconds.

Three moves, no prior setup.

• Drop — Drag your Excel or CSV file, or the link of a Google Sheets spreadsheet, into the drop zone.
• Read — The dashboard shows up right away: key figures, trend, breakdowns and source table.
• Adjust — If needed, change a colour, hide a widget or fix a column — everything is optional.

Accepted formats: .xlsx, .xls, .csv and Google Sheets spreadsheets shared by link.

The ideal format is a simple table: a single header row at the top, then one row per record (a sale, an invoice, an answer…). One column = one piece of information, with an explicit name.

Avoid merged cells, interleaved subtotals, titles above the table and empty “breathing” columns: Untable can often work around them, but a clean table always gives the best result.

Each column gets a role, detected from its name and its content.

• Dates — They drive the trend line and the “vs previous period” variations.
• Measures — Continuous numbers (amounts, quantities, rates) become key figures and charts.
• Categories — Columns with a few repeated values (region, channel, status) feed the breakdowns.
• Identifiers — Invoice numbers, emails, references: unique per row, they are never charted.

Some columns are deliberately ignored: empty or nearly empty columns, unique identifiers and free text (comments, descriptions). The “File reading” panel, below the filters, lists each column with its role: you can fix it in one click.

The same layout for every file, from the most important to the most detailed.

• KPI cards — The total (or average) of each main measure, the first one highlighted.
• Variation badge — The arrow compares the last period with the previous one; green = up, red = down.
• Trend line — The main measure over time, aggregated by day, month or quarter depending on the span.
• Bars and donut — The breakdown of a measure by category: bars for a ranking, donut for shares.
• Source table — Your data as it was read, paginated, to check a figure at a glance.

No configuration is required. Every setting is optional and reversible.

• Sum or average — On a KPI card, switch the aggregation if the detection does not match your use.
• Granularity — On the line, switch from day to week, month or quarter to smooth or detail.
• Hide — A useless widget hides in one click; it can be brought back at any time.
• Chart type — Swap bars and donut, or add a widget (goal, KPI, chart) from “Add”.

Filters, accent colour, year-on-year comparison and presentation mode sit in the same place, above the dashboard.

These situations are recognised and handled automatically.

• Several sheets — The largest sheet is selected; the others stay one click away above the dashboard.
• Total rows — “Total”, “Subtotal” or “Balance” rows at the end of the table are excluded from calculations.
• Balance columns — A balance or a running total is not summed: the last value is shown.
• Durations — Durations like “1:30” or “2 h 15” are converted to decimal hours.
• Mixed dates — French, English or ISO formats are recognised; ambiguous dates are settled by the consistency of the column.

The four most frequent causes, and their fix.

• Headers missing or on several rows — Gather the labels on a single row, right above the data.
• Several tables on the same sheet — Keep one table per sheet; the first block is read, the rest ignored.
• Empty or too small file — At least two data rows and one numeric column are needed to produce a dashboard.
• Text-only columns — Without a number or a date, there is nothing to measure: add an amount, quantity or date column.

Your data stays with you.

The file is read directly in your browser. Nothing is sent to a server to generate the dashboard. Nothing is kept after the tab is closed — only an explicit save, from an account, stores a dashboard.