Exporting an analysis
Maps to:
POST /v1/analyses/export
The export endpoint produces a downloadable file — Excel or CSV — from an
analysis definition. This differs from the .csv execute suffix, which returns
raw CSV inline; export returns a named file ready to save.
Request
POST /v1/analyses/export
{
"meta": { "version": "v1" },
"data": {
"analysis": { "queries": [ /* … */ ] },
"format": "excel",
"force_daily": false
}
}
analysis— the analysis definition (same structure as execute, but see the normalisation notes below).format—excel(default) orcsv.force_daily— whentrue, exports at daily granularity, clearing any resampling and moving average in the definition.
Export normalises the definition
Before executing, the export service adjusts the analysis, so a complex analysis can export differently from how it executes:
- Queries with a
confidence_boundare dropped. - Every query is forced onto the first query's
periodandresample— mixed-period or mixed-resample analyses are unified. - Duplicate queries are removed.
Response
A file download: the body is the file content, with a Content-Disposition
header giving the filename and a Content-Type matching the format. Save the
bytes to disk using that filename.
Excel structure. An Excel export contains a Meta tab summarising what was downloaded (brands, audiences, regions, time period), then one tab per metric (e.g. Buzz, Index, Consideration). All exported data is weighted to a nationally representative sample.
Why there is no "export by UUID"
Unlike execute, export cannot be driven by a
saved analysis's UUID alone — it needs export-only choices such as force_daily
that are not part of the stored definition. To export a saved analysis, first
fetch its definition by UUID, then build an
export request from it.
Export vs the .csv execute suffix
| Export endpoint | .csv execute suffix | |
|---|---|---|
| Output | A named file (Excel or CSV) | Raw CSV inline |
| Formats | Excel and CSV | CSV only |
| Best for | Sharing / spreadsheet downloads | Programmatic data pulls |